Abstract
This article looks to verbal valency analysis to help clarify the meaning of the difficult clause ובאו חמדת כל הגוים in Hag. 2.7b, and so the verse as a whole. Attention to the verb’s valency patterns (i.e., בוא [qal]) presents a fresh interpretive option that is both contextually fitting and (unlike previous proposals) linguistically robust. These patterns suggest that the cryptic phrase כל חמדת הגוים is not the subject (so most) but a descriptor for the Jerusalem temple naming the locative goal of the clause’s motion verb. This analysis identifies כל הגוים of the previous clause as the subject, yielding the following rendering of the verse: ‘And I (Yahweh) will shake all the nations so that they will come to “the desire of all the nations” (i.e., the temple), in order that I may fill this house with glory, declares the Lord of Hosts’.
In a previous article I showed how attention to verbal valence patterns may help clarify the ambiguous semantics and syntax of the difficult Ps. 73.24b. 1 Here I bring the same approach to bear upon another crux interpretum in the Hebrew Bible, Hag. 2.7b (ובאו חמדת כל הגוים).
Previous approaches to the Masoretic Text of Haggai 2.7b
The ambiguous syntax and semantics of the noun phrase חמדת כל הגוים have led to competing interpretations of the MT of Hag. 2.7b. 2 The majority view among modern interpreters is that the NP refers to the ‘treasures/precious things of all the nations’ and functions syntactically as the clause’s S (‘The treasures/precious things of all the nations shall come … ’). Most holding this view retain the singular form of the NP’s head noun in the MT (חֶמְדַּת) and, given the clause’s plural V (באו), interpret it collectively (‘treasures/precious things’). 3 Others overcome the lack of concord between the singular S and plural V by revocalizing the MT’s singular חֶמְדַּת to the plural חֲמֻדֹת, an attested derivative of the cognate verb חמד (see Gen. 27.15; Dan. 9.23; 10.3, 11, 19; 11.38, 43; Ezra 8.27; 1 Chron. 20.25). 4
A messianic reading of the singular חֶמְדַּת has been popular in the history of interpretation. 5 In contrast to the previous view, this approach interprets the NP that חֶמְדַּת heads (חמדת כל הגוים) personally (i.e., as referring to the messiah, not treasures). Most taking a messianic approach analyze the NP as the S of the clause’s plural V. Proponents of this view have overcome the resultant lack of grammatical concord by arguing that the plural V (באו) is accommodating the plural nomen rectum הגוים (‘the nations’). 6 The messianic approach reflected in the NKJV sidesteps this grammatical issue by interpreting חמדת כל הגוים instead as a NP-ADV that names where the S is headed: ‘And they [all the nations] shall come to the Desire of All Nations (i.e., the Messiah)’. 7 On this view, the S is the DO of the previous clause, i.e., כל הגוים (‘all the nations’).
A final interpretation resembles the above views but also has its own distinctive nuance. It agrees with the majority that the NP כל הגוים חמדת refers to ‘the treasures of all the nations’ but follows the NKJV in identifying כל הגוים (‘all the nations’) of Hag. 2.7a as the S. Parting ways with all other views, it takes the NP כל חמדת הגוים as a NP-ADV that specifies details of accompaniment: ‘And they shall come with the treasures of all the nations’. 8
Evaluation of previous proposals
Previous approaches to the NP חמדת כל הגוים in Hag. 2.7b (and so the clause as a whole) are problematic mainly because, whatever their contextual merits may be, they are not satisfying from a linguistic standpoint. The most obvious example is the view that the noun חמדה has a singular personal referent (i.e., the messiah). Notwithstanding attempts to defend its grammaticality, the resultant lack of concord between the clause’s S (singular) and V (plural) on this view is grammatically indefensible. As Koopmans points out (citing Keil), the suggestion that the clause’s plural V is accommodating the plural הגוים (‘the nations’) is not workable here. Such accommodation occurs only when, unlike Hag. 2.7b, ‘the governed noun contains the principal idea, so that there is a constructio ad sensum’. 9
Other interpretations are unappealing linguistically because they do not align with the valency patterns associated with the clause’s verb (בוא, qal) in Classical Hebrew. 10 Verbal valency theory highlights the close relationship between verbal semantics and syntax. 11 As Dyk, Glanz, and Oosting observe, a verb’s meaning is generally ‘portrayed through, expressed in, and carried by the formal patterns in which it occurs’. 12 Verbs therefore can (and most often do) require a distinctive number and nature of constituents to complete their meaning in any one of their different senses. 13
We can use the verb סמך to briefly illustrate this property of verbs. This verb requires the following pattern when it means ‘to lay something on someone’: (1) a NP naming the S; (2) a NP complement; and (3) a PP headed by על: ויסמכו ידיהם עליהם ‘And 1
But we find a different pattern associated with סמך when the verb means ‘to support someone/something’: (1) a NP naming the S; and (2) a complement naming the person or thing supported: היא סמכתהו ‘1
The single example of סמך illustrates the importance of verbal valency analysis for biblical interpretation. Interpreters are not free to construe verbal meaning and/or clausal syntax as they see fit, no matter how contextually attractive a given interpretation appears. Provided the necessary data are available (as in the case of בוא [qal]), 14 a verb’s valency patterns are an important linguistic control that helps define the range of viable interpretive options for a given clause. As such, they can help clarify or negotiate the meaning of difficult, ambiguous, or disputed clauses/clausal constituents. 15 Here I use valency analysis as a tool for assessing previous approaches. I look to it later for new light on Hag. 2.7b.
The NKJV rightly recognizes that בוא (qal) frequently governs a NP-ADV that specifies the locative goal of the verbal action (‘And they shall come [ובאו] to the Desire of All Nations’ [ חמדת כל הגוים] [NP-ADV]). 16 But to identify a single person as the locative goal (so NKJV), בוא (qal) invariably uses an oblique complement, most often a PP headed by אל or ל. 17 In the sole instance where a NP-ADV does name a person as the locative goal, a group of people are in view (Ezek. 38.11). This pattern for בוא (qal) is consonant with motion verbs more broadly. In their study of the valency patterns for this verb class, Oosting and Dyk ‘found no examples of a NP referring to an individual as the complement of a verb of motion’. 18 Neither do the data support taking the NP הגוים כל חמדת as furnishing details of accompaniment. The verb בוא (qal) consistently selects a PP headed by ב for this semantic function, not a NP-ADV. 19
The remaining view is the majority approach, namely, that חמדת כל הגוים refers to the ‘treasures/precious things of all the nations’ and functions syntactically as the clause’s S. The issue here is not that the proposal conflicts with the verb’s valency patterns. Its problems lie elsewhere.
Above we saw that some in this camp overcome the resultant lack of concord between a singular S and plural V by revocalizing the MT’s singular form (חֶמְדַּת) to the plural חֲמֻדֹת. This approach should be rejected for at least two reasons. First, no extant Hebrew manuscript has a plural form here. Second, as James Barr points out, while its emendation may sometimes be required, ‘the vocalization is historical evidence just as other aspects of the text are; it has to be explained not merely altered’. 20 In the present case, the ‘solution’ of emendation does not represent a serious attempt to explain the singular form in the text; it merely ‘corrects’ the vocalization to align the form with one possible interpretation of the grammar, namely, that חמדת כל הגוים is the S of the V. Below I show that the singular form can be explained when חמדת כל הגוים is read as a NP-ADV complement.
This leaves the collective interpretation. Notwithstanding its widespread acceptance, the linguistic argument for this view is surprisingly weak. Plural forms of the verb בוא (qal) do take a collective singular S elsewhere. 21 But it is not at all clear that the singular noun חֶמְדַּת is being used this way in Hag. 2.7b. For one thing, my research did not yield a single example where חמדה is unambiguously a collective noun. 22 On the other hand, when a singular S of a plural form of בוא (qal) is obviously a collective, that noun’s collective usage is clearly established. 23 These data points alone caution against too quickly selecting—and especially against simply assuming—a collective interpretation for חֶמְדַּת in Hag. 2.7b. 24
Another relevant caution arises from the research of scholars studying collectives in Classical Hebrew: interpreters should avoid generalizations about ‘collectives’, treating them as an undifferentiated category. 25 As Ian Young observes, ‘each collective tends to have its own profile in regard to grammatical agreement, and may even have different patterns of agreement with different linguistic elements. … ’ 26 The lack of concrete examples for a collective usage of חמדה shows the relevance of Young’s point here: it is methodologically unsound to read חֶמְדַּת in Hag. 2.7b collectively by analogy to singular nouns whose collective usage is clearly established and for which we can sketch a profile—such is precisely the data we lack in the case of חמדה. 27
Despite its widespread acceptance, then, the available linguistic data do not support the predominant view. Simply put, it is difficult to mount a compelling positive linguistic case for interpreting חֶמְדַּת collectively in Hag. 2.7b, and consequently for reading the NP it heads (חמדת כל הגוים) as the S of the clause’s plural V. Even the literature espousing this view bears this point out. Instead of arguments and evidence for a collective interpretation, one largely finds claims and assertions that this usage is, or must be, in view here. 28 This situation underscores that the interpretation’s popularity is not due to its intrinsic linguistic merits. Rather, it is due largely to the assumption that חֶמְדַּת is the S of the clause, and making this assumption the starting point of one’s analysis. 29 Working from this premise, the interpreter must choose one of two options given the clause’s plural V, both of which we have seen are less than satisfactory: assign חֶמְדַּת an unattested collective sense or conjecturally emend the singular חֶמְדַּת to the plural חֲמֻדֹת. Below I suggest that a better approach is to reassess the noun’s interpretation altogether (and so the NP it heads) based upon the valency patterns associated with the clause’s V.
All of this is not to say that the collective approach cannot be correct. The available data do not warrant such a conclusive verdict; חמדה is not a particularly high frequency noun in Classical Hebrew, and we must always keep in mind that we are dealing with an incomplete corpus. But it is to say: (1) based upon the available linguistic data, the majority view is far less appealing than its widespread acceptance suggests, and (2) this state of affairs should lead interpreters to ask whether a more linguistically satisfying option exists.
To summarize: previous views may be contextually workable, but they are linguistically wanting. A compelling interpretation of Hag. 2.7b will satisfy both criteria. I would suggest that a promising way forward is a method that centers interpretation on the valency patterns associated with the clause’s verb (בוא [qal]).
A fresh appraisal in light of verbal valency patterns
I take as my starting point the problems with reading חמדת כל הגוים as the clause’s S (discussed above). The alternative is that the NP is a complement or adjunct that defines the verbal action in some way. On this reading, the S is the plural כל הגוים (‘all the nations’) of the previous clause; no other candidate exists. 30 This construal of the syntax establishes the following clause pattern for Hag. 2.7b: בוא (qal) &0x002B; NP-S (כל הגוים) &0x002B; NP-ADV (חמדת כל הגוים). We can then search the available data for other clauses having this formal pattern—ones whose meanings are unambiguous—for potential insight into the NP-ADV in Hag. 2.7b (חמדת כל הגוים) and, as a consequence, the clause’s meaning as a whole.
An exhaustive analysis of בוא (qal) shows that the NP-ADV in such clauses functions in one of four ways: to specify details of (a) time (e.g., Gen. 19.5; 24.42; Num. 22.20; Judg. 7.19; 1 Sam. 26.7; 2 Sam. 19.21; 2 Kgs. 13.20); (b) manner (e.g., Gen. 24.35; 1 Sam. 16.4; 1 Kgs. 2.13); (c) cause or reason (Isa. 7.25); or (d) location. This last category further divides into: (d1) locative means (i.e., the means by which verbal motion is carried out) (‘by way of’) (e.g., Num. 21.1; 2 Kgs. 11.16)—the noun דרך (‘way, road’) heads the NP-ADV in all clear instances of this usage 31 —and (d2) locative goal (i.e., the place where motion is headed). 32
The nature of the NP-ADV in Hag. 2.7b (חמדת כל הגוים) immediately rules out (a) and (d1) as potential options; it is neither a temporal expression (excluding option [a]) nor headed by the noun דרך (‘way, road’) (excluding option [d1]). 33 The NP-ADV could potentially qualify for category (c). Elsewhere a NP-ADV headed by an abstract noun does specify details of cause or reason (יראת שׁמיר ושׁית [‘because of fear of briar and thorn’], Isa. 7.25). 34 The meaning of Hag. 2.7b would then be: ‘And they [all the nations] shall come because of the desire of all the nations (חמדת כל הגוים)’. But this reading does not comport with the literary context. Regarding option (b), the noun חמדה could theoretically describe how (manner) the verbal action is carried out (‘desirably’; see ‘peacefully’ [שׁלום] in 1 Kgs. 2.13). This meaning, however, is unworkable for חמדה as the head of the NP חמדת כל הגוים.
The final option remaining is (d2): the NP-ADV names the locative goal (impersonal, not personal; contra NKJV) of the verbal action (i.e., where the S is headed), as in the following examples: יבא יואב ירושׁלם ‘And 1 ובאו ציון ברנה ‘And 1 ויבא בית יהוה ‘And 1
In each of these clauses the verb בוא (qal) occurs in the same bivalent pattern: (1) NP-S; 35 (2) NP complement naming where the S is headed (locative goal). This option deserves serious consideration since it accounts for the vast majority of all NP-ADV’s governed by בוא (qal); the overwhelming number name the locative goal of the verbal action. Considering only the data of the Hebrew Bible, my research showed that 184 of the 211 NP-ADV’s governed by בוא (qal) (87%) have this semantic function. 36 This observation leads to the article’s main proposal: I would suggest that this well and widely attested valency pattern for בוא (qal) is a clue that the cryptic NP חמדת כל הגוים in Hag. 2.7b is a complement describing the locative goal of the verbal action like the other NP-ADV’s in category (d2).
But what location does חמדת כל הגוים have in view? In context, the most obvious choice is the temple. As discussed more below, the entire focus of Hag. 2.1–9 is on the temple. Most importantly, the temple is in view in the very next clause: ‘I will fill this house with glory’ (Hag. 2.7c). On this reading, חמדת כל הגוים in Hag. 2.7b would be the immediate discourse referent of ‘this house’ in Hag. 2.7c. Thus, recognizing the NP as a description of the temple is in close keeping with the immediate context.
Linguistic evidence for this interpretation comes from Ezek. 24.21 and 25: כה אמר אדני יהוה הנני מחלל את מקדשׁי גאון עזכם מחמד עיניכם ‘Thus says the Lord God: Behold I will profane my sanctuary (i.e., the temple), the pride of your strength, the desire of your eyes … ’ (Ezek. 24.21) ואתה בן אדם הלוא ביום קחתי מהם את מעוזם משׂושׂ תפארתם את מחמד עיניהם ‘And you, son of man, will (this) not (happen) on the day I take from them their stronghold (i.e., the temple)—the joy of their beauty, the desire of their eyes … ’(Ezek. 24.25)
37
In these verses, the temple descriptor ‘the delight/desire (מחמד) of your/their eyes’ employs a cognate of the noun חמדה (‘desire’), the head noun of the relevant NP in Hag. 2.7b. In context, this descriptor is involved in a parallel God is drawing between Ezekiel’s affection for his wife and the people’s affection for the temple. As Block notes, ‘What Ezekiel’s wife was to the prophet (cf. v. 16), the temple is to the people: an object of supreme delight’.
38
I would submit that the NP חמדת כל הגוים in Hag. 2.7b is an analogous descriptor of the temple, but with the agent of desire/delight being ‘all the nations’ rather than the exiles: מחמד עיניהם ‘the desire/delight of their eyes (i.e., the exiles)’ (Ezek. 24.25) חמדת כל הגוים ‘the desire/delight of all the nations’ (Hag. 2.7b)
Consider how the temple descriptor in Ezek. 24.21 and 25 disambiguates the semantics and syntax of the NP חמדת כל הגוים, which it parallels both lexically and grammatically. 39 The syntactic-semantic relationship between the nomen regens (מחמד [‘desire/delight of’]) and nomen rectum (עיניהם [‘their eyes’]) in the construct phrase מחמד עיניהם is: implied verbal notion–S, with the DO of the verbal notion (i.e., the temple) left unspecified (clear from the discourse context): 40 ‘the desire/delight of their eyes’ = (the object) their eyes (S) desire/delight in (implied verbal notion), i.e., the temple.
The corresponding analysis of the parallel חמדת כל הגוים is: implied verbal notion (חמדת)–S (כל הגוים), with the DO of the verbal notion left unspecified (clear from the discourse context): ‘the desire/delight of all the nations’ = ‘(the object) all the nations (S) desire/delight in (implied verbal notion), i.e., the temple’. Notably, the noun חמדה occurs in this type of construct relationship in Dan. 11.37: the desire of women (חמדת נשׁים) = (the god) (DO) women (S) desire (implied verbal notion). The verbless clause למי כל חמדת ישׂראל in 1 Sam. 9.20 provides a similar example: ‘For whom (DO) is all the desire (implied verbal notion) of Israel (S)?’, where the DO (Saul and his father’s house) is represented in the interrogative מי. 41
Thus, taking the NP חמדת כל הגוים as a locative complement describing the temple, the resultant meaning for Hag. 2.7b is: ובאו חמדת כל הגוים ‘And 1
On this construal, Hag. 2.7b exhibits the same common bivalent pattern for בוא (qal) as illustrated in the earlier examples from category (d2): (1) NP-S;
42
(2) NP complement naming where the S is headed (locative goal). Given that the vast majority of such complements are (unlike חמדת כל הגוים) obvious locatives/toponyms, it is important to point out that there are exceptions in other non-narrative texts. Consider, for example, the last clause of Ps. 105.18: ברזל באה נפשׁו ‘His neck came (into) iron (NP) (i.e., an iron collar)’.
Here the NP ברזל (‘iron’) is a non-obvious locative complement of אוב (qal).
43
Isa. 57.2 likely provides another example: יבוא שׁלום ינוחו על משׁכבותם הלך נכחו ‘Those who live uprightly enter a place of peace (NP); they rest (on) their beds’. (Isa. 57.2)
44
As this translation indicates, the non-obvious locative שׁלום (‘peace’) in the initial clause (יבוא שׁלום) probably names the verb’s locative goal (hence the translation ‘place of peace’ and in taking those (lit. the one) who walk uprightly). 45 Pointing in this direction is the corresponding element משׁכבותם (‘their beds’) in the parallel clause (‘they rest on/upon [על] their beds’), which describes the physical location where the dead rest. 46 In light of this contextual parallel and the fact that the vast majority of NP-ADV’s governed by בוא (qal) are locatives, it is best to analyze שׁלום as a locative complement referring to the grave or tomb by metonymy. 47 Thus, while most locative complements of בוא (qal) are obviously so—which is unsurprising, especially given their high frequency in historical narrative 48 —some are not. The non-obvious locative nature of חמדת כל הגוים does not, therefore, invalidate the locative interpretation that the verb’s valency patterns prompt.
This new understanding of Hag. 2.7b fits well in the larger context of Hag. 2.1–9, and has important implications for the (much debated) issue of how the nations are characterized in these verses. Hag. 2.7 is part of a message of encouragement the prophet delivered to the post-exilic community shortly after they embarked upon the daunting task of temple reconstruction. Such a message was necessary because discouragement and grief had set in just weeks after the temple’s foundations were laid. The prophet focuses particularly on the fact that those old enough to recall the splendor of Solomon’s temple soon realized that this temple was comparatively ‘like nothing’ (v. 3). 49
The prophet Haggai therefore encourages the people to ‘Be strong!’ and ‘Work!’ (v. 4) and adds the following motivation: ‘For/because (כי) I (Yahweh) am with you (אני אתכם)’. The prophet gives a third exhortation at the end of v. 5, ‘Do not fear!’ The two-part motive clause that follows in vv. 6–7 (see כי heading v. 6) fixes the community’s attention on God’s sovereign control over the current state of affairs. 50 Boda explains the relationship between the two parts: ‘God’s cataclysmic action first affects creation [v. 6], but then finds its goal in the context of human affairs when it causes a secondary shaking affecting “all nations” [v. 7a]’. 51 This ‘shaking’ will cause all nations to bring their ‘silver and gold’ (v. 8) in order that ‘I [Yahweh] will fill this house [the temple] with glory’ (i.e., material splendor as an expression of God’s manifest presence). 52 Yahweh even promises that the temple’s glory will surpass that of the first (v. 9), a promise that would have been especially encouraging for older generations (cf. v. 3). The end result, Yahweh declares, is that ‘I will give peace in this place’ (see במקום הזה אתן שׁלום), i.e., in both the temple and Jerusalem more broadly (v. 9). 53 The prophet’s aim in the immediate context of Hag. 2.7 (vv. 6–9), then, is to encourage the community in their present work by setting before them a picture of the temple’s/Zion’s future glory.
The temple descriptor ‘the desire/delight of all the nations’ in v. 7b serves this larger purpose. The beleaguered community’s ‘desirable/delightful land (ארץ חמדה)’ (Zech. 7.14; Jer. 3.19; 12.10) and ‘the desire/delight of their eyes’ (מחמד עיניהם) (i.e., the temple) (Ezek. 24.21, 25) had become a desolation. But the prophet assures them that Yahweh will sovereignly ‘shake’ the nations to transform Zion’s (currently dismal) temple into ‘the desire/delight of all the nations (חמדת כל הגוים)’—the object of supreme desire/delight not only for Israel (as in former times; Ezek. 24.21, 25) but also for all nations. Thus, this descriptor indicates that the Great King’s sovereign purpose is not only to transform the temple through the nations but to transform the (formerly hostile) nations themselves to worship God there, an image not unfamiliar to the latter prophets (cf. Isa. 2.1–4; 60.1–8; Zech. 2.15; 14.1–19 [esp. v. 16]). The resultant universal or international vision cast for Zion and her temple would be a great impetus for the difficult work of reconstruction.
A notable implication in light of previous scholarship is that, of the nations Yahweh is sovereignly bringing to Jerusalem, there will clearly be a ‘subgroup among them that recognizes YHWH’s unique deity’; 54 at least some of the wealth coming to Jerusalem for the temple’s glorification will be from tribute (not plunder). That is to say, on this reading חמדת כל הגוים constitutes an overt reference to foreign nations worshipping at Zion’s temple, a reference that other interpretations of the NP (unintentionally) obscure. 55
In light of the foregoing analysis, Hag. 2.7 may be rendered, ‘And I (Yahweh) will shake all the nations so that they will come to “the desire/delight of all the nations” (i.e., the temple), in order that I may fill this house with glory, declares the Lord of Hosts’. 56 On this reading, v. 7b describes the initial purpose of Yahweh’s ‘shaking’ all nations (‘so that they will come [ובאו] to “the desire of all the nations”’, i.e., the temple), and v. 7c the ultimate purpose (‘in order that I may fill [ומלאתי] this house with glory’). Consequently, on this reading the nations are in view in v. 7b, but their treasures do not enter the picture until v. 7c (‘I will fill this house with glory’), and then more explicitly in v. 8 (‘the silver and gold are mine’).
The appeal of this new approach is that it explains the MT of Hag. 2.7 in a way that is not only contextually fitting, but also linguistically robust. It does not require positing an unattested collective sense for the noun חמדה or a conjectural emendation to a plural form to create agreement between the clause’s S and V. Grammatical agreement is straightforward on this approach: the clause’s plural V (באו) simply has a plural S (כל הגוים). Additionally, the interpretation of חמדת כל הגוים as a descriptor for the temple (and so a physical location) is supported by the presence of a similar descriptor in Ezek. 24.21, 25, and is firmly grounded in the valency patterns for the clause’s V. I would, therefore, submit that the proposed analysis offers a more satisfactory reading of Hag. 2.7b, and so the verse as a whole, than previous interpretations.
