Isaiah 66:13 - On the noun אִישׁ

(The above rendering comes from the RJPS translation, an adaptation of the NJPS translation. Before accounting for this rendering, I will analyze the plain sense of the Hebrew term containing אִישׁ, by employing a situation-oriented construal as outlined in this introduction, pp. 11–16.)

The meaning of אִישׁ here is a longstanding interpretive crux, which will be resolved in this comment. The summary is: Depicting a situation schematically (as here) is a prototypical usage of אִישׁ as a situating noun. Here אִישׁ is employed to invoke a stereotypical situation, to illuminate the depicted situation. It is used as the label for a crying infant in need of maternal soothing, in accord with the preceding imagery.

To understand the plain sense of this verse, it must be read in the context of the preceding three verses. In those verses, which are addressed to Israel in exile, the topic is the people’s forthcoming good fortune, as promised by GOD. Israel is portrayed as a suckling child. Such imagery continues into this verse.

Numerous scholars construe this verse as beginning a new sentence that summarizes the preceding few verses. So Luzzatto (mid-19th century) wrote explicitly, ad loc.; so the renderings in KJV, NRSV, ESV; McKenzie (Anchor Bible) 1968:206; Blenkinsopp (Anchor Bible) 2003:302; Watts (Word Biblical Commentary) 2005:934.

The problem with such a con­strual is that it does not fit the Hebrew syntax. It therefore yields an incoherent text, in two ways:

1. The syntax of the verse’s second clause is the inverse of its first clause, which (if they are read together) makes them clash with each other in terms of semantic roles.
2. It makes the verse’s third clause seem merely tacked on. This explains why Elliger and Rudolph noted in BHS (1977:778) that the third clause is “perhaps added.”

Claim #1 requires some explanation. In the Bible, comparative constructions similar to this case are not uncommon (HALOT [1967] 2001:483; cf. Joüon 2006:604–5). A protasis introduced by the comparative preposition -כּ ‘like’ and an apodosis introduced by the discourse deictic כֵּן ‘so’ are indeed combined 60 times—as ostensibly here in verse 13. (On כֵּן as a discourse deictic term, see Forbes 2014; BHRG 2017:437.) That is how this verse is conventionally construed: the -כּ ‘like’ at the start of the first clause leads into the כֵּן ‘so’ at the start of the second clause.

Yet whereas in all the other cases, syntactic and semantic parallels in the two halves align with each other, here they do not align:

• the syntactic mate for אִישׁ in the purported protasis is customarily construed as אָנֹכִי ‘I’ in the supposed apodasis, but the semantic role of אִישׁ (as patient) is not matched in the second clause by that pronoun, but rather by the second-person plural object suffix כֶם-; and
• the semantic role of אֵם ‘mother’ in the first clause (as agent) is matched by the deity’s first-person pronoun אָנֹכִי in the second clause, while that noun in the first clause lacks any syntactic mate in the second clause.

Translators have handled the mismatch between those clauses in two unsatisfying ways. Many of them, such as McKenzie 1968:206, NJPS (whose Isaiah appeared in 1973), NRSV (1989; 2022), Childs (Old Testament Library) 2001:530, Blenkinsopp 2003:302, and Watts 2005:934, surreptitiously re­work the syntax of the first clause so as to match the second clause.

As a mother comforts her son / So I will comfort you… (NJPS)

Alternatively, KJV and ESV faithfully preserve the dissonance—thus resulting in an incoherent rendering.

As one whom his mother comforts, / so I will comfort you… (ESV)

Now, here is the solution from a situation-oriented approach. The situating noun אִישׁ is serving one of its standard discourse functions: it efficiently invokes a situation—in this case, it is a stereotypical one that is depicted schematically. The speaker nests that sketch inside a comparison of that situation with the one under discussion (so also, e.g., in Zech 4:1; Ps 38:15). According to the normal pattern of such comparisons, the simile illuminates the preceding description—not the following one.

Wait, you may ask, can the simile apply to the preceding description despite the intervening verse break? Yes. In the Hebrew Bible, what we moderns would call a sentence (a logically complete thought) occasionally transcends the boundary of a Masoretic verse—ending in the middle of a verse. In the book of Isaiah alone, see 10:1–2; 25:4–5; 29:5–6; 30:4–5; 43:20–21; 58:13–14; 65:6–7. See also Ginsburg, Introduction to the Massoretico-Critical Edition of the Hebrew Bible, 1897:68–108; and the NJPS translators’ cognizance of this biblical practice (Orlinsky, Notes on the New Translation of the Torah, 1970:21, citing the 10th-century exegesis of Saadia Gaon). That, I claim, is the case here, as well.

According to my proposal, the comparative preposition -כּ ‘like’ modifies the immediately preceding depiction. Then the discourse deictic כֵּן ‘so’ looks backward, taking God’s whole quoted promise as its anaphor, starting with הִנְנִי נֹטֶה ‘I will extend…’ (v. 12).

I.e., contrary to the conventional construal, the particle כֵּן is not relying upon the clause headed by -כּ to serve as protasis. Rather, it stands on its own, to introduce a summary. This usage of כֵּן is likewise a conventional one—occurring 120 times, in which the particle roughly corresponds to ‘thus, as had just been told’ (HALOT [1967] 2001:482), e.g., Exod 10:29; Job 8:13.

In this respect, my proposal resembles the NJPS construal of Isa 63:13–14, which renders clauses that begin with -כּ and with כֵּן in separate sentences:

[13] “…So that they did not stumble—

[14] …Like a beast descending to the plain?”

’Twas the spirit of the Lord gave them rest;

Thus did You shepherd Your people

To win for Yourself a glorious name.

See also Ezek 23:44 (NJPS; so also NRSV, ESV).

As a result of parsing the first two clauses of v. 13 in this manner, the last clause now provides closure to the passage, by making a satisfying concluding statement.

Meanwhile, in this usage, what information does אִישׁ tell us about its referent? Commentators and translators are divided over the referent’s implied age. Given the imagery of a suckling child in the preceding verses, it is not surprising that some scholars understand אִישׁ contextually as ‘child’ (NRSV; Childs) or ‘son’ (Ibn Ezra; NJPS; Blenkinsopp). Others render אִישׁ in more conventional terms that are more consistent with its usage elsewhere: ‘a man’ (McKenzie 1968:206), ‘a person’ (Watts 2005:934), or ‘someone’ or ‘one’ (LXX, KJV, ESV; Malbim, ad loc.).

A situating noun that is functioning on the discourse level contributes minimal meaning on the informational level. As such, there is nothing to prevent אִישׁ from being employed in nonspecific initial (anaphoric) reference to an infant. Thus here in Isa 66:13, even though אִישׁ refers to (denotes) a suckling child, I do not claim that it means “infant.” Rather, its principal meaning is its discourse function: to efficiently evoke a familiar situation in terms of its principal participant. (For other instances of some form of אִישׁ as denoting an infant or child, see Gen 4:1; Num 30:4; 31:18, 35; Judg 21:14.)

Such a usage of אִישׁ minimally constrains its referent’s gender. Because the reference is nonspecific in nature, neither אִישׁ nor the possessive pronoun that it governs can ascribe gender to their stereotypical referent, beyond the mild constraint of “not exclusively female” (Stein 2013). In effect, the reference is gender inclusive. And there is nothing in the imagery of a crying infant that constrains its gender, either. Hence there is no warrant for rendering in gendered terms.

The above construal, which relies upon the conventional understanding of אִישׁ as performing its prototypical situating function, readily yields a coherent and informative Hebrew text. Consequently, it must be the plain sense of this passage.

Note, too, that my construal breaks the verse into two thoughts at the major disjunctive accent (etnaḥta), which seems more consistent with the pattern of Masoretic accents than does the conventional parsing. Although I am admittedly not an expert in this area, I would not expect the major disjunction to occur between the protasis and the apodosis of a comparison—if that were indeed the intended syntax—when a third part of the verse follows, as here. Rather, I would expect the major break after that comparison. (Compare Deut 8:20; Josh 1:17; Isa 31:5; Jer 2:26; 5:27; Song 2:3.)

As for rendering into English, the NJPS rendering of this verse misses the mark, as discussed above. The revised rendering recasts the first clause to match the Hebrew semantic roles, with a gender-neutral indefinite pronoun for אִישׁ. The modified punctuation now connects that clause with the preceding verse.