Lecture 6-1: Prophetic Forms: Oracle, Woe, Vision Report

The specimen text for the lecture is the first of the eight oracles against the nations:

And the eighth in the series, which is the one the other seven exist for:

The Hebrew of the opening of 1:3:

כֹּה אָמַר יְהוָה עַל־שְׁלֹשָׁה פִּשְׁעֵי דַמֶּשֶׂק וְעַל־אַרְבָּעָה לֹא אֲשִׁיבֶנּוּ

koh amar YHWH / al-sheloshah pish'e dammeseq ve-al-arba'ah lo ashivennu — "Thus said Yahweh: for three transgressions of Damascus, and for four, I will not turn it back."

Set that against WEB's "I will not turn away its punishment" and you have already found this lecture's first working problem, which §3.4 takes up: the Hebrew verb has an object suffix, "it," and the Hebrew does not say what "it" is.

Two smaller notes on the printed Hebrew. אָמַר (amar) is a qatal, a completed form — "said," not "says." Every English version prints a present tense, and §3.2 explains why that is defensible and what it costs. And פִּשְׁעֵי (pish'e) is the construct plural of פֶּשַׁע (pesha), the first noun of a construct chain — the Hebrew way of joining two nouns so that the first depends on the second — so the phrase is "transgressions-of Damascus." The noun does not mean "sin" in general. Its home is the language of rebellion against a superior, and it is used of a vassal breaking with an overlord; "transgression" and "crime" are both attempts at it, and neither carries the political colour.

3.1 Why forms, and not just sentences

Take a page of Amos and try to divide it into units without any knowledge of form. Where does one oracle stop? Chapter divisions are medieval and verse divisions later still; the paragraph markings in a printed Hebrew Bible are ancient but not original either. In practice, readers divide prophetic books by recognising openings and closings — a messenger formula at the front, a signature formula at the back, a הוֹי that starts something new — and these are exactly the forms this lecture describes. Form recognition is not a decoration on top of reading. It is how you find out what you are reading.

The forms also tell you what kind of claim is being made. "Thus said Yahweh" is not a rhetorical flourish; it is a specific claim about where the following words come from, and it makes the speaker a conduit rather than an author. A cry of woe is not an insult; it has a shape borrowed from somewhere else in Israel's life, and hearing that borrowing is part of hearing the line. None of this is visible in a translation that renders every clause correctly.

3.2 The messenger formula

The formula is כֹּה אָמַר יְהוָה (koh amar YHWH). כֹּה is an adverb, "thus, in this way"; אָמַר is a third masculine singular qatal of אמר, "he said." Literally: "thus said Yahweh."

The reason this formula matters is that it is not a special religious phrase. It is the ordinary Hebrew way of introducing a message carried by a human messenger. When Jacob sends word ahead to Esau, the messengers are told to say כֹּה אָמַר עַבְדְּךָ יַעֲקֹב (koh amar avdekha ya'aqov), "thus said your servant Jacob." The messenger then speaks in the first person — "I have sojourned with Laban" — although the messenger is not Jacob. This is the shape prophetic speech takes. The prophet says "I" and means Yahweh, and the formula is what licenses the shift.

Ludwig Köhler set this out in 1923, and Claus Westermann made it the foundation of Basic Forms of Prophetic Speech in 1960: the primary form of prophetic address is messenger speech, and its parts are the commissioning formula, the messenger formula, and the message. The observation is straightforwardly checkable from the Hebrew Bible's narrative books, and it is one of the more secure results in the field.

The tense. אָמַר is qatal. Chapter 3-1 discussed at length what a qatal does and how much disagreement there is about it; here the practical point is that the form is not a present tense, and "thus says Yahweh," which is what nearly every English version prints, is an interpretation. The two common analyses are that the qatal reports the moment the prophet received the message, which is genuinely past by the time he speaks it, or that it is what grammars sometimes call a performative or an instantaneous perfect, the speech act coinciding with the utterance. The King James Version's "Thus saith the LORD" preserves the ambiguity in older English because "saith" there is not obviously a narrow present. Modern versions have to choose, and they choose the present because English needs a tense and the past would suggest the message is stale. You should be able to say that the Hebrew is a qatal. You should not say that the English versions have mistranslated it.

The closing signature. The counterpart at the end of a unit is נְאֻם־יְהוָה (ne'um YHWH), conventionally "declares Yahweh" or "says Yahweh." נְאֻם is a noun, not a verb — an "utterance" — so the phrase is really "utterance of Yahweh," a construct chain again. In Amos it often ends an oracle, and its position is one of the more reliable markers for where a unit stops.

3.3 The two-part judgement oracle

The commonest large form in Amos has two parts. First an indictment: what has been done, usually stated with a qatal or a participle, and usually introduced by a preposition or a particle meaning "because." Then a sentence: what will follow, usually first-person, introduced by לָכֵן (lakhen, "therefore") or by a simple conjunction.

Amos 1:3-4 is exactly this. The indictment is עַל־דּוּשָׁם … אֶת־הַגִּלְעָד — a preposition עַל with an infinitive construct and a suffix, "because of their threshing… Gilead" — and the sentence follows in verse 4 with the burning of Hazael's house. Amos 2:6-16 is the same shape at greater length. When you meet לָכֵן in a prophetic book, stop and look back: the sentence is beginning, and everything before it was the case for the prosecution.

The order matters more than it looks. Announcing a punishment first and the reason afterwards would make the reason a justification added under pressure. This form makes the offence the ground and the sentence the consequence, and the grammar carries that: the causal preposition subordinates the indictment to what follows.

3.4 Amos 1-2: the same words seven times, and then not

Now the set piece. Eight oracles run from 1:3 to 2:16, against Damascus, Gaza, Tyre, Edom, Ammon, Moab, Judah and Israel. Each begins with the identical Hebrew, changing only the name:

כֹּה אָמַר יְהוָה עַל־שְׁלֹשָׁה פִּשְׁעֵי [X] וְעַל־אַרְבָּעָה לֹא אֲשִׁיבֶנּוּ

koh amar YHWH al-sheloshah pish'e [X] ve-al-arba'ah lo ashivennu

The graded numerical formula. "For three… and for four" is a fixed Hebrew pattern, x followed by x+1, familiar from wisdom literature — Proverbs 30 runs several of them, "three things… four things." Wolfgang Roth's study of the form in 1965 established how widely it is spread. What it means is disputed in detail but agreed in outline: it is not arithmetic. Nobody is counting to seven. The effect is roughly "more than enough, and one more than that," an indefinite quantity presented as though it were being tallied.

לֹא אֲשִׁיבֶנּוּ, and the missing antecedent. The verb is a hiphil yiqtol, first person singular, of שׁוב, "turn, return"; the hiphil is causative, so "I will cause to turn back, I will bring back, I will revoke." Attached is a third masculine singular object suffix: "it." And the Hebrew never says what "it" is.

This is a point where readers will not agree. Four proposals have been current:

1. The punishment, supplied from context. This is what WEB prints ("I will not turn away its punishment"), and what the King James Version prints ("I will not turn away the punishment thereof"). The word "punishment" is not in the Hebrew.
2. The decree or the word — Yahweh will not revoke what he has already said. Shalom Paul argues along these lines, and it has the advantage of giving the suffix a plausible masculine antecedent in the prophetic word itself.
3. Yahweh's anger, which is masculine in Hebrew and is the object of the same verb elsewhere.
4. The nation named, taken as the object: "I will not bring it back," that is, restore it.

The grammar excludes none of these. What the grammar does establish is that there is a suffix, that it is singular and masculine, and that the versions which print "its punishment" have supplied a noun. Modern versions vary: some print "I will not revoke the punishment," some "I will not revoke it," some restructure the clause. A reader who knows only English cannot tell that anything has been added. That is the whole reason for this course, in one clause.

The turn. Seven oracles fall on Israel's neighbours, and there is good reason to think an Israelite audience would have heard them with approval — the crimes listed are atrocities committed against others, several against Israel itself. The Hebrew of the eighth oracle opens identically, word for word, with עַל־שְׁלֹשָׁה פִּשְׁעֵי יִשְׂרָאֵל (al-sheloshah pish'e yisra'el). And then the indictment does not describe a war crime. It describes selling a righteous man for silver, and a poor man for a pair of sandals, and it goes on at four times the length of any of the seven.

Two features of the Hebrew make the trap visible on the page. The identical formula. The repetition is exact; nothing prepares the hearer. And the length: the seven foreign oracles run two or three verses each; the Israel oracle runs eleven, nearly four times the longest of them, and its sentence section does not end where the pattern says it should. The form has been used as a delivery mechanism and then abandoned. John Barton's study of these oracles argues, against a common assumption, that Amos is not appealing to a known code of international law that the nations had signed; the case is made rather by getting the audience to agree to a principle and then applying it to them. Whether or not you follow Barton on that, the rhetorical mechanism is legible from the Hebrew alone.

3.5 The woe cry

The word is הוֹי (hoy), conventionally "woe" or "alas." Grammatically it is an interjection, and the construction after it is remarkably stable across the prophets: הוֹי followed by a participle in the plural, naming the addressees by what they do, and then a description or a threat.

Amos 6:1 begins הוֹי הַשַּׁאֲנַנִּים בְּצִיּוֹן (hoy ha-sha'ananim be-tsiyyon), "Woe, those at ease in Zion" — though here what follows הוֹי is an adjective with the article rather than a participle. Amos 5:18, the passage lecture 6-2 reads, begins הוֹי (hoy) with a hitpael participle, "those who long for." Isaiah 5 runs a whole series of them.

This is a point where readers will not agree. Where does the woe cry come from?

• The funeral lament. הוֹי is the cry raised over a dead person in some narrative texts, and on this view the prophets are pronouncing a funeral over the living: the addressee is treated as already dead. Claus Westermann and Günther Wanke argued in this direction, and Waldemar Janzen developed it at length in Mourning Cry and Woe Oracle (1972). It gives the form a strong and immediately intelligible force.
• Wisdom instruction. Erhard Gerstenberger argued in a 1962 article that the woe cry belongs with the curse-and-blessing language of the wisdom teachers rather than with mourning, pointing to the participial construction, which describes a type of person — the one who does such-and-such — exactly as a proverb does.
• A caution against both. Meir Weiss questioned whether the funerary use is early enough or frequent enough to be the source at all, and argued that the prophetic use may simply be an attention-getting cry with no single institutional home.

What can settle this? Only distribution and usage, and the corpus is small. The honest position is that the funeral connection is well attested for the word and not demonstrated for the form, and that a preacher who says "in Hebrew, hoy is a funeral cry, so the prophet is holding a funeral for Israel" is stating a hypothesis as a fact. It may be a good hypothesis. It is not a fact about the word.

3.6 The vision report

Amos contains five vision reports (7:1-3, 7:4-6, 7:7-9, 8:1-3, 9:1-4), and the first four share a pattern tight enough to be laid out as a template:

1. An introduction with the verb ראה in the hiphil: הִרְאַנִי (hir'ani), "he showed me," or in the first two כֹּה הִרְאַנִי אֲדֹנָי יְהוִה (koh hir'ani adonai YHWH), "thus the Lord Yahweh showed me."
2. The thing seen, introduced by וְהִנֵּה (ve-hinneh, "and behold"), the particle lecture 3-3 discussed: it points, and it puts the reader at the prophet's eye.
3. A question and an answer — in the third and fourth visions, מָה־אַתָּה רֹאֶה עָמוֹס (mah-attah ro'eh amos), "what do you see, Amos?", answered with a single word.
4. An interpretation, which is where the vision turns into a judgement.

The third and fourth visions do something with step 3 that a reader of a translation cannot see. In the fourth, the thing seen is כְּלוּב קַיִץ (keluv qayits), a basket of summer fruit; and the interpretation announces that בָּא הַקֵּץ (ba ha-qets), "the end has come." קַיִץ (qayits, "summer fruit") and קֵץ (qets, "end") are different words from different roots that sound almost alike. The vision is a pun. WEB prints "a basket of summer fruit… the end has come," which is accurate and in which nothing whatever happens.

Notice what this does and does not license you to say. It is a real feature of the Hebrew, unmistakable, and it is a fair example of chapter 2-3's point that some things do not survive translation and a footnote is the only remedy. It does not license the claim that the two words are etymologically connected or that one "really means" the other. They are a sound-play, and sound-play is a literary device, not a semantic argument.

In the third vision the thing seen is אֲנָךְ (anakh), traditionally "plumb line," and that word is a genuine problem which lecture 6-3 takes up in detail. Note here only the structural point: the pattern of the vision reports leads you to expect a pun in the third vision too, and readers who assume there must be one have sometimes let that expectation drive their lexical decision. An expectation generated by a form is not evidence about a word.

3.7 What form criticism establishes, and where it stops

Hermann Gunkel's programme, from the turn of the twentieth century, had three parts: identify the recurring form (Gattung), reconstruct the setting in the life of the community where it belonged (Sitz im Leben), and trace the history of its use. The first part is descriptive and checkable. The second and third are inferential, and the inference is often long.

For prophecy, the first part has held up well. The messenger formula, the woe cry, the vision report, the two-part oracle, the disputation, the summons to hear — these are real and recurrent, and a student who learns them reads better. The second part has held up less well. We have very little independent evidence about the institutional settings of prophetic speech in eighth-century Israel, and a setting reconstructed from the form and then used to explain the form is circular. Hans Walter Wolff's proposal that Amos's language shows the influence of clan wisdom — his Amos' geistige Heimat, "the intellectual home of Amos" — is a careful example of this kind of argument, and it has been both extended and disputed. Robert Wilson's later work on prophecy and society tried to put the sociological side on firmer comparative ground.

Three limits to keep in front of you:

• A form is a description of the text, not a report about its history. Recognising the shape of a woe cry tells you what the unit is doing. It does not tell you whether Amos composed it, quoted it, or inherited it.
• Deviation from a form is not evidence of corruption or of later editing. It is a live option, but a text that departs from its pattern may be departing on purpose — Amos 2:6-16 is the standing proof.
• Forms are not airtight. Real texts mix them, truncate them and run them together. A student who has just learned the categories will be tempted to force every unit into one; resist that, and be willing to say that a passage is a woe cry that turns into something else halfway through, because sometimes it is.

The diagram shows the mechanism of Amos 1-2 as a single figure: the invariant formula across the top, the seven foreign oracles in one panel with their crimes and their verse counts, and the eighth set apart in the warning colour because it is where the pattern is turned. The Hebrew of the formula and of the name Israel is given with transliteration underneath, so the figure remains readable where the Hebrew font is unavailable.

6. Further Reading

• Claus Westermann, Basic Forms of Prophetic Speech, trans. Hugh Clayton White (Westminster, 1967; German 1960). The classic treatment of the messenger formula and the judgement oracle; read at least the section on messenger speech.
• Hans Walter Wolff, Joel and Amos, Hermeneia (Fortress, 1977). Form-critical analysis of every unit in Amos, and the source of the "clan wisdom" proposal discussed in §3.7.
• Shalom M. Paul, Amos, Hermeneia (Fortress, 1991), on 1:3 and the object of אֲשִׁיבֶנּוּ, and on the graded numerical formula.
• Waldemar Janzen, Mourning Cry and Woe Oracle, BZAW 125 (De Gruyter, 1972), for the funerary account of הוֹי.
• Erhard Gerstenberger, "The Woe-Oracles of the Prophets," Journal of Biblical Literature 81 (1962), 249-263, for the wisdom account.
• Meir Weiss, "The Origin of the 'Day of the Lord' — Reconsidered," Hebrew Union College Annual 37 (1966), and his related work questioning the form-critical derivations.
• John Barton, Amos's Oracles Against the Nations (Cambridge University Press, 1980), on what the eight oracles presuppose about their audience.
• Wolfgang M. W. Roth, Numerical Sayings in the Old Testament, VTSup 13 (Brill, 1965), for the x / x+1 pattern.
• Marvin A. Sweeney and Ehud Ben Zvi, eds., The Changing Face of Form Criticism for the Twenty-First Century (Eerdmans, 2003), for the state of the method and its critics.
• Bruce K. Waltke and M. O'Connor, An Introduction to Biblical Hebrew Syntax (Eisenbrauns, 1990), §30.5 on the performative use of the qatal, for §3.2.