"Blessed shall he be that taketh thy little ones"

What the psalm is doing

Psalm 137 is short. Nine verses. The first four set the scene by the rivers of Babylon. The captors demand a song of Zion. The exiles will not give one. The next two verses are a self-targeting oath: if the singer ever forgets Jerusalem, let his right hand forget its skill and his tongue stick to his palate. The last three verses are imprecation. Verse 7 calls on God to remember Edom on the day of Jerusalem's fall. Verse 8 calls Babylon the 'daughter doomed to destruction' and pronounces a blessing on whoever repays her in kind. Verse 9 names the kind: 'Blessed shall he be, that taketh and dasheth thy little ones against the stones.'

Two things about the closing line are not in dispute. First, it is the climax of the poem, not an outburst that breaks the structure. The whole psalm has been building toward retribution, with the oath in verses 5-6 anchoring the singer's loyalty to Jerusalem before the curse on Babylon arrives. Second, the violence described in verse 9 is the violence Babylon committed against Judah. Jeremiah 19:9 records mothers eating their children inside the besieged Jerusalem. Lamentations 2:11-12 and 4:4 describe infants dying in the streets. 2 Kings 8:12 has the Aramean threat of 'ripping up' the children of besieged cities. Verse 9 is asking God for symmetry, not for novelty.

What is in dispute is what kind of speech act this is. Is the psalm modeling honest prayer, including raw rage, before God? Is it the legitimate cry of victims that God authorizes as scripture without endorsing as policy? Is it a prophetic invocation of a judgment Isaiah 13:16 had already pronounced on Babylon? Or is it speech that the church should retire from public worship as morally unfit, even if it stays in the canon as testimony to suffering?

Psalm 137 next to the other imprecation psalms

Psalm 137 is the most famous of the imprecatory psalms but not the only one. Three other psalms (58, 69, and 109) carry curses substantial enough that the same liturgical and ethical questions arise. Reading them side by side surfaces what is and is not unusual about Psalm 137.

Reading the four columns together surfaces what Psalm 137 has and does not have. It is the only one of the four where the curse names infants as the specific target rather than as collateral consequence of judgment on the enemy. It is the only one set explicitly inside imperial conquest (the others are royal-court or personal-enemy settings). And it is the one without a New Testament citation that would re-attribute the speech to Christ or to the apostolic community. Psalms 69 and 109 enter Christian liturgy with built-in re-attribution; Psalm 137 does not.

What Babylon actually did to Judah's infants

Verse 9 is not asking for a novel atrocity. It is asking for what Babylon had done back. The historical record on the inside of Jerusalem's siege (587-586 BC) is preserved in three places.

Jeremiah 19:9, the prophet's foretelling of the siege: 'I will cause them to eat the flesh of their sons and the flesh of their daughters.' Lamentations 2:11-12 records the same event from inside the city: 'My liver is poured upon the earth, for the destruction of the daughter of my people, because the children and the sucklings swoon in the streets of the city. They say to their mothers, where is corn and wine? when they swooned as the wounded in the streets of the city, when their soul was poured out into their mothers' bosom.' Lamentations 4:4: 'The tongue of the sucking child cleaveth to the roof of his mouth for thirst: the young children ask bread, and no man breaketh it unto them.' 2 Kings 8:12 records the standard ANE siege threat: 'I know the evil that thou wilt do unto the children of Israel: their strong holds wilt thou set on fire, and their young men wilt thou slay with the sword, and wilt dash their children, and rip up their women with child.' The verb 'dash' (Hebrew rāṭaš) in 2 Kings 8:12 is the same verb in Psalm 137:9.

The four positions above agree on this background. They disagree on what the background means. The lex-talionis position takes the symmetry as the point of the verse: the psalmist asks God for what Babylon did. The liturgical-corporate position takes the symmetry as evidence that the verse is a cry of the wronged rather than a license for new violence. The unfit-for-liturgy position grants the symmetry but argues the church's prayer cannot ask for symmetry of this kind. The protest-speech position takes the symmetry as the form the witness has to take, since softening it would be softening what was done.

How the Christian liturgy has actually used the psalm

The reception history of Psalm 137 in Christian worship is the most concrete place to see the four positions playing out. The choices made at each stage of liturgical history are not abstract; they are practical decisions about what congregations actually pray out loud.

Augustine in the early fifth century reads the whole psalm allegorically. The 'rivers of Babylon' are the disordered loves of the present age; the 'little ones' of verse 9 are nascent sinful thoughts; the 'rock' on which they are dashed is Christ (Augustine connects Ps 137:9 to 1 Cor 10:4, 'that rock was Christ'). The reading lets the church keep the verse in its prayer life by reinterpreting what the words refer to. Cassiodorus, Bede, and the medieval Latin tradition broadly inherit this move. The whole psalm stays in the Divine Office. The literal sense recedes.

The medieval and Reformation tradition keeps the psalm in the daily office without much change. Calvin in his commentary on Psalms treats verse 9 as the legitimate cry of victims for divine justice while warning against personal vindictiveness. The same psalm is sung in the monastic offices and in the Reformed Genevan Psalter. The Westminster Directory for Public Worship (1645) does not single out the imprecatory psalms for omission.

The shift comes in the twentieth century. The Roman Catholic Liturgy of the Hours, revised after Vatican II (the Liturgia Horarum was promulgated by Paul VI in 1971), explicitly omits Psalms 58 and 83 entirely and excises verses 7-9 of Psalm 137 from the four-week psalter cycle. The stated reason in the General Instruction is the difficulty of these psalms 'for use in the liturgy of an age of pluralism' and the inability of the Christian prayer to be 'in conformity with the spirit of the New Testament.' The Church of England's Common Worship (2000) marks Psalm 137:7-9 as optional. The Episcopal Book of Common Prayer (1979) places the verses in brackets for the daily office.

Several Reformed and Lutheran traditions made the opposite call. The Lutheran Service Book (2006) retains Psalm 137 in full. The Presbyterian Trinity Psalter Hymnal (2018) includes a metrical version of the whole psalm, including verses 7-9. Bonhoeffer's argument from The Prayerbook of the Bible (1937) is regularly invoked in defense of retention: the imprecatory psalms can be prayed by Christians only in Christ, who absorbs the judgment, but they must be prayed because they belong to the church's witness against atrocity. The Eastern Orthodox liturgy retains the full psalm in the Lenten cycle, and verse 9 appears in the Bridegroom Matins service of Holy Week as part of the Great Reading of the Psalter.

The 'little ones' as nascent vice

The patristic and medieval allegorical reading of verse 9 deserves a closer look, because it is the move that kept the verse in Christian liturgy for thirteen hundred years before the twentieth-century excision started. Augustine's Enarrationes in Psalmos 136 sets out the framework. Babylon is the city of disordered love; the 'little ones' (parvuli) are the early stages of sinful thoughts; dashing them on the rock (which is Christ, per 1 Cor 10:4) is the ascetic discipline of resisting temptation before it grows into established sin.

The reading is theologically careful: the violence in the verse is not directed outward at people but inward at the self's own incipient sins. Cassiodorus's sixth-century Expositio Psalmorum and Bede's eighth-century commentary both extend the allegory. Aquinas in the thirteenth century carries it forward. The reading is the standard medieval Christian way of praying the verse: not as a curse on Babylonian infants, but as a prayer for one's own moral vigilance.

Modern readers in the lex-talionis and protest-speech positions split on this move. Some (Wenham, Zenger) treat the allegorical reading as one legitimate way the church has prayed the verse, alongside the literal-sense reading as protest against historical Babylon. Others (Brueggemann, Volf) argue the allegorical move evacuates what the verse is doing and effectively hides the suffering of imperial victims behind a piety of inner moral struggle. The unfit-for-liturgy position is hardest on the allegory: Lewis explicitly calls it evasion. Whatever one makes of the allegorical reading, it is the historical answer to the question of how the church has prayed verse 9 for most of its existence.

Where the positions actually disagree

Stepping back from the four readings, the disagreements cluster around three questions. First, what kind of speech is the imprecation. The lex-talionis position reads it as authorized prayer that gives rage to God. The liturgical-corporate position reads it as communal lament that surrenders vengeance to God. The unfit-for-liturgy position reads it as testimony that does not function as Christian prayer. The protest-speech position reads it as authorized witness that refuses to hide what was done.

Second, who is the speaker. The lex-talionis and liturgical-corporate readings can keep the historical exilic community as the speaker. The unfit-for-liturgy reading argues the church cannot make that speaker its own. The protest-speech reading, especially in Bonhoeffer's form, places the speech in Christ's mouth and re-attributes the prayer at the Christological level.

Third, what the verse does in the canon if the church does not pray it publicly. The unfit-for-liturgy position is comfortable with omission in worship and retention in study and preaching. The other three positions resist excision, on the grounds that what the church refuses to pray, the church gradually stops hearing. The four positions agree that the verse stays in scripture; they disagree on whether scripture and liturgy are the same thing.

Reading the psalm with the question open

Psalm 137 is not the only text in the canon that refuses to soften the cost of imperial violence. It is one of the most visible, because it ends on a single line that cannot be glossed past. The four positions above name four ways Christian and Jewish readers have stayed with the verse. None of them removes the difficulty. What they do is make the difficulty visible at different points: in what kind of speech act the verse is, in who the church names as the speaker, in whether the church prays the verse out loud, and in what the church does with a canon that includes the testimony of victims who wanted infants dashed against the rocks.

Most readers who stay with the psalm for any length of time end up holding pieces of more than one position. Few who keep the verse in their personal prayer life pray it without Bonhoeffer's Christological re-attribution. Few who excise it from public worship are comfortable saying the church has nothing to learn from it. What the psalm requires is that the reader take a position knowing what each one costs.