Wednesday 8 November 2023

Textual errors in Lassus' Prophetiae Sibyllarum

It isn’t often that I get to combine my interests as a choral singer and a classicist. Recently I got interested in the Prophetiae Sibyllarum or ‘Sibylline prophecies’ by Roland de Lassus, also known as Orlando di Lasso. These are twelve short pieces for unaccompanied choir, in Latin, composed around 1560.

This story is about how I noticed an error in the text.

Lassus’ pieces are startling, polytonal, chromatic compositions. They’re not often performed. Here’s a selection of scores on the International Music Score Library. And here’s a 1994 recording by the German group Cantus Cölln.

Roland de Lassus

Why Cantus Cölln, rather than one of the more recent recordings by De Labyrintho (2007), Vocalconsort Berlin (2015), or La Main Harmonique (2022)?

Because the Cantus Cölln recording is one of the few to use the correct text. The others all use a faulty text. So does every score hosted by the Choral Public Domain Library. But then, if everyone used the Bärenreiter edition (Schlötter 1990), we wouldn’t be having this conversation.

Dactylic hexameter

As I went through the Prophetiae, I slowly realised that the text uses an ancient verse form called dactylic hexameter, the same metre used in ancient epic and oracular verse. The poems that Lassus has set aren’t ancient: the text was written around 1500. Some neo-Latin poets in the modern era have used dactylic hexameter — John Milton, for example — but this was my first time hearing hexameter in Renaissance choral music.

It was in the fourth movement, the ‘Cimmerian Sibyl’, that things went horribly wrong.

In teneris annis facie insignis honore ...

Of tender years, with glorious face, in honour ...

This text is impossible.

Hexameter is supposed to have a steady — ⏑⏑ pulse (or in musical notation, 𝅘𝅥 𝅘𝅥𝅮𝅘𝅥𝅮). The problem is with the words facie insignis. At the place in the line where they sit, they ought to have the rhythm ⏑⏑ — — — ⏑. But they don’t. They can’t.

Latin hexameter has strict rules. One of those rules is that the last syllable of facie has to be abbreviated: faci’ insignis. The text of the Prophetiae Sibyllarum follows that rule rigidly elsewhere. But then facie insignis has the rhythm ⏑⏑ — — ⏑. And that isn’t hexameter.

Is it possible that most of the time the poet had a detailed, skilled knowledge of the intricacies of Latin dactylic hexameter, but when it came to this one line, everything suddenly went pear-shaped?

No, of course not. It’s a textual error. And it isn’t Lassus’ error: he got it right. It’s every edition from 1600 up until 1979 that got it wrong. I find it wild that in that period not a single editor noticed the problem.

The Sibylline oracles

It’d be wise to fill in some background here. There are real ancient poems called the Sibylline oracles, but these are not they. The poems set by Lassus are modern.

The ancient Oracles are a set of Judaeo-Christian poems ranging from the 2nd century BCE to the 4th century CE. They too are in dactylic hexameter, but the ancient poems are in Greek, and they’re hundreds of lines long. They’re fascinating poems in their own right. They present themselves as being prophecies spoken by actual oracles, prophesying Jewish history, the coming of Christ, the end times, and so on.

Michelangelo didn’t include the Cimmerian Sibyl on the Sistine Chapel ceiling: here’s the Libyan Sibyl instead, along with her counterpart in the 1505 edition. (sources: Barbieri ca. 1505, ch. 3;

From late antiquity onward, Christian readers sometimes assumed the poems genuinely were oracular pronouncements. As a result, both the poems and the Sybils acquired a lot of Christian street cred. The Sibyls were reckoned among the virtuous heathen because they supposedly foretold the coming of Christ.

Some ancient Christian writers celebrated them, especially Lactantius, so their fame came to be independent of the Greek poems. Long after the ancient poems faded from public consciousness, the Sibyls continued to enjoy celebrity status.

This is why the legendary Sibyls appear among Michelangelo’s paintings on the ceiling of the Sistine Chapel, alongside prophets of the Hebrew Bible. They all get counted as forerunners to Christianity, preparing the way of the Lord.

In the late 1400s, a tract by the Italian humanist Filippo Barbieri gave a tally of twelve famous Sibyls. Barbieri’s tract was the occasion for the first appearance of the short Latin poems that Lassus ended up using in his choral setting. An article by Peter Bergquist (1979) traces the poems to an edition of Barbieri’s tract published around 1505, titled Quattuor hic compressa opuscula (‘four short works compiled here’). This 1505 edition has engravings of the twelve Sibyls, and adds the poems as captions, on facing pages.

Note. Some online facsimiles: Barberi ca. 1505; Birck 1555, an edition of the ancient Sibylline oracles which includes the Latin poems as an appendix; Lassus ca. 1560, the manuscript of Lassus’ choral settings, soprano alto tenor bass (the Prophetiae are around p. 50, with different pagination in each part); and Lassus 1600, the first publication of Lassus’ choral settings (soprano and bass parts). I haven’t been able to trace an online facsimile of Birck’s 1545 edition of the ancient Oracles, which reportedly contains the neo-Latin poems too: Google Books claims to have copies (1 2), but they’re mislabelled.

The unmetrical hexameter

Here’s the faulty text that is used in every copy on the CPDL website, and on nearly every recording: De Labyrintho (2007), Vocalconsort Berlin (2015), Trinity College Music Society (2021), and La Main Harmonique (2022). They all sing the text that Lassus didn’t write.

In teneris annis facie insignis honore
Militiae aeternae regem sacra virgo cibavit
Lacte suo, per quem gaudebunt pectore summo
Omnia, et Eoo lucebit sydus ab ore
Mirificum. Sua dona magi cum laude ferentes
Objicient puero myrrham, aurum, thura Sabaea.

Of tender years, with glorious face, in honour,
the sacred virgin has fed the king of the eternal host
with her milk. All things will rejoice in him
with uplifted heart, and from the eastern face will shine a star
of wonder. The magi will bring their gifts with praise
and offer them to the boy: myrrh, gold, (and) Sabaean incense.

There are actually several textual errors. The simple ones are:

  • Line 2 cibavit: should be future tense cibabit (‘the virgin will feed’, not ‘has fed’).
  • Line 4 ore: should be orbe (‘from the eastern sphere’ or ‘eastern world’, not ‘eastern face’).
  • Line 6 objicient: should be obijcient (same meaning, different pronunciation).

Lassus’ manuscript is held in Vienna at the Österreichische Nationalbibliothek, Mus.Hs. 18744. All four parts have orbe in line 4 and obijcient in line 6. Cibavit is Lassus’ own error: he correctly wrote cibabit in the tenor part, but cibavit in the soprano, alto, and bass parts.

The variation objicient–obijcient is a philosophical difference. Objicient is more linguistically sound, and it appears in the 1555 edition of the poem and in the Lassus texts of Bergquist (1979) and Schlötter (1990). But in the manuscript it’s clear that obijcient is what Lassus wrote; it also appears (correctly) in the 1600 edition. Most likely Lassus was using an edition that had obiicient, which is how the 1505 edition prints it.

But it’s line 1 facie insignis that set me on this trail. Take a look at the following scansions. Macron and breve marks (— and ⏑) represent long and short syllables in the hexameter (not the rhythms used in Lassus’ setting); the anceps mark at line-end (×) indicates indifferent syllable length.

(a) The rhythm of dactylic hexameter. (b) The first line from Lassus’ prelude movement, scanned, and illustrating elision on quae. (c) The faulty first line from the ‘Sibylla Cimmerica’ movement in two possible interpretations, one with hiatus, the other observing elision.

I also give the first line from Lassus’ prelude movement (‘The songs you hear, modulated with a chromatic tenor ...’) to illustrate how the words should correspond to the hexameter rhythm, and also to illustrate the correct observance of elision. The text of this opening movement is likely by Lassus himself. Notice incidentally that this line is what metrical stylists call a ‘golden line’, in the form noun A adjective B verb adjective A noun B. It’s clearly calculated to show off Lassus’ skill with Latin verse, as well as with music.

The scansion of the line from the fourth movement shows that the rhythm cannot possibly correspond to the hexameter rhythm, whether facie is elided or not.

The correct reading is facie praesignis, which is similar in meaning: ‘with exceptionally glorious face’, as opposed to insignis ‘with glorious face’. It doesn’t suffer from the problems with insignis:

  • Praesignis starts with a consonant, so there’s no need to worry about elision.
  • Another principle of hexameter is that any vowel followed by a consonant cluster is treated as metrically long. So before the pr- of praesignis, facie acquires a long final syllable, and the phrase has the correct metrical rhythm, ⏑⏑ — — — ⏑.
The manuscript of Lassus’ Prophetiae Sibyllarum, 4th movement, tenor part, with the correct readings clearly visible. (Vienna, Mus.Hs.18744 iii, p. 27)

There are other occasional differences between the text in Lassus’ manuscript and the poems as they appear in the 1505 and 1555 editions. Lassus’ title for his fourth movement is Sibylla Cymmeria (not Cimmeria). ‘Cimmerian Sibyl’ would be Sibylla Cimmerica, and that’s how it appears in the 1555 edition of the poem: Lassus’ title is literally just the words ‘Sibyl’ and ‘Cimmeria’, which is an infelicity. (The 1505 edition is worse: it prints Sibylla Chimeria in one place, Sibylla Chimicha in another.)

The best published text for Lassus’ settings is Bergquist’s (1979), which is also used in the Bärenreiter edition (Schlötter 1990). Even that doesn’t agree perfectly with the Lassus manuscript. Lassus uses forms like obijcient and eximij, rather than objicient and eximii, and he doesn’t usually capitalise the first letter of each hexameter line. You aren’t going to hear most of these things in performance, though objicient is audible.

The only recordings I’ve found that use the Bergquist–Bärenreiter text are the ones by Cantus Cölln (2002) and Gallicantus (2018). (The latter puts Lassus’ settings alongside some modern compositions — ‘Sibylla Clevelandiae’, ‘Sibylla Chicagonis’, and so on.) The Hilliard Ensemble recording (1998) does have the correct text, but their consonants are so under-enunciated that it takes some careful listening to tell.


  • Barbieri, F. circa 1505. Quattuor hic compressa opuscula. ... Venice. [Internet Archive]
  • Bergquist, P. 1979. ‘The poems of Orlando di Lasso’s Prophetiae Sibyllarum and their sources.’ Journal of the American Musicological Society 32.3: 516–538. [DOI | JSTOR]
  • Castellion, S. 1555. Σιβυλλιακῶν χρησμῶν λόγοι ὀκτώ. Sibyllinorum oraculorum libri viii. Ed. Sixt Birck (alias Xystus Betuleius), Sébastien Castellion (alias Sebastianus Castalio). Basel. [Internet Archive]
  • Lassus, R. circa 1560. Prophetiae Sibyllarum. Manuscript. Vienna Mus.Hs.18744. [Österreichische Nationalbibliothek]
  • Lassus, R. 1600. Prophetiae Sibyllarum. Ab Orlando de Lasso, piae memoriae (Cantus, Bassus). Munich. []
  • Schlötter, R. (ed.) 1990. Orlando di Lasso. Sämtliche Werke (Neue Reihe) 21. Prophetiae Sibyllarum. Kassel. [IMSLP (PDF)]

No comments:

Post a Comment