Three Ravens and Twa Corbies: the transformations of a traditional song from the 17th to the 20th century

There were three Ravens first appeared in print in Melismata, a book of songs compiled by Thomas Ravenscroft, published in 1611. A previous article, There were three Ravens: sublime love and ridiculous riddles, explored the layers of meaning in the song, found through its contemporaneous cultural references, and the comically bizarre interpretations offered by some modern authors.

This second article traces the many transformations of the song in the oral tradition from the 17^th to the 20^th century, with videos, soundfiles and music in staff notation. This includes the version printed by Thomas Ravenscroft in 1611; a different version included in a quodlibet (song medley) by William Cobbold in c. 1620; its Scottish metamorphosis, Twa Corbies, in the 18^th and 19^th century; 19^th century variants in Lincolnshire, Derbyshire, and Scotland; a 19^th century American minstrel show parody, sung to the tune of Bonnie Doon, which then spread in the oral tradition in the USA and England; a 19^th century US version sung to several variants of When Johnny Comes Marching Home; and 20^th century variants sung in Scotland, the USA, and Canada.   

We conclude with some reflections on the changes the song has undergone over the centuries.

The first appearance in print of There were three Ravens

There were three Ravens first appeared in Thomas Ravenscroft’s Melismata, 1611. This was his third book of songs, his first book being Pammelia, 1609, and the second Deuteromelia, also 1609.

There had never been commercially printed books like Ravenscroft’s. In the introduction to Pammelia, he explains his purpose. Songs for the highly skilled singer and musician were being published in lute-song books such as those by John Dowland, but the diversity of music is such that she “bestow[s] her melodious gifts” to “any one whatsoever”, and the more humble forms of music commonly sung among the populace had never been in print. As Ravenscroft put it, it may seem miserly and unkind “in never (as yet) publikely communicating, but always privately retaining … this more familiar mirth and jocund melodie”, so “that fault being now mended, this kind of Musicke also is now commended to all mens kind acceptation.”

Pammelia consists entirely of catches, totalling 100, for between 3 and 11 voices, a catch being a round or canon, meaning everyone sings the same, but begins at staggered intervals, thus creating polyphony. Deuteromelia is divided into “Freemens Songs” for 3 and for 4 voices, and “Rounds or Catches” for 3 and for 4 voices. Melismata is organised differently, into “Court Varieties”, “Citie Rounds”, “Citie Conceits”, “Country Rounds”, and “Country Pastimes”. No reasoning is given for the organisation of Melismata, but the obvious explanation is the origins of the music, the places from which they were collected, as his three books are not of songs he composed, but music collected from various sources. In the introduction, the Apologie, of his 1614 treatise, A Briefe Discourse, Ravenscroft states this, that “those former Harmonies by mee published in my Infancy … those Workes for the most part were not Compos’d by My selfe, but by divers and sundry Authors, which I never the lesse compil’d together, in regard of the generall delight men tooke in them. And although very many of them were Defective in their Composition when they came into my hands : yet according to my knowledge then, I corrected them and commended them to the world.” His statement that he collected and edited rather than composed the contents of the books is verified by the fact that a large proportion of the songs in Pammelia, Deuteromelia and Melismata can be traced back to their antecedents in the 16^th century, as will be demonstrated in the dedicated article on Ravenscroft’s life and work, to be published on this website later this year.

Ravenscroft’s editorial role, his correcting, is not easy to discern, and is not specified in his publications. He states that “those Workes for the most part were not Compos’d by My selfe”, but if any of the pieces were his own composition, they are not marked as such. It could in theory be that some of these songs were monophonic and he added parts to make them polyphonic so that, for example, the harmonies and accompaniment of Ravens were his additions to an originally monophonic song, but this is speculation: it may already have been polyphonic so that he published it unaltered. What is clear is that Ravens was circulating in the oral song tradition, and that around the time Melismata was published in 1611 there was at least one other variant of the song, as we will see below.

There were three Ravens appears in Melismata as one of the “Country Pastimes”, as follows.

We see above that there are 4 voices: treble (which carries the melody), medius, tenor and bassus. There are 10 verses, which I made 5 in the video which begins this article by collapsing every 2 verses into 1, replacing repetitions in odd-numbered verses with the words of the even-numbered next verse. For example, the first 2 stanzas …

There were three Ravens sat on a tree,
Downe a downe, hay downe, hay downe.
There were three Ravens sat on a tree
with a downe,
There were three Ravens sat on a tree,
they were as blacke as they might be,
with a downe derrie, derrie, derrie, downe, downe.

The one of them said to his mate,
Downe adowne, hay downe [hay downe],
The one of them said to his mate,
with adowne :
The one of them said to his mate
Where shall we our breakefast take?
with adowne dery [dery dery downe] downe.

… becomes the first …

There were three Ravens sat on a tree,
Downe a downe, hay downe, hay downe.
they were as blacke as they might be,
with a downe,
The one of them said to his mate,
Where shall we our breakefast take?
with a downe derrie, derrie, derrie, downe, downe.

There is only 1 note for the 2 syllable word “Ravens”. This is not a mistake: in subsequent verses, this note is sung to 1 syllable – “them”, “der” of “yonder”, “lie”, “flie”, and so on – so it would be more confusing in subsequent verses to print 2 notes where there is 1 syllable.

There are some repeated words missing from the treble voice part in the non-lexical vocables (nonsense syllables), reinserted in the typed-out verse above in square brackets. (For an explanation of non-lexical vocables, also called diddling, lilting, mouth music, portaireacht bhéil, port à beul, or jigging, see the previous article on Ravens here.) Similarly, we see below that there is a missing “hay downe” in the medius and a missing “hey downe” in the bassus: the red boxes indicate the notes for which there are obviously missing words, which are easily reinstated.

The differential spelling of “hay downe” and “hey downe”, “a downe” and “adowne”, is to be expected and is not significant. This was the era before standard spelling, when the same writer in the same document might spell the same word more than one way. This was the case even with proper names. William Shakespeare, for example, famously spelled his own name several different ways.

The non-lexical vocables do not always agree between the voices. For example, at the end of each verse, we have the following clashing words – click the soundfile to hear the line.

In the treble: “with a downe derrie, derrie, derrie, downe, downe.”

 

In the medius: “with a downe, hey derrie derrie, downe, downe, downe.”

 

In the tenor: “with a downe derrie, derrie downe a downe.”

 

In the bassus: “with hey downe downe, derrie downe downe.”

 

There is no reason to change this, to assume error and standardise the words to make them agree. Though the words are different, they do fit together musically, as we hear below:

 

Listeners were used to hearing different words sung at the same time since the 13^th century, when there is the first evidence of the catch or round (roundelay), then called a rota (wheel), as evidenced in Sumer is icumen in / Perspice Christicola, a tradition which continued with the sacred rounds of the Llibre Vermell de Montserrat (Red Book of Montserrat) from Catalonia, c. 1399, as well as secular motets with 2 or 3 songs with different melodies and different words sung simultaneously.

As we see on the right, only the non-lexical vocables of the medius, tenor and bassus parts are printed, which we might wrongly assume to be a method to help singers orient themselves, to ensure that they are singing the rest of the unprinted words to the correct notes. However, in common with the other “Country Pastimes” in the book, all for 4 voices, there are insufficient notes for the words not indicated, which therefore cannot be sung. This can only mean that those notes which do not fit the syllables and for which no words are given in the medius, tenor and bassus parts are to be played by instruments. Ravenscroft therefore printed these songs for 4 voices accompanied by 3 instruments, or perhaps 4, if an instrument is to double the treble just as an instrument logically doubles the notes of the non-lexical vocables in the other 3 voices, rather than stopping at every point the voice comes in.

There were three Ravens in the oral tradition

First, we set the terms: traditional music, the oral tradition, folk music, art music.

By traditional music, we mean songs and tunes that are passed from person to person orally, down the generations, hence the oral tradition. A song or tune spreads in this way across counties, countries, and sometimes across continents, as people migrate and take their music with them. This music does not rely on literacy and the written note, but on memory, performance and communal participation.

The origin of a particular traditional song or tune can sometimes be traced to a written form, such as a broadside ballad; but in the majority of cases it is impossible to claim an ‘original version’, as there are usually only variants with unknown origins. The music is passed from person to person, from generation to generation, within families, among friends, at social gatherings, and the many singers and musicians who have sung and played a piece add to, subtract from, and otherwise change aspects of the words and/or music, sometimes through preference, sometimes unconsciously.

This type of music is therefore traditional, being passed down the generations; it is communal, being shared by as many as will and owned by no one; anonymous, being of unknown authorship and communally produced; and it is constantly variable, changing over time and place. It is for these reasons it is called folk music, the music of the common people, sung unaccompanied or with whatever instrument(s) a performer chooses. This makes it distinct from art music, with a known composer, written down in a set form for voices and/or instruments, and always therefore played as written and arranged, as the composer intended.

A song may move between these two worlds of folk and art music, as we see, for example, with fixed and formal orchestral arrangements of traditional tunes, but once it is fixed and final, that version of the song or instrumental tune is not in the oral tradition. This difference between folk song and art song has generated debates among collectors of traditional song: once a song or tune is written down, the danger is that this captures, formalises and ossifies it, stopping the oral transmission process in its tracks, preventing the natural emergence of further variants by those who sing it as written on the page. However, the English folk song collectors of the late 19^th and early 20^th century were of the view that song traditions were dying with the oldest rural generation of their day, since the generations below them were largely ignorant of or apathetic to the existence of their elders’ traditional music. Their assessment was not entirely correct, as pockets of the oral tradition continued, but they were largely right. The song collectors’ visits to old singers resulted in manuscripts, phonograph recordings, and book publications which preserved musical material that would otherwise have been lost forever.

In Thomas Ravenscroft’s time, the early 17^th century, the situation was both different and similar to that faced by the folk song collectors of the 19^th and 20^th century. It was different in that traditional music was alive and well in the country and the city, as Thomas Ravenscroft’s song collections, Pammelia, Deuteromelia, Melismata, and their traceable variant antecedent songs testify. The situation was also similar in that, as Ravenscroft expressed in Melismata, traditional song was considered low music by the practitioners of art music whose songs and consort music were being published, which meant that folk traditions were sung privately but not communicated in printed books, and “that fault being now mended, this kind of Musicke also is now commended to all mens kind acceptation” and, without Thomas Ravenscroft, the versions of the songs he collected and published would have been lost forever, just as the variant folk songs of the 19^th and 20^th century would have been lost without the work of the collectors.

Ravens was in the oral tradition when Ravenscroft collected and published it. Does his act of printing it with a fixed arrangement for 4 voices and 3 or 4 instruments make it art music? On the basis of the definitions above, it did, but it did no harm to its place in the oral tradition, as it continued as a folk song spawning many variants for four and a half centuries after he printed it. This article tracks and compares collected versions of the song from the early 17^th until the mid 20^th century. Where appropriate, pitches of transcriptions have been transposed to allow for easy comparison between melodies.

The versions and variants of the song are presented below in groups, as type A, B, C, etc. Each type is given in chronological order, starting from the earliest date the variant is known, with four caveats:

i. Similar variants are presented together, the earliest version of that variant first.

ii. When a song is collected from a traditional singer there isn’t usually a way of knowing how long the song has existed in that form, so completely accurate chronology is impossible.

iii. To include all collected versions of Ravens would literally be a complete book, and many versions of variant words or tunes differ only in minor details. I therefore give typical examples of all variant types or branches of the song’s family tree.

iv. The classification of a song variant is a matter of judgement. If a version of a song has more or less the same words as another, but is sung to a different tune, is it a different variant or the same? If a version of a song has more or less the same tune as another, but is sung with quite different words, is it classed by the tune or the words? If the words of a song have 3 stanzas in common with another but has 2 new and different stanzas, are they classed within the same variant subcategory or is the latter a new variant? To these questions there are no cast iron solutions, and each case is judged on its own merits.

A. There were three Ravens in Thomas Ravenscroft, Melismata, England, 1611.

We begin by restating the earliest known version. The first verse is sung in the soundfile below as recorded in Melismata, without collapsing two verses into one, as the video does. Thomas Ravenscroft gives no details of his source.

The first verse is heard in the soundfile below.

 

There were three Ravens sat on a tree,
Downe a downe, hay downe, hay downe.
There were three Ravens sat on a tree
with a downe,
There were three Ravens sat on a tree,
they were as blacke as they might be,
with a downe derrie, derrie, derrie, downe, downe.

The following stanzas follow in the same pattern, with new couplets:

The one of them said to his mate,
Where shall we our breakefast take?

Downe in yonder greene field,
There lies a Knight slain under his shield

His hounds they lie downe at his feete,
So well they can their Master keepe,

His hawks they flie so eagerly
There’s no fowle dare him come nie

Down there comes a fallow Doe
As great with yong as she might goe

She lift up his bloody hed
And kist his wounds that were so red

She got him up upon her backe,
And carried him to earthen lake,

She buried him before the prime,
She was dead her selfe ere even-song time

God send every gentleman
such haukes, such hounds and such a Leman

The first article about Ravens gives all the historical and theological references in the song, followed by a précis as follows.

In summary, the ravens represent “the blackness of sin or unfaithfulness” (MS Bodley 764). They wished to predate on the dead body of a knight, who represents Christian sacrifice: he has been faithful to his lady by defending her, and has been slain by so doing. He lay in a green field, symbol of the shortness of life. His body was guarded by his hounds, symbolising the Christian virtues of faithfulness and steadfastness, and his hawks, symbols of holy self-discipline and shedding of past sins. The woman he died for was a fallow doe, meaning a fertile and attractive woman, his leman, his wife, and she was pregnant with their child. She kissed the wounds on his head, their redness a symbol of his Christ-like self-sacrifice, her kissing them a symbol of her saintliness. She repaid the depth of his love with the depth of her own, not leaving his body for carrion-feeders, but she carried and buried his body though heavily pregnant. As a result of the physical strain and emotional heartbreak, she died by the end of the day, repaying his self-sacrifice with her own.

The chief theme is the long-established metaphor of Christ as the lover-knight, and this song takes it one stage further. Thus …

God send every gentleman
such haukes, such hounds and such a Leman

… is a wish for everyone who is faithful to the ultimate degree, who gives their life for another as Christ did, to be rewarded in kind with the faithfulness of such exemplars as the knight’s hawks, hounds, and leman. Given the cultural context of the song, this is expressed in Christian terms: the passion of Christ is imitated by the knight, and the knight’s sacrifice is imitated by his leman. Surely this message of selfless protection of those we love is one that anyone of any background can appreciate. This is sublime love.

B. There were three rav’ns, extract, in William Cobbold, New Fashions, England, c. 1620.

A quodlibet is a song which combines elements of other songs in counterpoint. One such is New Fashions for voices and viols by William Cobbold, organist of Norwich Cathedral. On a foundation of the melody, Browning, and variations on it, the words and tunes of several popular songs are layered over in succession, including (among many more) The leaves be green (which is the verse to the melody of Browning), Greensleeves, The shaking of the sheets, and There were three rav’ns. No details are given of the sources for the songs.

New Fashions is written in British Library Additional MSS 18936-39 and Royal College of Music MS 684. Neither can be dated conclusively, but the first manuscript was probably complete by c. 1620. Since the timing of There were three ravn’s in the quodlibet is made to fit the counterpoint, two versions are given below: first, as it appears in Cobbold’s composition; second, the timing of the assumed original, which is easily deduced.

Both the tune and the non-lexical vocables of There were three ravn’s in New Fashions are different to that recorded by Ravenscroft. Since it includes only the first stanza, the rest of the verses cannot be compared.

Below is a soundfile of the portion of There were three rav’ns in New Fashions, taken from the Fries Lachrimae Consort performance in the video above.

 

Below is a soundfile of the presumed original timing of the song.

 

C. Ther wer three ravens in Sir Robert Gordon, Straloch lute manuscript, Scotland, 1627–29.

Between 1627 and 1629, Sir Robert Gordon, a geographer, cartographer, poet, mathematician and antiquary of Straloch, Perthshire, Scotland, compiled a manuscript of 90 popular tunes arranged in lute tablature. It is now known as the Straloch lute manuscript.

In his copy of Joseph Ritson’s Scotish Song. In Two Volumes, published in 1794, English writer and antiquary Joseph Haslewood (1769–1833) wrote a marginal comment about a manuscript lute book, presented in 1781 by Dr. Skene of Marischal College to Dr. C. Burney, which contained airs “noted and collected by Robert Gordon, at Aberdeen, in the year of our Lord 1627”, one of which is Ther wer three ravens.   

In January 1839, George Farquhar Graham made a complete list of the titles of tunes Sir Robert Gordon wrote lute tablature for, and he copied the tablature for 30 of the tunes, those he “thought most important”. His list of titles included Ther wer three ravens, but he did not copy the music for it. 

The original manuscript changed hands and became the property of one Mr. Chalmers. On Mr. Chalmers’ death, his library was sold. In September 1845, George Farquhar Graham’s friend William Chappel of London wrote to him with the news that he had intended to buy the Straloch lute manuscript in the sale, but was put off by the high asking prices of some of the books. He regretted his reticence, as the Straloch manuscript “went for a mere trifle”. It has since been untraceable.

For our purpose, all that remains now is the title, so we cannot know if the melody was that of Ravenscroft, or Cobbold, or related to either, or an altogether different tune. George Graham’s partial copy of the manuscript is now in the National Library of Scotland, Adv. MS 5.2.18. All of the surviving arrangements by Robert Gordon are lute only, without voice or song words, so it is likely Ther wer three ravens was written in this form. This is the first evidence of the song in Scotland.

D. Twa Corbies, recorded from traditional singers in Scotland from c. 1783.

Our next evidence of Ravens is a Scottish transformation of the story and a re-titling of the song, now Twa Corbies. A corbie in the Scots language is a raven, Corvus corax, considered a bird of ill omen, and sometimes the carrion crow, Corvus corone, or the hooded crow, Corvus cornix.

The earliest evidence of this variant is in one of the manuscripts supplied by Mary Fraser Tytler of Scotland to American folklorist, Francis James Child. The particular manuscript is not specified and could be one of several he lists, with song texts taken down from singers between the dates of 1783 and 1801. Child published Twa Corbies in his The English and Scottish Popular Ballads. This 5 volume work in 10 parts (each volume consists of 2 parts), published between 1882 and 1898, was a collection of the words only of traditional ballads, comparing variant texts. Francis Child was not a collector of folk song – he did not travel to visit singers and notate songs as Cecil Sharp, Ralph Vaughan Williams et al did – but a collator, complier and commentator of previously collected songs. It may seem odd now that the tune, so fundamental to the performance of a song, should be discarded by Child: he was a Professor of English, and his concern was philology, not melody. It was left to Bertrand Harris Bronson to correct Child’s omission in The Traditional Tunes of the Child Ballads, published in 4 volumes between 1959 and 1972.

Professor Child’s classification and numbering system has become a standard for folk song enthusiasts. What he called The Three Ravens is Child Ballad 26, its main entry in his Volume I, with more commentary in the Additions and Corrections of Volumes IV and V.

In the Scots verses of Twa Corbies, all the key players in the drama remain: the ravens or corbies, the slain knight, his hounds, his hawks, and his leman. There is a minor change to the personnel – there are now twa (two) corbies, not three – and to the language – the knight’s fallow doe or leman is now his “lady fair”. The change in the story is radical: the final verse extolling the virtues of the knight’s faithful hounds, hawks and leman is gone, as in the Scottish story they have all abandoned his body to the scavenging ravens. In this version, the non-lexical vocables have been removed.

As I was walking all alane,
I heard twa corbies making a mane;          [mane: moan, complaint]
The tane unto the t’other say,
Where sall we gang and dine today?          [gang: go]  

In behint yon auld fail dyke,                          [auld fail dyke: old turf dyke]
I wot there lies a new slain knight;              [wot: know]
And naebody kens that he lies there,          [kens: knows]
But his hawk, his hound, and lady fair. 

His hound is to the hunting gane,
His hawk to fetch the wild-fowl hame,
His lady’s ta’en another mate,
So we may mak our dinner sweet. 

Ye’ll sit on his white hause-bane,                 [hause-bane: neck bone]
And I’ll pike out his bonny blue een;            [een: eyes]
Wi ae lock o his gowden hair
We’ll theek our nest when it grows bare.   [theek: cover, thatch] 

Mony a one for him makes mane,
But nane sall ken where he is gane;
Oer his white banes, when they are bare,
The wind sall blaw for evermair.

In Twa Corbies, the responses of the faithful hounds, hawks and wife are inverted, turned from a tale of sublime love to one of wretched desertion. Such a fundamental change in the meaning of the story can only have taken place because a singer or succession of singers disregarded the symbolism and meaning of Ravens and wished to change it to a more savage and cynical song, or characters were remembered but verses forgotten and re-composed, so it was reconstructed in the way we see above. Such is the folk process that creates the variants of traditional song.

In his Minstrelsy of the Scottish Border, Vol. III, 1803, Sir Walter Scott replicated the words supplied by Mary Fraser Tytler, but gave them a different origin. Of Twa Corbies, Scott wrote: “This poem was communicated to me by Charles Kirkpatrick Sharpe, esq. jun. of Hoddom, as written down, from tradition, by a lady.” Scott didn’t include the music for this or any other song, nor did he name the lady. In his Volume IV, Child does name her, citing a letter from “C[harles] K[irkpatrick] Sharpe to Scott, August 8, 1802 … “The song of ‘The Twa Corbies’ was given to me by Miss Erskine of Alva (now Mrs Kerr), who, I think, said that she had written it down from the recitation of an old woman at Alva.”” It is so unlikely in an oral tradition that the texts supplied by Tytler to Child and Sharpe to Scott would be identical in every detail that either the source was the same singer in both cases, or conformity was achieved by editorial means, a common practice at the time.

A clearly related version of the Scottish Twa Corbies, but with significant differences and this time complete with music, was published by Alexander Campbell in his Albyn’s Anthology, Volume 2, 1818, “from the singing of Mr. Thomas Shortreed, of Jedburgh, as sung and recited by his mother.” The melody and first verse are below. As you’ll hear, not only is the tune utterly unlike those from England, the rhythm and intervals give it a distinctive Scottish character.

 

The first 3 stanzas are similar in tone and language to the previous version of Twa Corbies: two ravens look for a meal, they see a slain knight and understand that only his hawks, hounds and lady know he’s there, but there is no mention of their desertion.

As I cam’ by yon auld house end
I saw twa corbies sittin thereon,
The tane unto the t’other did say,
O whare sall we gae dine the day?
O whare sall we gae dine the day?

Whare but by yon new fa’en birk,                       [fa’en birk: fallen birch tree]
There, there lies a new slain knight;
Nae mortal kens that he lies there                     [kens: knows]
But his hawks and hounds, and his ladye fair.
But his hawks and hounds, and his ladye fair.

We’ll sit upon his bonny breast bane,                [bane: bone]
And we’ll pick out his bonny gray een;               [een: eyes]
We’ll set our claws intil’ his yallow hair
And big our bow’r, it’s a’ blawn bare.                  [big our bow’r: build our dwelling, i.e. nest]
And big our bow’r, it’s a’ blawn bare.

The final 2 stanzas are completely unrelated to previous renderings of Ravens or Corbies:

My mother clekit me o’ an egg,                          [clekit me o’: hatched me out of]
And brought me up i’ the feathers gray,            [gray: hatchlings are lighter in colour than adults]
And bade me flee where’er I wad,                      [wad: would]
For winter wad be my dying day.
For winter wad be my dying day.

Now winter it is come and past,
And a’ the birds are biggin’ their nests,             [biggin’: building]
But I’ll flee high aboon them a’                           [aboon: above]
And sing a sang for summer’s sake.
And sing a sang for summer’s sake.

Was this the work of a singer who disliked the bleakness of the previous Twa Corbies? Or who had a fondness for birds, so composed new verses about the lives of young ravens? Or who forgot the final verses but liked the song and so created a new ending? The motives of such changes in the oral tradition are unknowable.

In his Ancient Scots Ballads, 1894, George Eyre-Todd prints a version of Twa Corbies with text from Sir Walter Scott’s Minstrelsy of the Scottish Border, which is the same as that in the Mary Fraser Tytler manuscript, both referenced above. The music is referred to vaguely as being from “tradition”, as below.

 

The vagueness of the attribution of the music, and the fact that the text is copied from Sir Walter Scott, together with the fact that the music is so close to that of another well-known song, raises the suspicion that this version of Twa Corbies was the invention of George Eyre-Todd rather than genuinely traditional. The music sounds like a Scotched-up version of Buffalo Gals, which can be heard here.

E. There was three ravens, sung by Mr. Peacock, c. 1803–08, and Harry Richard, 1892, Lincolnshire, England.

We return to England for the next variant, with words that are close to those printed by Ravenscroft. This version was sung by two singers 80–90 years apart, two and nearly three centuries after Ravenscroft. The non-lexical vocables are still intact, but there are fewer of them.

In the Additions and Corrections of his Volume V, Francis Child states: “The copy which follows was communicated by E. L. K. to Notes and Queries, Eighth Series, II, 437, 1892, and has been sent me lately in Manuscript by Mr. R. Brimley Johnson, of Cambridge, England, with this note: “From E. Peacock, Esq., F. S. A., of Dunstan House, Kirton-in-Lindsay, Lincolnshire, whose father, born in 1793, heard it as a boy at harvest-suppers and sheep-shearings, and took down a copy from the recitation of Harry Richard, a laborer, who could not read, and had learnt it ‘from his fore-elders.’ He lived at Northorpe, where a grass-field joining a little stream, called Ea, Ee, and Hay, is pointed out as the scene of the tragedy.””

Since Mr. Peacock was born in 1793 and heard the song “as a boy at harvest-suppers and sheep-shearings”, we may surmise that to have a good memory of it he was between 10 and 15 years old at the time, putting his hearing of it between 1803 and 1808. The labourer Harry Richard must presumably have heard and learned the same or similar version of the song, which he still sang in or near to 1892, 84 to 89 years later.

It is not clear whether the manuscript by Mr. R. Brimley Johnson of Cambridge recorded the melody, but Child did not. Due to the scansion, it cannot have been the same tune as recorded in previous iterations. The words are as follows.

There was three ravens in a tree,
As black as any jet could be.
A down a derry down

Says the middlemost raven to his mate,
Where shall we go to get ought to eat?
A down a derry down

It’s down in yonder grass-green field
There lies a squire dead and kill[e]d.
A down a derry down

His horse all standing by his side,
Thinking he’ll get up and ride.
A down a derry down

His hounds all standing at his feet,
Licking his wounds that run so deep.
A down a derry down

Then comes a lady, full of woe,
As big wi bairn as she can go.
A down a derry down

She lifted up his bloody head,
And kissd his lips that were so red.
A down a derry down

She laid her down all by his side,
And for the love of him she died.
A down a derry down

Nearly all the essential characters remain from Ravenscroft’s version of 1611, now sung nearly 200 and nearly 300 years later. We still have the ravens, the killed man, the hounds, and the pregnant lady. By this time, all the cultural associations of the symbolic animals and the knight as a symbol of Christ’s sacrificial death (see the first article) have faded from knowledge, to the extent that the singer, Harry Richard, thinks the song is about a real event that happened locally. In this version, the man has been updated from a slain knight to a squire, and his hawks – associated with the gentleman knight of the 16^th and 17^th century – have been replaced in the 19^th century by a squire’s horse. The pathos of the “lady, full of woe” remains, as does the message in the final verse that “for the love of him she died.”

F. The Three Ravens, sung by John Holmes, Derbyshire, England, c. 1825.

We stay in England for a variant with words that again are recognisably related to those printed by Ravenscroft, sung over 200 years later, retaining non-lexical vocables that are different to Ravenscroft’s, but clearly descended from them.

This version was published in Frank Kidson, Traditional Tunes, 1891. Frank Kidson commented, “I am favoured with the copy here presented by Mr. John Holmes, of Roundhay, who first heard it about 1825 from his mother’s singing. This was in a remote village among the Derbyshire hills, most aptly named Stoney Middleton … Mr. Holmes tells me that he heard a Danish gentleman play this same tune, so familiar to him, on the violin, and was informed by him that it had been current among the people in Denmark.”

The first stanza and melody are as follows.

 

Frank Kidson states that the words are “by no means complete”. This could sometimes mean the singer had not been able to remember all the verses, as appears to be the case here, with words missing in the second stanza (where Ravenscroft’s “As great with yong as she might goe” would be a perfect fit); or sometimes a collector would only notate the words of the first verse and melody if the remaining words were close to those of an already-collected version of the song.

Kidson prints the following three verses, which retain the barest bones of the story in Ravenscroft, with the addition of the “lady full of woe” cursing her true love’s killer.

There were three ravens on a tree,
A-down, a-down, a derry down,
There were three ravens on a tree,
Heigh ho.
The middlemost raven said to me,
There lies a dead man at yon tree,
A-down, a-down, a derry down,
Heigh ho.

There comes his lady full of woe,
A-down, a-down, a derry down,
There comes his lady full of woe,
Heigh ho.
There comes his lady full of woe,
. . . . . as she could go,
A-down, a-down, a derry down,
Heigh ho.

Who’s this that’s killed my own true love,
A-down, a-down, a derry down,
Who’s this that’s killed my own true love,
Heigh ho.
I hope in heaven he’ll never rest,
Nor e’er enjoy that blessed place.
A-down, a-down, a derry down,
Heigh ho.

G. The Two Ravens, rewritten by Allan Cunningham, Scotland, 1825, circulating in the USA until the 1950s.

Allan Cunningham (1784–1842) was a Scottish poet who published a version of the song, without music, in his The Songs of Scotland, Ancient and Modern. Vol. I, 1825. It is clearly Cunningham’s own work, based on the traditional words in verses 1, 3, 4 and 5, with his own original verses 2 and 6, but he is coy about admitting it: “The present version is made up from various readings and recitations. It is difficult to say how much is of England or of Scotland; or how much is new, or how much is old” (p. 291).

There were two ravens sat on a tree,
Large and black as black might be,
And one unto the other ‘gan say,
Where shall we go and dine to-day?
Shall we go dine by the wild salt sea?
Shall we go dine ‘neath the greenwood tree?

As I sat on the deep sea sand,
I saw a fair ship nigh at land,
I waved my wings, I bent my beak,
The ship sank, and I heard a shriek;
There lie the sailors, one, two, three:
I shall dine by the wild salt sea.

Come, I will show ye a sweeter sight,
A lonesome glen and a new slain knight;
His blood yet on the grass is hot,
His sword half drawn, his shafts unshot,
And no one kens that he lies there,
But his hawk, his hound, and his lady fair.

His hound is to the hunting gane,
His hawk to fetch the wild fowl hame,
His lady’s away with another mate,
So we shall make our dinner sweet;
Our dinner’s sure, our feasting free,
Come, and dine by the greenwood tree.

Ye shall sit on his white hause-bane,
I will pike out his bonnie blue een;
Ye’ll take a tress of his yellow hair,
To theak yere nest when it grows bare;
The gowden down on his young chin
Will do to rowe my young ones in.

O cauld and bare will his bed be,
When winter storms sing in the tree;
At his head a turf, at his feet a stone,
He will sleep nor hear the maiden’s moan;
O’er his white bones the birds shall fly,
The wild deer bound and foxes cry.

In his Minstrelsy: Ancient and Modern, 1827, the Scottish author William Motherwell printed a version of Ravens with music (described below, variant H) and Corbies without music. Motherwell’s Corbies is simply Allan Cunningham’s work, The Two Ravens, but he called it “a Scottish ballad”, didn’t credit Cunningham, and changed Cunningham’s first line, “There were two ravens sat on a tree”, to “There were twa corbies sat on a tree”.

The Motherwell modification of “two ravens” to “twa corbies” then appeared in Charles Dexter Cleveland’s A Compendium of English Literature, published in Philadelphia in 1862, with the English-Scots language now meeting in the middle as “two” rather than “twa corbies”. Cleveland calls it “One of the most poetical and picturesque ballads existing” and, in a book of English literature, makes no mention of its Scottish language or Allan Cunningham’s hand in rewriting it.

Henry W. Shoemaker, Mountain Minstrelsy of Pennsylvania, wrote that this version – again without mentioning Cunningham’s name – entered the oral tradition, that it was “Long popular in Clinton County” and “One of [singer] Clarence Walton’s favorites.” However, no melody is given, and the text is identical to the previously-printed Cunningham verses, which is not credible for a song circulating in the oral tradition.

The Cunningham verses also appear, once again uncredited, in Mellinger E. Henry’s Still More Ballads and Folk-Songs from the Southern Highlands, 1932, again without music. In this case, the author was aware the piece was not traditional, and therefore debated its inclusion in the book. He wrote (p. 8–10):

“The Twa Corbies. Obtained from Mrs. Henry C. Gray … Terre Haute, Indiana. The ballad, therefore, does not come from the southern highlands, but came as a result of meeting Mrs. Gray while on a ballad-quest in the Great Smoky Mountains. It was copied by the writer from an end-paper apparently of an old bound volume of magazines in the possession of Mrs. Gray. Just as the text was about to be sent to the printer, Mr. Phillips [sic] Barry pointed out that it is identical with the version in Cleveland’s Compendium [Cunningham’s verses, uncredited]. It was then decided not to reprint the text. However, on Mr. Barry’s suggestion it is again printed. He says in a letter of June 26, 1931: “It seems to me that Mrs. Gray’s text of ‘The Two Corbies’ might well be included in your collection with the other two texts [both of them versions of the variant Billy Magee Magar – see variant K below]. The use of Cleveland’s Compendium was so universal in American high schools that it is not likely that Mrs. Gray’s grandfather was the only person who learned the ‘Two Corbies’ from it. There is, after all, not so very much difference between a school-book and a broadside or a songster, when it is a question of giving a particular song text a new start in oral tradition.” As Mrs. Gray was not certain that the ballad was copied down by her grandfather, it will be just as well to quote what she has to say about the song: “I am very much afraid I can’t help you greatly on the ‘Twa Corbies’. As I have told you, it was in an old volume of bound magazines, that was among those given me by my quaker great-aunt … She read constantly and remembered all she read. She was a great one for clipping, and her books are full of clippings. This book that the ballad was found in, I believe, was among the hundred or so she bought of a church. Some one in a town north from here willed a lot of books to a church. They were stamped ‘Good Shephard [sic] Library, Linton, Ind.’ Some way or other they got down here to St. Luke’s, an Episcopal mission. They were such books that the rector thought were not altogether fitting for a church library and at a church sale one time he sold them all. My Aunt, true to form, bought them all. This end-paper may have been in the book when she bought it or she may have put it in for safe keeping. It appears to be a fly leaf of an old volume; the hand writing resembles hers a tiny bit … If it is Aunt Libbie’s writing it was written at least thirty-five years ago.”

The Cunningham verses appear again, this time with music, in Richard Chase, American Folk Tales and Songs, 1956. He wrote: “This extraordinarily good text came to me through Mrs. Willard Brooks, now of Washington, D. C. She could not remember where she learned it. Artus Moser collected it on Gashes Creek, Hickory Nut Gap, near Asheville, North Carolina, and Annabel Morris Buchanan has found it in Virginia. An “original” text appeared in a high school book in 1859, A Compendium of English Literature by Charles D. Cleveland.”

Richard Chase’s use of quotation marks for “original” indicates that he did not know that Charles Cleveland in 1862 simply reprinted Allan Cunningham’s poem of 1825, Cunningham’s own work based on traditional material, so those quotation marks (which grammatically should be inverted commas) are misplaced: it was an original text, not traditional. The words Richard Chase prints are simply Cunningham’s poem, but with spelling anglicised. It is not credible that the words would remain static in the oral tradition, which raises the likelihood that his source, Mrs. Willard Brooks, sang from printed or handwritten words, or that Richard Chase himself used a printed source for the words: it was common practice for folk song collectors to edit, clean up, and interpolate verses not from the source singer. There is no clear evidence, then, that the Cunningham version was ever sung in the oral tradition.

The tune used, presumably that sung by Mrs. Willard Brooks, is taken from a black-face minstrel version, sung in a different style and to different words, circulating from 1863. This is described below, variant J. The tune is Scottish, The Caledonian Hunt’s Delight, also known as Ye banks and braes o’ bonnie Doon, or Ye banks and braes, or Bonnie Doon, as we will hear below.

H. The Three Ravens, in William Motherwell, Minstrelsy: Ancient and Modern, Scotland, 1827.

In the Introduction to his Minstrelsy: Ancient and Modern, 1827, William Motherwell makes clear that the contents of his book “are to be understood as referring exclusively to the Ancient Romantick and Historick Ballad of Scotland.” Since his book includes a version of both Three Ravens and Twa Corbies, this should indicate that both branches of the song were sung in Scotland, but all is not as it seems. As noted above, Motherwell’s Corbies is simply Allan Cunningham’s rewrite, The Two Ravens, but with “two ravens” in the first line changed to “twa corbies”, and without crediting Cunningham.

Motherwell’s Three Ravens was collected for him by Andrew Blaikie of Paisley, Scotland. It “is given from the singing of a traditionary version of the ballad very popular in Scotland, and the words of which set differ little from those given in Ritson’s Ancient Song” (p. xviii). This is a reference to Joseph Ritson, Ancient Songs and Ballads, published in 1790, which simply reprints the words published by Thomas Ravenscroft in Melismata, 1611. Both Ravenscroft and Ritson are discussed in detail in the first article.

Motherwell gives the very fine melody below and only the first verse, so the rest of the song is not available for comparison with other versions. There is no mention of the name or location of the singer, nor the date of collection.

 

I. Variants missing from William Chappell, Popular Music of the Olden Time, 1859.

There is an oddity in tracing the various permutations of Ravens and Corbies in the commentary by William Chappell in his Popular Music of the Olden Time, 1859. Chappell cites Ritson’s unevidenced and vague assertion of the oldness of the song prior to its publication by Ravenscroft in 1611 (discussed in the first article on Ravens), then adds that “It is nevertheless still so popular in some parts of the country, that I have been favored with a variety of copies of it, written down from memory; and all differing in some respects, both as to words and tune, but with sufficient resemblance to prove a similar origin” (p. 59).

Does this include any variants not previously published? We cannot know, as Chappell doesn’t print any of them. And the music he gives for Ravens is the treble line from Ravenscroft’s Melismata, 1611, with his own arrangement for the other three voices. This means that in a book entirely about historical music, he dismisses the historical arrangement in favour of his own, and fails to include traditional variants he has access to.

J. The Three Crows, minstrel version, USA, 1863, collected from oral tradition in the USA, Canada and England until c. 1960s.

We have seen that the oral song tradition was the source for Ravens as printed by Thomas Ravenscroft in 1611, and that a different version from tradition was included in William Cobbold’s quodlibet in 1620 or earlier. In 1863 and 1868, the reverse process took place: a printed version entered the oral song tradition and, unlike the claims about the Allan Cunningham verses, the evidence is clear and credible.

E. Byron Christy and William E. Christy’s book, Christy’s New Songster and Black Joker, 1863, was a tie-in to the successful stage act of Christy’s Minstrels, formed in 1843 in New York by Edwin Pearce Christy. Their act was a variety show of songs, dances and comic turns, performed by white men in burnt-cork black-face. It is a measure of their success and popularity that their run at Mechanics’ Hall, New York City, ran for over 7 years from March 1847 until July 1854.

After the original Christy’s Minstrels disbanded, many of the members formed a new troupe led by J. W. Raynor and Earl Pierce, called Raynor & Pierce’s Christy Minstrels. They toured the cities and provinces of England from 1857 to 1860, including some extended residencies, 10 months at the Polygraphic Hall in London, 4 months at St. James’s Hall, Liverpool.

The show evidently featured The Three Crows, printed in the book without music, shown below. Anyone buying the book would have been familiar with the melody from the stage show and, as we will see, another publication included the song with the music.

The purpose of speaking each verse followed by singing the same was parody, a mocking imitation of a church practice known as lining out. Among congregations that either could not afford books or could not read, lining out was a song leader speaking the line of a hymn which the congregation then sang. Over time, this became the leader chanting or intoning the line which the congregation then sang. That the song was about three scavenging crows rather than Jesus was the point of the intended humour.

The same song, performed in the same satirical manner, but with more verses, was included in another black-face minstrel book, Frank Brower’s Black Diamond Songster, also in 1863. Like Christy’s New Songster, Frank Brower’s book was the merchandise of a minstrel show that toured the USA and the British Isles. Brower’s version of the song is called The Four Vultures, despite the song being about three crows: did the four singers named in the book dress as or otherwise imitate vultures for the performance? The book notes that it was “Always received with shouts of laughter”, presumably at its mockery of church practice. As we see below, it was printed without music.

Five years later, in 1868, the song appeared again in Henry Randall Waite’s book, Carmina Collegensia, a compendium of songs sung in colleges. Three Crows is complete with music in 4 part harmony and an indication that it should be lined out. There are 4 verses, the same in content as in Christy’s New Songster, but worded differently.

 

Since the scansion is the same and the manner of singing the same in the three minstrel publications, it is a fair assumption that the tune is the same in each. The melody is a portion of a Scottish tune that went by several titles. In James Manson’s Hamilton‘s Universal Tune Book, Vol. II, published in Glasgow in 1846, it was The Caledonian Hunt’s Delight, credited to James Miller and Stephen Clarke (below). Since Scottish poet Robert Burns (1759–96) used it as the melody for his verse, Ye banks and braes o’ bonnie Doon, it was also known by that title, or by one of two shortened titles, Ye banks and braes or Bonnie Doon.

 

The idea to use The Caledonian Hunt’s Delight for the Three Crows parody probably came from its previous use as the melody for a shape note hymn, When Marshall’d On The Nightly Plain, in The Hesperian Harp, a hymn book published in 1848 by Dr. William Hauser of Jefferson County, Georgia, in which the tune is called Bonnie Doon, shown below.

The minstrel parody of Ravens or, in this case, Crows, entered the oral tradition and was widely collected in the USA and England, with or without the lining out. There follow some examples selected from the many.

John Harrington Cox of Morgantown, Monongalia County, West Virginia, was the Archivist and General Editor of the West Virginia Folk-Lore Society. In his book, Folk-Songs of the South, 1925, he published a version he learned from his father in Illinois in 1915, which combines the Bonnie Doon parody melody with a refrain related to the Billy Magee Magar (variously spelt) variant from 1868 on, which used the melody of When Johnny Comes Marching Home (for which, see variant K below).

 

There were three crows sat on a tree,
And they were black as black could be.
Philly McGee McGaw!

One of them said unto his mate,
What shall we do for meat to ate?
Philly McGee McGaw!

There lies a steed on yonder plain
That by his master has been slain.
Philly McGee McGaw!

We’ll perch ourselves on his backbone
And pluck his eyes out one by one.
Philly McGee McGaw!

In Orford, New Hampshire, USA, on 19^th November 1942, Jonathan Moses sang the following variant to the collector, Marguerite Olney. She noted it as The Three Crows, probably because other versions mention that number, but the words sung by Jonathan Moses mention no number, so I have named it after the opening words, as is common practice. The melody is still clearly based on Bonnie Doon, with some variations. 

 

This is a very simplified version of the story, with only three stanzas and much repetition, each verse having only two lines, each of which is sung twice and three times respectively. These lines will now be familiar from previous variants. Without the repetition, the stanzas are:

Oh one old crow said to his mate
What shall we do for grub to eat?

There is a horse on yanders plains
already dressed, already slain.

We’ll perch us and pick us on his bare backbone
And pick his eyes out one be one.

Three black crows, sung by William Nelson of Kinsac of Nova Scotia, Canada, is clearly related, and reintroduces non-lexical vocables. His version was published in 1950, among the songs collected by Helen Creighton and Doreen H. Senior in Traditional Songs from Nova Scotia.

 

Three black crows sat on a tree
Ca beelya geelya gaw ye
And they were black as crows could be
Ca beelya geelya gaw ye.

One black crow said unto his mate
What shall we do for something to eat?

An old red horse in yonder lane
Who very lately has been slain

We’ll pick his eyes out one by one
And pick the meat from off his bones

It was collected many more times: twice, for example, by collector Helen Hartness Flanders, sung to her in Autumn 1934 by Mrs. Ralph Edwin Jacquith of Vermont, USA, learned from her father, Henry W. Wallace, of Woodstock, and in the early 1930s by the unnamed mother and brother of Miss Sylvia Bliss of Vermont.

It was also sung in England, which is unsurprising since, as noted above, Raynor & Pierce’s Christy Minstrels toured the cities and provinces of England. Charlie Clissold of Gloucestershire was recorded by Mike Yates on an unspecified date between 1964 and 1985, included on The Horkey Load, Volume Two, released by Veteran Tapes in 1988 (VT 109). In this version, Charlie sings solo and lines out as if the choir are with him, with a new final stanza.

 

There were three crows sat on a tree,
and they were as black as crows could be.

And one old crow said to his mate,
What shall we have this day for bait?

They flew across the burning plain
to where an oxen had been slain.

They perched upon his big backbone
and pecked his eyes out one by one.

And this old crow flew into a tree and said,
You old bugger, you shan’t shoot me.

In this version, the story is even more disintegrated than all previous variants: the slain knight, later a squire, later a horse, is now an oxen, and the only players remaining in the drama are the three ravens, now crows. Lyrically, these versions from the USA, Canada and England follow the pattern of many variants collected in the USA, leaving only the slightest trace of the early 17^th century story and none of its meaning.

K. The Billy Magee Magar variant, USA, 1868, collected from the oral tradition in the USA until 1945.

The previously referenced Henry Randall Waite compendium of college songs, Carmina Collegensia, 1868, not only has the church parody Three Crows sung to Bonnie Doon, given above, it has another version: Crow Song, in 4 part harmony, the tune based on that for the American Civil War song, When Johnny Comes Marching Home, which was first published 5 years earlier in 1863.

 

This version became widespread in the USA, often sung, often collected, often printed, with various spellings of the refrain, “Billy Magee Magar”. There are far too many versions, only slightly different, to be interesting, so what follows is a small sample of the most noteworthy variants.

There were three crows, sometimes named by the refrain, Billy MacGee MacGore, was sung by Mrs. Jane Gentry of Hot Springs, North Carolina, to folk song collector Cecil Sharp on 27^th July 1917 (in Cecil Sharp MSS 3808/2799).

 

The subsequent verses extend the details of the Carmina Collegensia version slightly: there is a crow talking to his mate, a horse lying on yonder plain, by some cruel butcher slain, they eat it before it’s stale, till naught remains but bones and tail. The 17^th century story has been reduced to non-existence.

The same words with only minor differences were given in a manuscript on 8^th December 1941 by Mrs. Alice Robie of Pittsburg, New Hampshire, to collector Marguerite Olney. The handwritten sheets had belonged to Alice Robie’s sister, Mrs. Frank Luther. There is no melody, but the words would fit the tune above without the “Caw caw caw” and the repeat of “And they all flapped their wings and cried”.

A variant with very similar words but a different scansion was sung by Charles Tillett of Wanchese, North Carolina, printed in Louis W. Chappell, Folk-Songs of Roanoke and the Albemarle, 1939. The text is noted as 1924 and the tune as 1935, with no explanation for the discrepancy. The words are similar to those in Carmina Collegensia, with only minor changes, and the melody is clearly based on When Johnny Comes Marching Home, but with significant and interesting deviations, as follows.

 

Another rendering, with one fewer lines and no tune noted, was sung by Mrs. Myra P. Daniels of Hardwick, Vermont, to Marguerite Olney, on 1^st July 1954. Mrs. Daniels said “I heard it sung often as a child”, but she didn’t have a recollection of where she learned it. She probably heard it at school or at a party, as it appears exactly as she sang it in Benjamin Albert Botkin’s book, The American Play-Party Song, 1937, with the following instructions but without music (p. 63): “Here the players form a circle and move around three in the centre who are the crows and sit flapping their wings and cawing, to the singing of the lines:

There were three crows sit on a tree,
Billy Magee, Magaw.
And they were black as crows could be,
And they all flapped their wings and cried
Billy Magee, Magaw,
And they all flapped their wings and cried,
Caw! Caw! Caw!

If this was a game, the rules of the game, its progression and resolution, is not explained.

A further variant in this group was sung to Marguerite Olney on 12^th August 1945 by Amos Eaton of South Royalton, Vermont, learned from his mother. The melody is a variant of that above, still recognisably related to When Johnny Comes Marching Home.

 

Lyrically, the version sung by Amos Eaton in 1945 is very simple. It has only four stanzas, which are very close to the minstrel version sung to Bonnie Doon from 1863 (see above, variant J), and to those sung to a different melody and recorded by Kenneth Larson, learned in 1912 (see below, variant M). If we remove the repetitions and refrains, Amos Eaton’s words are:

There were three crows sat on one tree
And they were black as crows could be

Said one old crow unto his mate
What shall we do for grub to ate?

There lies a horse on yonder plain
Who was by some cruel butcher slain

We’ll perch ourselves on his backbone
And peck his eyes out one be one

L. Three Crows, sung by Mrs. McLeod of Dumfries, Scotland, 1906.

This version was collected in 1906 by Claude H. Eldred, an undergraduate at the University of Wisconsin, from Mrs. McLeod of Dumfries, Scotland, who was visiting relatives at Lake Mills, Wisconsin. She had learned the song from her parents. Neither she nor her parents could read or write, so in both cases it can only have been learned orally. If the music was written down or recorded, it did not appear in Arthur Beatty’s article report of it in The Journal of American Folklore (see bibliography).

This version is noteworthy because:

(i) from the 1611 story, it retains black corvids in a tree discussing where they will eat, spotting a body, and something of the “God send every gentleman” final verse, now “God grant that each lambkin”;
(ii) it retains non-lexical vocables;
(iii) the humans have been removed from the story, with the effect that instead of being an exhortation to self-sacrificial love, as in the Ravenscroft version, it is a story of crow predation any rural villager would recognise.

Three crows there were who sat on a tree,
Fa la, hay down, derrie down down,
And they were as black as crow ever could be,
Fa la, hay down, derrie down down.

With the same non-lexical vocables for the second and fourth lines, subsequent verses are:

Then one of them said to his mate on the tree,
Where do you think our supper will be?

Down in the meadow behind a gray stone,
A newly born lambkin is lying alone.

We’ll pluck his eyes out of him one by one,
And let him go blinding ere day is done.

But a gay little birdie far down in the tree
Heard them declare that the lambkin should dee

So quickly she flew to the cold gray stone
Where the poor little lambkin was lying alone

And told him that quickly away he must flee
Or else the three crawens would make him to dee.

God grant that each lambkin that is in our flock
Be told of his danger as he lies by the rock.

M. Two Crows, learned by Kenneth Larson, Songs and Ballads, USA, c. 1912.

Kenneth Larson, a High School Teacher in McCammon, Idaho, USA, typed out a collection in c. 1933 he called Songs and Ballads (Folk Material and Old Favorites) which, as far as I have been able to find, has never been commercially published, and is now in the Idaho State Archives. In the Preface, he states, “I have used no sources other than those of oral tradition, by which all folk literature is necessarily secured; I have scorned referring to song books or to the columns of newspaper captializing on the modern fad of collecting and reviving old songs.” He includes the verses and music of Two Crows (A South-Idaho Version), of which he comments, “Learned in childhood, about 1912, at Malad, Idaho, from my parents.”

 

There were two crows sat in a tree,
In a tree,
There were two crows sat in a tree,
In a tree,
There were two crows sat in a tree,
In a tree;
As black as any crows could be,
Crows could be.

With the same pattern of repetition, subsequent verses are:

The one crow whispered to his mate,
Have you seen anything to eat,

There lies a horse in yonder field,
And there we’ll have a merry meal,

We’ll perch upon his old backbone,
And peck his eyes out one be one.

As we have seen, the Lincolnshire version (variant E), sung c. 1803–08 and 1892, replaced the knight with a squire and his hawks with a horse. Now, in Idaho in 1912, only the horse remains. The ravens are now crows, and missing are the knight or squire, his hounds, his hawks, his leman or lady, the final message, and the corpse is a horse rather than a man. In this, it follows the minstrel and the Billy Magee Magar variants. It is at best a degenerate version of There were three Ravens or Twa Corbies, its story obliterated and its melody reminiscent of a nursery song.

N. Two Crows, sung by Patrick Ward Gainer’s grandfather, USA, c. 1909–14.

In his Folk Songs from the West Virginia Hills, 1975, the collector and editor Patrick Ward Gainer includes a version of Two Crows “My grandfather Gainer used to sing this song to me when I was small.” Since the author was born in 1904, we can place this at c. 1909–14.

The words of the verses are, the reader will no doubt see, standard fare by now, except the non-lexical vocables, which are unusual. The simple and repetitive tune is unique to this version.

 

There were two crows sat on a tree,
Tattery nan de tario,
And they were black as crows could be.
Faira nay tareno, tattery nan de tario.

And the one he said unto his mate,
What shall we do for grub to ate?

There lies a man on yonder plain,
Whose body has been lately slain.

We’ll perch upon his long backbone,
And pick his eyes out one by one.

O. Three Old Crows, Three Crows and Two Crows, sung in the Southern Appalachians, USA, 1916 and 1918.

In their book, English Folk Songs from the Southern Appalachians, 1917, Olive Campbell and Cecil Sharp include a song they call The Three Ravens, despite the fact that the singer Ben Burgess sings of crows. He sang Three old crows (my title, since that’s what the song says) on 28^th September 1916 in Charlottesville, Virginia. It can be dated back several decades, as he learned it as a boy from his Italian great-grandfather, Genini. Genini came to the USA in 1789 at the invitation of Thomas Jefferson, at the end of Jefferson’s travels in France and Italy, so that he could introduce Italian methods of viticulture (winegrowing) to the USA. Ben Burgess remembered the tune and only some of the words of the first verse.

 

The revised edition of English Folk Songs from the Southern Appalachians, published in 1932, is quite different, now credited to Cecil J. Sharp, edited by Maud Karpeles, and it includes two new versions of the song, each with music. I am grouping these three tunes together because they have features which indicate a common melodic ancestor, made more likely by the fact that they were all collected in the same region.

Queenie Woods remembered only the first stanza, as follows.

 

A day later, Ada Maddox sang a related melody. The idiosyncratic timing is how it is notated in the book, doubtless reflecting the fact that traditional singers unaccompanied by an instrument can be rather free with their timing.

 

There were two crows sat on a tree,
Lardy hip tie hoddy ho ho,
There were two crows sat on a tree,
And they were black as crows could be.
Lardy hardy hip tie hoddy ho ho.

The old he-crow said to his mate:
What shall we have today to eat?

There lies a horse in yonders lane,
Whose body has not very long been slain.

We’ll press our feet on his breast-bone,
And pick his eyes out one by one.

These three melodies suggest a dependent relationship or a common ancestor. The tune sung by Queenie Woods is a version of the Ada Maddox tune, with very similar non-lexical vocables, and melodic repetitions by Queenie that appear to be the result of either Queenie, or the person she heard it from, simplifying a version of the Ada Maddox melody. All three tunes heavily feature a repeated interval of a third. In these transcriptions, transposed to the same key signature for easy comparison, there is a repeating major third in all tunes from g’ to b’. In the tune sung by Ben Burgess there is also a repeating minor third from d’’ to f’’ in his second section, and in Ada Maddox’s tune a repeating minor third from b’ to d’’ in her second section.

P. Three Crows, sung by Lucy and Mary McAllister, Virginia, 1935.

Four variants, here designated P and Q, are as good examples as any to illustrate how the folk process works in the oral tradition.

The manuscripts of Professor Winston Wilkinson, now kept by the University of Virginia, comprise songs he collected while research fellow in folk music at the University of Virginia in 1936–37. This includes two closely-related versions of the song, sung by two closely-related people, Mrs. Lucy McAllister and Mrs. Mary McAllister, presumably the wives of two brothers. Their versions were collected 21 days apart, and illustrate how words and melody become personalised to the performance style of the singer. When there are enough of these small changes in rhythm and pitch when passed from singer to singer, eventually there is a melody entirely different to that in the beginning.

Mrs. Lucy McAllister’s Three Crows:

 

Mrs. Mary McAllister’s Three Crows:

 

Q. Two Crows, sung by James Chisholm, Virginia, 1918 and 1936.

The same point about the development of variants in the oral tradition can be made with the same singer. On 21^st May 1918, Cecil Sharp wrote down the words and melody for Two Crows, as sung by James Chisholm of Nellysford, Virginia. 18 years later, on 12^th March 1936, Winston Wilkinson visited the same singer, now in Greenwood, Virginia, and wrote down the same song, with a melody that had changed and the verse extended over the intervening years.

As James Chisholm sang it in 1918 …

 

… and as he sang it in 1936.

 

Anyone who sings traditional songs from memory will, like me, likely have had the same experience. We learn a song and sing it for, say, 5 years, then it falls out of the repertoire. 5 years later we try to revive it but find we have forgotten a portion of the music. If it wasn’t written down, we would have to fill in the gap with the notes and words we think best. If it is written down, we return to the written notation from 10 years ago, and find to our surprise that what we first learned had changed in our performance over time, without us realising.

R. Two Crows and Three Blackbirds, Connecticut and Vermont, USA, collected 1939 and 1942.

Mrs. G. C. Erskine of Cheshire, Connecticut, USA, sang this version to Helen Hartness Flanders on 1^st October 1939. It was learned from her grandmother, Orinda Townsend of Dixfield, Maine, born in 1828. This is a good example to show that one can date the collection of a song from the oral tradition, but finding a date for the origin of a song variant is usually impossible. If we could date when Mrs. Erskine learned it from her grandmother, we would then have to find out whether she sang it exactly as she learned it, or whether she changed it in some way, which would be an impossible task. Even if we could show that she had made no changes, we would then have to trace where her grandmother Orinda Townsend learned it, try to answer the same question, and so on back in time, so the impossibility of finding a date of origin multiplies.

 

There were two crows sat on a tree
Hi-dum, die-dum derry I-O
There were two crows sat on a tree
Hi, derry O
There were two crows sat on a tree
And they were black as crows could be
Hi-dum, die-dum, derry I-O

The succeeding stanzas follow the pattern of the first stanza.

Said one old crow unto his mate
What shall we have for grub to ate?

There lies a horse on yonder plain
’Tis is just six weeks since he was slain

We’ll sit upon his bare backbone
And pick his eyes out one by one

The next stanza retains the sorrowful woman omitted from many variants examined so far, though her lover is not mentioned until the following verse. 

There comes a lady full of woe
She’s full of grief as she can go

She sat down by the horse’s side
And for the love of his rider died

The last two stanzas are noteworthy. The language of Mrs. Erskine’s couplet of 1939 …

There comes a lady full of woe
She’s full of grief as she can go

… is an echo of Ravenscroft’s couplet of 1611 …

Down there comes a fallow Doe
As great with yong as she might goe

Mrs. Erskine’s final couplet is reminiscent of so many final verses in traditional song about tragic individuals or couples who die for love …

She sat down by the horse’s side
And for the love of his rider died

… and is very close to the final stanza of the Lincolnshire version above (variant E), sung by Mr. Peacock in c. 1803–08 and by Harry Richard in 1892:

She laid her down all by his side,
And for the love of him she died.
A down a derry down

In Cadyville, New York, USA, on 18^th June 1942, Mrs. Lily Delorne sang a close variant of the Connecticut version to the song collector, Marguerite Olney. Mrs. Delorne stated that the song was handed down in her family from Starksboro, Vermont. Her version has a very similar scansion to that above with very similar non-lexical vocables, so although no melody was recorded, we could reasonably surmise it was sung to the same tune or one closely related.

Now the ravens or crows have become blackbirds, and the dead man with a horse and hounds is not a knight or a squire, but a soldier. The words are different enough to be worth citing in full, particularly as it retains significant elements of the 1611 version recorded by Thomas Ravenscroft, and ends with two traditional floating verses typical of songs of tragic lovers, such as Barbara Allen.

There were three blackbirds on one tree,
I-dum, I-dum, derrie I-aye.
There were three blackbirds on one tree,
I-dum, derrie I-aye.
There were three blackbirds on one tree,
Saying, Where shall we go dine today?
I-dum, derrie I-aye.

In yonders meadow there behold
A soldier lying dead and cold.

His horse is standing by his side
Waiting for him to get on and ride.

His hounds are lying at his feet
Licking the wounds that are so deep.

There came a maiden full of woe.
She turn-ed o’er his bloody head
And kissed the lips that once were red.

She laid herself down by his side
And there she lay until she died.

They dug a grave both long and side
And placed this couple side by side.

Out of his grave a red rose grew
And out of her a lily too.

S. Twa Corbies, Scotland, from c. 1783, fitted to An Alarc’h, Brittany, 1839, in c. 1956.

The only supposedly traditional melody surviving for the Mary Fraser Tytler manuscript words for Twa Corbies, made more well-known by Sir Walter Scott, is that printed by George Eyre-Todd in Ancient Scots Ballads, 1894. As noted for variant D above, Eyre-Todd’s vague attribution for the melody, and its closeness to another tune, raises the suspicion that this version was his invention rather than genuinely traditional. Connoisseurs of folk song may have spotted that Twa Corbies to a different tune, often sung by folk revival performers in the 20^th century, was missing from section D. This last variant is the version with that tune. 

As we have seen in the first article, commentators on traditional song are wont to make unwarranted, fanciful, and un-evidenced claims about the age of a song, and performers have regularly done the same on album sleeve notes, sometimes just picking a random century for the date of the song’s origin. So it is with Twa Corbies on Hark! The Village Wait, the debut album of folk-rock band, Steeleye Span, released in 1970. Ashley Hutchings, the band’s bass player, stated in the sleeve notes for its CD re-release in 1991, “This goes back to the 13^th century at least”, for which there is not the slightest jot of evidence, nor does he offer any: there is no record of the song before 1611. John Tobler, who wrote the re-release album notes, also gets the author of the printed source wrong: “First printed in [William] Motherwell’s Minstrelsy of the Scottish Border in 1803”. This was a book by Sir Walter Scott, not William Motherwell. William Motherwell’s book with Twa Corbies was Minstrelsy: Ancient and Modern, published in 1827. What neither the original sleeve notes nor the re-release notes for the album state is that the tune Steeleye Span used for Twa Corbies was borrowed from another song.

The Scots poet, Morris Blythman, who wrote under the pseudonym Thurso Berwick, was central to the Scottish folk song revival. In circa 1956 he fitted the words of Twa Corbies to a traditional Breton tune, An Alarc’h (The Swan), learned from the Breton folk-singer Zaig Montjarret, who must by some route have learned it from Barzaz Breiz (Ballads of Brittany), a book of traditional Breton songs collected by Théodore Hersart de la Villemarqué, published in 1839.

The Breton An Alarc’h and the tune’s adaptation for Twa Corbies can be heard in the videos below. 

There were three Ravens: Here are three Observations

The previous article, There were three Ravens: sublime love and ridiculous riddles, concluded with the following two observations.

i. There were three Ravens is a cleverly deceptive song. On the surface, it is a simple story, its appearance made simpler by the repetition of lines and the interjection of non-lexical vocables. Under the surface, it carries a long history of symbolic meaning that would have been instantly understood by its singers and listeners in the early 17^th century. Due to a culture shift in the intervening centuries, the symbolic language and story of the song require detailed analysis to be understood in the modern day. Doing so reveals a song about sublime love couched in Christian symbolism. I do not believe one needs to be a Christian to appreciate its depth and beauty.

ii. From the sublime to the ridiculous. The commentary of Bertrand Harris Bronson, and especially of Vernon V. Chatman III, illustrates the tendency among some 20^th century commentators on traditional song to invent backgrounds for ballads that are not based on evidence and fly in the face of credibility, imposing fanciful interpretations that have no basis in the song’s historical context. Not all commentators were of this tendency, and one of the chief English folk song collectors, Cecil Sharp, explicitly argues against such baseless speculation in his English folk song, some conclusions (1907).

To these observations, we can now add a third.

iii. In the oral tradition, the story in the song went through many transformations:

• from the faithfulness of the slain knight’s hounds, hawks and lover, who protected his body from predation by three scavenging ravens, and the pregnant woman who died from the strain of burying his body, based on the medieval parable of Christ the lover-knight;
• to the knight being a squire with a horse, and the woman laid down by his side and died without burying him;
• or the same story, but the three ravens were blackbirds and the squire was a soldier;
• to the woman cursing the killer of her true love;
• to the complete desertion of the slain knight by his hounds, hawks and lover, who left his body to predation by two ravens;
• to the three ravens being crows, the removal of all characters in the story except the horse, now dead, there being no narrative or meaning, and it is now a black-face minstrel parody of lining out in church, sung to Bonnie Doon, which entered the oral tradition and either kept the horse or changed it to an oxen;
• to the same degenerated non-story as the minstrel version, sung to When Johnny Comes Marching Home, with added imitative crow calls;
• and the same degenerated non-story to a variety of other melodies, with two or three crows;
• to a more compassionate version, with “a gay little birdie” warning a “lambkin” (rather than a horse or oxen) to get away from three crows, “Or else the three crawens would make him to dee”.

We can observe that:

• Over time, creativity in the oral tradition is wild and unpredictable, bringing about huge changes in melody and story.
• These changes are always contextual, made only according to the one version the singer heard, not to the longer tradition sometimes stretching back centuries.
• Thus over time a song can change its story, tone and meaning, so former melodies may change altogether or be swapped in from another song, and earlier poetic profundities may degenerate into a homeopathic reduction of its former self, just as the rich symbolism of There were three Ravens degraded over time into merely a song about two or three hungry crows looking for a meal.
• The degeneration of a song is only possible to recognise if one knows of its earlier, richer versions. Seen in their own right and without any other reference points, later variants such as Billy Magee Magar may be seen simply as fun participatory songs.

 

© Ian Pittaway. Not to be reproduced in any form without permission. All rights reserved.

 

Bibliography

Baldridge, Terry L. (undated) Irish Lilting.

Beatty, Arthur (1907) Some New Ballad Variants. The Journal of American Folklore, Vol. 20, No. 77 (April–June 1907), pp. 154–156.

Botkin, Benjamin Albert (1937) The American Play-Party Song. New York: Frederick Ungar.

Brewster, Paul G. (1940) Ballads and Songs of Indiana. Collected and Edited by Paul G. Brewster. Bloomington, Indiana: Indiana University Bookstore.

Bronson, Bertrand Harris (1959) The Traditional Tunes of the Child Ballads. Volume 1. Princeton, New Jersey: Princeton university Press.

Brower, Frank (1863) Frank Brower’s Black Diamond Songster. New York: Dick and Fitzgerald. Included in a composite book of 1864, The Universal Book of Songs and Singer’s Companion: containing an Extensive Collection of Patriotic, Sentimental, Irish, Ethiopian, Comic and Jolly Convivial Songs, embracing all the most popular vocal gems extant. New York: Dick and Fitzgerald.

Campbell, Olive Dame & Sharp, Cecil J. (1917) English Folk Songs from the Southern Appalachians. New York and London: G. P. Putnam’s Sons. Also see Sharp, Cecil J. & Karpeles, Maud (1932). 

Chambers, Robert (1829) The Scottish Ballads. Edinburgh: William Tait. 

Chappell, William (1859) Popular Music of the Olden Time. London: Chappell & Co.

Chase, Richard (1956) American Folk Tales and Songs. New York: New American Library.

Child, Francis James (1882) The English and Scottish Popular Ballads, Volume I (of V). New York: Dover Publications.

Child, Francis James (1890) The English and Scottish Popular Ballads, Volume IV (of V). New York: Dover Publications.

Child, Francis James (1898) The English and Scottish Popular Ballads, Volume V (of V). New York: Dover Publications.

Christy, E. Byron & Christy, William E. (1863) Christy’s New Songster and Black Joker. New York: Dick & Fitzgerald.

Cleveland, Charles Dexter (1862) A Compendium of English literature. Philadelphia: E.C. & J. Biddle.

Cobbold, William & Payne, Ian (c. 1620 & 2002) Cobbold New Fashions. Reconstructed by Ian Payne.

Cobbold, William – also see Payne, Ian (2002)

Coffin, Tristram Potter (1950) The British Traditional Ballad in North America. Philadelphia: American Folklore Society. 

Cox, John Harrington (1925) Folk-Songs of the South. Cambridge: Harvard University Press. 

Cripps, Wayne (transcriber) (1995) The Straloch Manuscript. Fort Worth, Texas: Lyre Music Publications.

Cryer, A. B. (2026) Christy’s Minstrels Explained.

Cunningham, Allan (1825) The Songs of Scotland, Ancient and Modern. Vol. I. & II. London: John Taylor.

Dictionaries of the Scots Language (undated) CORBIE, Corby, Korbi, n. [′kɔrbɪ̢, ′korbi].

Duffin, Ross W. (2014) The Music Treatises of Thomas Ravenscroft. Treatise of Practicall Musicke and A Briefe Discourse. London: Routledge.

Flanders, Helen Hartness (1960) Ancient Ballads Traditionally Sung in New England. Vol. I Child Ballads 1–51. (Philadelphia: University of Pennsylvania Press).

Gainer, Patrick W. (1975) Folk songs from the West Virginia Hills. Grantsville, West Virginia : Seneca Books.

Hauser, William (1848) The Hesperian Harp. Philadelphia: S. C. Collins.

Henry, Mellinger E. (1932) Still More Ballads and Folk-Songs from the Southern Highlands. The Journal of American Folklore, Vol. 45, No. 175 (Jan–Mar 1932), pp. 1–176.

Horton, Randall Keith (1985) American Music: A Discussion on “Lining Out”.

Kidson, Frank (1891) Traditional Tunes. Oxford: Chas. Taphouse & Son.

Larson, Kenneth (c. 1933) Songs and Ballads (Folk Material and Old Favorites) Collected By kenneth Larson, High School Teacher, McCammon, Idaho.

Mabillard, Amanda (2011) Playing Fast and Loose with Shakespeare’s Name.

Manson, James (1846) Hamilton‘s Universal Tune Book, Vol. II. Glasgow: W. Hamilton.

Matteson Jr., Richard L. (2014) 26. The Three Ravens (or Twa Corbies).

Morehen, John & Mateer, David (transcribers and editors) (2012) Thomas Ravenscroft. Rounds, Canons and Songs from Printed Sources. London: Stainer and Bell.

Motherwell, William (1827) Minstrelsy: Ancient and Modern. Glasgow: John Wylie.

Payne, Ian (2002) New Light on ‘New Fashions’ by William Cobbold (1560–1639) of Norwich. The Journal of the Viola da Gamba Society, Volume 30,  pp. 11–37.

Ravenscroft, Thomas (1609) Pammelia. Musicks Miscellanie. Or, Mixed Varietie of Pleasant Roundelayes, and delightfull Catches, of 3.4.5.6.7.8.9 10. Parts in one. London: William Barley.

Ravenscroft, Thomas (1609) Deuteromelia: Or The Second part of Musicks melodie, or melodius Musicke. Of Pleasant Roundelaies; K. H. mirth, or Freemens Songs. And such delighfull Catches. London: Thomas Adams. Click the title or here.

Ravenscroft, Thomas (1611) Melismata, Musicall Phansies. Fitting the Court, Citie, and Countrey Humours. To 3, 4, and 5. Voyces. London: William Stansby.

Ravenscroft, Thomas (1614) A Briefe Discourse Of the true (but neglected) use of Charact’ring the Degrees by their Perfection, Imperfection, and Diminution in measurable Musicke, against the Common Practise and Custome of these Times. London: Thomas Adams.

Ritson, Joseph (1790) Ancient Songs and Ballads From The Reign of King Henry the Third, To The Revolution. London: J. Johnson. Click the title to read this online with the paper in its current aged colour, or here to read or download a cleaned-up PDF.

Scott, Sir Walter (1803) Minstrelsy of the Scottish Border. Vol. III. Edinburgh: Longman and Rees.

Sharp, Cecil J. (1907) English folk song, some conclusions. London: Simpkin & Co. 

Sharp, Cecil J. & Karpeles, Maud (1932) English Folk Songs from the Southern Appalachians. London: Oxford University Press.

Shoemaker, Henry W. (1931) Mountain Minstrelsy of Pennsylvania. Philadelphia: Newman F. McGirr.

Simpson, Elske (2024) What is lilting?

Waite, Henry Randall (1868, this 2^nd edition 1876) Carmina Collegensia: a complete collection of the songs of the American colleges, with selections from the student songs of the English and German universities, and popular songs adapted to college singing. Boston: Oliver Ditson & Co.