Other Prominent Names: "Oh, the Wind and Rain","Binnorie","The Cruel Sister","The Twa Sisters",The Miller and the King's Daughter',"The Bonny Swans","The Bonny Bows of London"
Regional Traditions: Scottish,English,Appalachian
Style: Ballad
Sub-style: Murder Ballad
Oldest Surviving Version: 1658
Child Number: 10
Roud Number: 8
Key: E
Justin's Commentary
Although it's hard to assess the age of traditionals with certainty, The Two Sisters probably has the furthest-reaching clear lineage of any song we play, with its oldest surviving written version dated 1658.
I've seen several scholars opine that the variation of lyrics throughout the history of this song is high (for exmaple, Robert B. Waltz and David G. Engle at Ballad Index assert that "The refrains sung with this ballad vary tremendously").
...and maybe this is my bias as a bluegrasser, but in my reading, I've been surprised as how consistent the lyrics seem to be. The variations in almost four centuries of this song are not much more than we find in just a single century of evolution of bluegrass standards like Rueben's Train or Shady Grove.
Consider that the 1658 version begins "There were two sisters who went a' playin'", while the most common 2024 version begins "There were two sisters walking down a street". That's pretty serious staying power.
Anyway, throughout the history of this song, we find three common elements, with most versions having at least two, and many all three:
Two Sisters, the older of whom kills the younger by drowning her, usually in a navigable stream, but sometimes a sea.
A miller, a family of Millers, or a body of water known to be on land owned by millers. Often, the millers' son is a suitor seeking the affection of one of the sisters, usually the younger one.
An instrument, usually a fiddle/violin/viol but sometimes a harp, which is either made from body parts of the murdered sister, or which can only play a song telling of her fate, or both. In some versions, this instrument is played at the older sister's wedding (sometimes, to the miller's son), and during the performance, it snitches on her.
The third of these is remarkably steady, appearing in essentially all versions for the first two centuries of this song, and the vast majority thereafter as well. Francis James Child (whose scholarship we'll examine in a moment) notes that the fabrication of an instrument from body parts appears in "all complete and uncorrupted forms of the ballad" published prior to 1882. In fact, if a version does not have this narrative component, I view it as a good prompt to examine whether that version might have been derived from a different song entirely. The history of this song is fascinating in this respect; let's examine it in detail.
History
Earliest Scholarship, uncovering the 1658 text
In 1852, in Volume 5 of the Oxford University Press journal "Notes and Queries", we find research into a song called "The Miller's Melody", which by that time was already, and is expressly referred to as, "an old ballad". The entry includes an account from British hymnologist and organ player Edward Francis Rimbault, who says:
The excerpt from an 1852 volume of "Notes and Queries", describing the broadside with the verses to the song. (scan from archive.org)
I have an old "broadside" copy of the ballad in question, "Printed for Francis Grove, 1656," which is here transcribed, verbatim et literatim, for the especial benefit of your numerous readers. It may also be found in a rare poetical volume, entitled Wit Restored, 1658, and in Dryden's Miscellany Poems (second edition, which differs materially from the first).
As far as I can tell, the 1656 broadside has never been recovered, and no known version of it survives.
However, the verses included in Rimbault's letter precisely match those in the 1658 book "Wit Restor’d" to which he refers, and this book has survived, with a remarkably legible scan on archive.org.
The title of the book is given in five lines:
The oldest known surviving written version of the song, from the 1658 book "Wit Restor'd". (scan from archive.org)
Wit Restor’d. In severall Select Poems Not formerly publish’t.
The book begins with the following inscription:
Mr. Smith, 'to Captain Mennis then commanding a Troop of Horſe' in the North, against the Scots.
The "Mr. Smith" here is apparently a reference to the one of the book's authors/collectors, James Smith, a clergyman who published four books - apparently mostly of satirical verse - from 1640 to 1663. He became Archdeacon of Barnstaple in 1660, and died in 1667.
The song appears on page 51, under the heading "The Miller and the King's Daughter". It's attributed to "Mr. Smith", though, as with other publications of the time, that appears to be giving credit to Smith as the archivist, not the author.
Confusion over Smith's role has led to some conjecture either that the song might be entirely parody, or that Smith might have added verses intended as burlesque interpretation of the rest of it.
An appendix of the Project Gutenberg Book of English and Scottish Ballads, Volume II, has this to say about the matter of whether the entry is an indication of parody (note the song is assssed in long-form in the text proper on page 231):
From Wit Restor'd, (1658,) reprinted, London, 1817, i. 153. It is there ascribed to "Mr. Smith," (Dr. James Smith, the author of many of the pieces in that collection,) who may have written it down from tradition, and perhaps added a verse or two. Mr. Rimbault has printed the same piece from a broadside dated 1656, in Notes and Queries, v. 591. A fragment of it is given from recitation at p. 316 of that volume, and a copy quite different from any before published, at p. 102 of vol. vi. Although two or three stanzas are ludicrous, and were probably intended for burlesque, this ballad is by no means to be regarded as a parody.
This seems to me a fair assessment. Rimbault's phrasing suggests that he too regarded Wit Restor'd as an archival copy of an older song, not as an original work. And thumbing through Wit Restor'd on archive.org, which I encourage you to do if this kind of content is interesting to you, one finds that very few of the other 'poems' are written in verse form as if to convey a traditional song in rhythm. Indeed, it seems that inclusion of this piece was meant as a separate exercise from most of the rest of the book.
1700s, many titles
In the mid-1700s, we start to see the song appear with many names: "Damned mil-damm" in 1749, "The Cruel Sister" in 1760, and in 1770, "Bonny Bows of London" and "The King Lived in the North Country" (more on this, the "Bow Down" variant, below).
An illustration from the 1890 fairy tale version of "Binnorie"
Then in 1781, we see the song appear for the first time as "Binnorie", a name with incredible staying power, as versions with that name are still being played and published today. "Binnorie" is also the name given to a fairy tale adaptation in 1890.
In the later 1700s and 1800s, the song became a popular folk ballad throughout Scandinavia, and here, we see the younger sister's "long yellow hair" become a much more frequent theme. During our 2024 Europe tour, I felt a certain connection to the history of the piece while playing it in the streets of Copenhagen. We included a couple of verses that we don't normally sing, including the one about stringing the bow with her long yellow hair.
Child lists Twenty-one variations and then provides about nine pages of exposition on these 21, comparing and contrasting lyrical content - quite a lift for pre-internet research. He also notes several sub-variations among these, including what appear to be the beginnings of some of the lyrical distinctions, such as which instrument ia made, whether the sisters are rich or poor, etc., which I'll discuss below.
Interestingly, none of these 21 variations appear to have been commonly called by any variation of "Wind and Rain" at that time - this name appears to have gained traction some time after Child's writing.
Most of the scholarly historical accounts of this song conclude with variants of the late 1800s, or focus only on the English and Scottish traditions of the early 1900s, but the part of the lineage which we celebrate most is yet ahead.
Appalachian Evolution
Around 1900, we see start to see a version in England and Appalachia that has come to be called the "Bow Down variant", which has several elements which are starkly different from either the early history of the song or our more recent "Wind and Rain" versions:
There is no instrument made or augmented from the body parts of the younger sister
The miller is either the murderer, or is wrongly accused of being the murderer
The miller is hanged at the end of the song
A good example of this variant is this 1928 recording by a man named Bradley Kincaid from Garrard County, KY.
Although scholars seem to unflinchingly accept "Bow Down" as the same song, it has deep enough structural and thematic differences that I might be inclined believe it is an entirely different song which was at some point partially lost, and the missing pieces were filled in with stanzas from The Two Sisters. For example, I can't help but notice that the final line of each stanza bears a rhythmic resemblance to "Crawdad Song":
"He bought the youngest a fine fur hat; and the older sister, she didn't like that"
"You get a line, I'll get a pole and we'll all go down to the crawdad hole"
In fact, if you listen to the above version, it's fairly easy to sing "Crawdad Song" in your head as you hear each verse. Hints of other songs strike the ear as well.
However... there is apparently a single data point from over a cenutry earlier - in 1770 - that, if authentic, makes a clear case that "Bow Down" is indeed rooted in The Two Sisters. It's a song called "There was a King Lived in the North Country", and unlike many of the early 1900s versions, it has most of the common historical elements: two sisters, a gay gold ring (albeit given by their father), a drowning younger sister... it even has the line "Oh father, oh father, there swims a swan" which we'll see later picked up by Grateful Dead and Crooked Still and others.
So whether Bow Down is The Two Sisters, or whether it has another lost root, it's clear that the two songs are closely related.
In any case, shortly thereafter, we start to see elements emerge that represent the quintessence of the Appalachian (and later, Grateful Dead) lineage of this song. For example, a 1933version cut by Richard Hayward as "Binnorie" and distributed by Decca Records has the iconic line "a sound that melted hearts of stone", sung with a cadence and rhythm that almost exactly matches Jerry Garcia's delivery of this line 50 years later. My best guess is that Jerry Garcia had access to this recording, or to a version of the song which was influenced by it.
After this, studio versions become more sparse as the song takes on more Appalachian qualities, but also becomes more obscure. Almost all recordings of the next few decades are made by archivists or folklorists, and not by touring musicians.
Of personal interest to me as a newly minted Florida musician, there is a noted version recorded by an archivist visiting a home in High Springs, Florida, in June 1937. It is sung by a Mrs. C. S. MacClellan (Library of Congress entry: https://www.loc.gov/item/afc9999005.5234). This version has a simlar structure to the "Binnorie" versions above (though it says "By Nowling"), but it also tells the story much more in keeping with both the oldest versions and the "Wind and Rain" versions ahead.
In the second half of the 20th century, surviving versions continue the descent into an underground folk vibration. In fact, it's not clear to me that this song was ever played to a large crowd from about 1860 to 1980. But a few prominent and lovely versions exist from this period of the song's fade to temporary obscurity.
The first is the beautiful 1958 Peggy Seeger version - the last published pre-Dead version that sounds like a distinctly different branch - it is mostly a Bow Down rather than a Wind and Rain. In fact, this might be the last version in this lineage which does not contain the phrase "wind and rain". And although this version sounds quite different (and selects very different verses) from Garcia's, the banjo part is somewhat reminiscent of Jerry's banjo style as we hear it in slower songs; I wonder if he might have heard this version and focused more on the banjo than the structure and verses.
After this point, as the Appalachian approaches the psychedelic, we see a rapid and near-universal resurrection of many of the very old verses, and particularly those which refer to the making of the instrument from the girl's body, after nearly half a century in which most versions omit them.
In the 1960's, Smithsonian Folkways Recordings has in its collection two versions: in 1961, Chicago musicians George and Gerry Armstrong cut a version, and a twangy, floaty 1969 version by an autoharpist named Kilby Snow. These versions appear to be the first that have the clear versewise refrain, "Oh the wind and rain" / "Oh, the dreadful wind and rain".
Two versions in the 1970s are, to my ear, most dispositive on the evolution of the song into the Grateful Dead version which has become the variant of note for many people listening in this era:
In 1975we have the Red Clay Ramblers version which, I'll note as an aside for later, describes the younger sister's "long curly hair" (rather than fine or yellow hair).
Jody Stecher's version cut in 1977 is the very obvious precusror to the Jerry Garcia version, with nearly identical verses and melody. In fact, Stecher probably deserves credit as the arranger of note for what is today the canonical version of the song, and his version is marvelous. I can't help but notice that my good friends East Nash Grass play the song almost identically to this record.
However, as best I can tell, this was for Stecher largely a labor of love and not part of a touring effort or public exposition of traditional music. I don't know how often he played the song live in this era, but I suspect it never had the light shone on it that it was about to.
The Dead
(Note: here, and in most commentary, when I refer to The Grateful Dead, I mean all of its outgrowths and not strictly the band of that name.)
As with many traditionals which find themselves refreshed at campfires and festivals, the stewardship of The Grateful Dead has been defining for this song.
Jerry Garcia and David Grisman made The Two Sisters (usually as "Oh, the Wind and Rain") a staple of their live shows in the early 1990s. In fact - don't tell anyone - they played it three nights in a row at The Warfield Theater in San Francisco in 1991 - December 7, December 8, and December 9.
If you've read this far, and if you love this song, I encourage you to take a moment to set aside whatever else you're doing and listen to this version by Jerry Garcia and David Grisman from 1992 at Warfield Theater - and consider that this song traveled 334 years as documented above to land in this room, with these instruments, and this packed-in crowd. There's decidedly nothing original here; this is a pure traditional. Taste the acid on the tongue of the fiddle, as Joe Craven enters the scene on cue. This is what rebirth looks like:
Understand that this room is something of a mecca for deadheads - in 1980, The Grateful Dead played a run of 15 nights there, celebrating their 15th anniversary. For The Two Sisters to land there after three and a half centuries is an exercise in genuine human time and space travel.
The DeadDisc entry for The Two Sisters notes a stunning 151 versions of the song leading up to, and then influenced by, the moment captured here.
(As an aside: DeadDisc is an incredible repository of information the appalachian traditionals played by The Grateful Dead, Jerry Garcia, and Old and In The Way, and a quintessential example of the bridge between traditional music and cypherpunk culture. It's a great next step if you're doing research on the appalachian tradition of this song.)
Given that The Two Sisters is well-documented for centuries, and appears in several prominent research repositories and collections, it was never at risk of being forgotten.
However, Jerry Garcia's discovery and development of the song was certainly instrumental in recovering it from relative obscurity.
Around the same time as the Warfield performance above, Garcia and Grisman recorded what I surmise is the most widely heard version among our fanbase, for their Shady Grove record. The record was released in 1996, posthumously in the case of Garica, who died in 1995.
Some things I want to note about the Shady Grove version:
Garcia sings here with very little drama or pathos; it has the sound of a straight Appalachian traditional (which on one hand I adore and aspire to be able to do, but on the other is different from some of my other favorite versions, which I'll describe below).
At the end, you can hear him saying, presumably to his fellow session players or the audio engineer (David Dennison), "that's all there is to that tune."
Bryan Bowers plays an autoharp here that is clearly included as an homage to the 1969 Kilby Snow version, but which has its own lovely soul and does not borrow heavily from Snow's solos.
Nearly every version after this point has one of two qualities:
It is a direct homage to the Garcia/Grisman version, with the same verses and melody, or
It is an intentional departure from the Garcia/Grisman version, with verses and sound from other lineages (especially "Binnorie"), and fit to play in Old Time circles.
There is one version that has both: a 2009 cut by Julie Fowlis, a Scottish Gaelic rivival icon and multi-instrumentalist powerhouse. Fowlis' version has the structure and verse of the Garcia version, but with every other verse sung in Scottish Gaelic, with new-agey harmonies.
Another compelling version with a harmonic focus is the one by Gillian Welch and David Rawlings for the Soundcatcher sountrack. This version includes a dose of Welch and Rawlings' commitment to haunting dissonance, and also includes some lyrics which are quite old, and others which appear, as far as I can tell, for the first time. For example, in this version, the sisters are from County Clare, Ireland. And while I can find references in reviews to the sisters being from County Clare, unless I'm missing something, this 2001 version is the first time this detail appears in the song.
In 2005, the Irish folk band Altan used the Welch/Rawlings lyrics in their version.
The version that squarely put this song on my radar, and which remains one of my favorites, is the 2015Billy Strings version with Greensky Bluegrass. Strings' version is a direct homage to the Garcia/Grisman version, but with a longing and lonesome sound. As might be expected from Billy Strings, the guitar solos are kept very close to the Garcia version, but with a thick and acute bluegrass picking style.
Another "sad live" version - perhaps the best in class for that category - is the 2019 version by Crooked Still with Lula Wiles. Aoife O'Donovan's voice is a certainly a match for the song, and the band's unusual arrangement (with a Cello-driven intro) is a perfect blend of the old and the new. Lyrically, this version stays close to the Garcia version, including the line "Father, oh Father there swims a swan", borrowed all the way from the 1770 "King Lived in the North Country" version.
Another notable Billy Strings performance of this song for me personally is from November 12, 2022 at The Met in Philadelphia. I, my son Fibonacci, and my parents were in the audience (it was my parents' first Billy Strings show). Toward the end of the first set, Billy Strings said, "Why don't we do... mandolin player's choice!"
At this point, mandolin player Jarrod Walker (who happens to be the younger brother of our frequent studio collaborator and one-of-a-kind banjo phenom Cory Walker) said something into the talkback mic, to which Billy looked surprised and replied, "that's not what I expected of you! I was expecting like Ashes of Love or something like that, but I think this is what Jarrod Walker wants to hear."
And sure enough, Jarrod's choice had been The Two Sisters (which they typically write on setlists as "Oh The Wind And Rain"). Jarrod knows that I like and play this song, and also knew I was going to be in the audience that night, but whether my presence had any effect I don't know; almost certainly coincidence I reckon. Either way, I was thankful because it was a lovely version, and it was a great moment for me and my family.
As an aside: after that show, the band's bassist Royal Masat came over to my AirBnb, where we shared a bottle of wine and talked about music late into the night. He was very encouraging about my music and my career, and it was a very validating moment, just when I needed a reason to keep going. Thanks Royal.
As cryptograss
Although I don't closely keep track of this statistic, I suspect that The Two Sisters is the traditional song which we are most often requested to play at various cypherpunk events. I have noted and described the incredible thirst for traditional music among the cypherpunk community at many shows and podcast appearances, and the age and many threads of this one seems to be captivating to the crypto native mind.
We played this song at six show with our dear friend and kickass fiddler Kuba Hejhal, who left this world in September 2024. Kuba was a special cryptograsser; he took very seriously the task of doing justice to this song.
As a band, we've played it at 12 shows so far, and it's always interesting to see how y'all respond. For whatever reason(s), it seems to be counter-intuitive to many people to suppose that these old songs might hit so hard, and be such a quintessential piece of the cypherpunk puzzle. But this song has survived and surged for three and a half centuries with no need for copyrights or centralized authorities to protect it. It serves as an easy and obvious example of what we hope decentralized society to be.
Lyrical variations, and choosing which verses to sing
What instrument? Made from what body part(s)?
I have always sung "he made a little fiddle", from her breastbone. This seems to me to be the oldest, most common, and most compelling story to tell to an audience, especially when we're playing with a great fiddler who can bring the song home.
The Scots Language Centre claims this song as "One of Scotland's oldest ballads", and notes that the song ends with a "wonderful magical ending, in which a harp or fiddle (or even a piano) is made from the sisters white bones and yellow hair, and taken to play at the wedding, where it sings the story of the murder."
However, I can find no version where a piano is made. I also find it slightly unlikely that any version which is contemporaneous with the establishment of the three themes I've outlined above might explicitly mention a piano, since the word "piano" doesn't appear to have been coined until after 1800 (and its predesseor, "pianoforte", from the 1760s), well after these themes had been established in dozens of versions or more.
The sound and song of the fiddle / harp
One of the most frissionate moments of the versions we've come to love, from Jerry Garcia to Crooked Still to Billy Strings, is the line, "with a sound that melted hearts of stone."
This line is old, going back to at least an 1802 version (see Child C in the Child balads link above), and very likely even further - nearly all "Binnorie" variants have a line nearly exactly matching this one:
He's made a harp of her breistbane, Binorie, O Binorie, Whase soons wad melt a hairt o stane, By the bonnie milldams o Binorie.
In most variants, the only song that the fiddle can play is the story of the younger sister's death, and in many, it is a self-reference to the very song being played. So for example, in the "Wind and Rain" versions, we almost always have a line resembling:
"The only tune that the fiddle could play was 'Oh, the dreadful wind and rain'"
With that in mind, I think it's interesting that the 1658 version has the following lines:
"What did he doe with her two shinnes?"
"Unto the violl they danc'd Mall Syms;"
"Mall syms" here is a reference to a popular dance of the time, an extant notation of which exists from 1626 - is it possible that The Two Sisters was once set to the music of this dance, long before the events described in this document? This is one of the possible avenues by which this song might be discovered to be even older than 1658 - if some enterprising reader wants to dig into 'Mall Syms', I'm happy to hear what y'all find. :-)
Are the sisters rich? Or poor?
In some versions, the sisters are unambiguously heirs to a fortune. In the 1658 version, for example, they appear to be daughters of the king, and the pretense that the elder sister gives for bringing the younger one to the water is to watch their father's fleet of ships come in.
Child writes (using his alphabetical notation here for different variations):
The sisters are king's daughters in English A, B, C, H, O (?), P, Q, R a, and in Swed ish B and two others of Afzelius's versions. They are an earl's daughters in Swedish F, and sink to farmer's daughters in English R b, c,* Swedish A, G, Norwegian C.
A 1931 version has the sisters going to the water to sse their father's fishing boat, which seems to support the idea that the story varied with regard to their wealth.
In many "Bow Down" versions, including the 1928 version linked above, the sisters are decidedly not royalty, and instead we have something like:
"An old woman lived by the sea shore, and she had daughters three or four"
(which seems strange on its own, and possibly another indication of an independent origin of "Bow Down" - if it's "The Two Sisters", why do these variants describe three of four daughters?)
In many versions, but almost every "poor sisters" version, we see the suitor court one or both of them with a "gay gold ring" or "gay gold pin", which makes sense - a young poor girl might presumably marvel at a gold ring.
However, we also see the ring in older versions where it's not clear whether the sisters are royalty or not, which suggests to me that there may be lost "poor sisters" versions predating the 1900s "Wind and Rain" versions.
The barefoot younger sister
I have often included in our versions a reference to the younger sister being barefoot.
An 1828 book, "Ballads of the North of Scotland" II, 128, lists verses for a version written as "The Bonny Bows O' London". This version mentions the younger sister not wearing shoes:
They could not ken her foot sae fair, Hey wi' the gay and the grinding; The shoes o' gowd they were so rare, At the bonny, bonny bows o' London.
A 1931 version, written as "Bilnorie", also seems to suggest that she was barefoot, referring to her "lily feet" in verse 14.
A very different 1932 version, as "The Twa Sisters", also mentions a barefoot younger sister:"
"Pit your fit (feet) on yonder marble stone," Heigh, ho my nannie O! An' sae slyly she dung her in, An' the swan swims sae bonnie, O.
An 1860 version written as 'Benorie', documented in the Scottish Highlands, also mentions the younger sister being barefoot. amd the miller pulling her out of the water by her bare feet.
Some other versions mention her having silver slippers (in some caes, which fall off during her float toward the miller's dam/pond) or leather shoes.
I sometimes sing as follows:
"There were two sisters of County Luce" (keeping the reference from the Welch/Rawlings version of their country of origin, but changing it to a county in Michigan, near my own roots, and Billy Strings' as well)
"They were only able to afford the older one's shoes" (which nods to the many "poor sisters" and "barefoot younger sister" variants)
Her hair: what is it used for? Is it fine or dready?
In most versions, the younger sister's hair is used to make either the strings of the instrument or to sring the the bow. In some versions, it is used to make the fiddle itself.
The most common quality of her hair we find is its color: yellow, gold, or "gowd", though, as I mentioned above, this quality only gains quasi-ubiquity once the songs becomes ppopular in Scandinavia.
Several modern versions, including both the Jerry Garcia versions and the Gillian Welch version, say, "he strung his fiddle bow with her long yella' hair."
As I mentioned above, some versions expressly describe her hair as "curly", while others say it is "fine".
And then, several of the oldest versions have the fiddler/harpist taking a "tate" or "tait" or her hair. What is a tate? It appears to me to be an old Scottish word for a dreadlock. (https://www.dsl.ac.uk/entry/snd/tait)
And since it doesn't make much sense to make a fiddle string out of a dreadlock - and because I love dreadlocks - I generally sing "He made a fiddle case of her long nappy/dready hair". However, I must admit that I can find no historical version mentioning that the fiddler makes a case for the fiddle.
So yeah, barefoot and dreadlocked. What can I say? I like stories about hippies.
Although it's hard to assess the age of traditionals with certainty, The Two Sisters probably has the furthest-reaching clear lineage of any song we play, with its oldest surviving written version dated 1658.
I've seen several scholars opine that the variation of lyrics throughout the history of this song is high (for exmaple, Robert B. Waltz and David G. Engle at Ballad Index assert that "The refrains sung with this ballad vary tremendously").
...and maybe this is my bias as a bluegrasser, but in my reading, I've been surprised as how consistent the lyrics seem to be. The variations in almost four centuries of this song are not much more than we find in just a single century of evolution of bluegrass standards like Rueben's Train or Shady Grove.
Consider that the 1658 version begins "There were two sisters who went a' playin'", while the most common 2024 version begins "There were two sisters walking down a street". That's pretty serious staying power.
Anyway, throughout the history of this song, we find three common elements, with most versions having at least two, and many all three:
The third of these is remarkably steady, appearing in essentially all versions for the first two centuries of this song, and the vast majority thereafter as well. Francis James Child (whose scholarship we'll examine in a moment) notes that the fabrication of an instrument from body parts appears in "all complete and uncorrupted forms of the ballad" published prior to 1882. In fact, if a version does not have this narrative component, I view it as a good prompt to examine whether that version might have been derived from a different song entirely. The history of this song is fascinating in this respect; let's examine it in detail.
History
Earliest Scholarship, uncovering the 1658 text
In 1852, in Volume 5 of the Oxford University Press journal "Notes and Queries", we find research into a song called "The Miller's Melody", which by that time was already, and is expressly referred to as, "an old ballad". The entry includes an account from British hymnologist and organ player Edward Francis Rimbault, who says:
As far as I can tell, the 1656 broadside has never been recovered, and no known version of it survives.
However, the verses included in Rimbault's letter precisely match those in the 1658 book "Wit Restor’d" to which he refers, and this book has survived, with a remarkably legible scan on archive.org.
The title of the book is given in five lines:
The book begins with the following inscription:
The "Mr. Smith" here is apparently a reference to the one of the book's authors/collectors, James Smith, a clergyman who published four books - apparently mostly of satirical verse - from 1640 to 1663. He became Archdeacon of Barnstaple in 1660, and died in 1667.
The song appears on page 51, under the heading "The Miller and the King's Daughter". It's attributed to "Mr. Smith", though, as with other publications of the time, that appears to be giving credit to Smith as the archivist, not the author.
Confusion over Smith's role has led to some conjecture either that the song might be entirely parody, or that Smith might have added verses intended as burlesque interpretation of the rest of it.
An appendix of the Project Gutenberg Book of English and Scottish Ballads, Volume II, has this to say about the matter of whether the entry is an indication of parody (note the song is assssed in long-form in the text proper on page 231):
This seems to me a fair assessment. Rimbault's phrasing suggests that he too regarded Wit Restor'd as an archival copy of an older song, not as an original work. And thumbing through Wit Restor'd on archive.org, which I encourage you to do if this kind of content is interesting to you, one finds that very few of the other 'poems' are written in verse form as if to convey a traditional song in rhythm. Indeed, it seems that inclusion of this piece was meant as a separate exercise from most of the rest of the book.
1700s, many titles
In the mid-1700s, we start to see the song appear with many names: "Damned mil-damm" in 1749, "The Cruel Sister" in 1760, and in 1770, "Bonny Bows of London" and "The King Lived in the North Country" (more on this, the "Bow Down" variant, below).
Then in 1781, we see the song appear for the first time as "Binnorie", a name with incredible staying power, as versions with that name are still being played and published today. "Binnorie" is also the name given to a fairy tale adaptation in 1890.
In the later 1700s and 1800s, the song became a popular folk ballad throughout Scandinavia, and here, we see the younger sister's "long yellow hair" become a much more frequent theme. During our 2024 Europe tour, I felt a certain connection to the history of the piece while playing it in the streets of Copenhagen. We included a couple of verses that we don't normally sing, including the one about stringing the bow with her long yellow hair.
As 'Child 10'
In 1882, Francis James Child included the song in his collection, titled "The English and Scottish Popular Ballads", as "10: The Twa Sisters". This was, and arguably remains, the most comprehensive accounting of the song's early history (the other contender is the amazing and much more recent scholarship of Richard Matteson at Bluegrass Messengers).
Child lists Twenty-one variations and then provides about nine pages of exposition on these 21, comparing and contrasting lyrical content - quite a lift for pre-internet research. He also notes several sub-variations among these, including what appear to be the beginnings of some of the lyrical distinctions, such as which instrument ia made, whether the sisters are rich or poor, etc., which I'll discuss below.
Interestingly, none of these 21 variations appear to have been commonly called by any variation of "Wind and Rain" at that time - this name appears to have gained traction some time after Child's writing.
Most of the scholarly historical accounts of this song conclude with variants of the late 1800s, or focus only on the English and Scottish traditions of the early 1900s, but the part of the lineage which we celebrate most is yet ahead.
Appalachian Evolution
Around 1900, we see start to see a version in England and Appalachia that has come to be called the "Bow Down variant", which has several elements which are starkly different from either the early history of the song or our more recent "Wind and Rain" versions:
A good example of this variant is this 1928 recording by a man named Bradley Kincaid from Garrard County, KY.
Although scholars seem to unflinchingly accept "Bow Down" as the same song, it has deep enough structural and thematic differences that I might be inclined believe it is an entirely different song which was at some point partially lost, and the missing pieces were filled in with stanzas from The Two Sisters. For example, I can't help but notice that the final line of each stanza bears a rhythmic resemblance to "Crawdad Song":
In fact, if you listen to the above version, it's fairly easy to sing "Crawdad Song" in your head as you hear each verse. Hints of other songs strike the ear as well.
However... there is apparently a single data point from over a cenutry earlier - in 1770 - that, if authentic, makes a clear case that "Bow Down" is indeed rooted in The Two Sisters. It's a song called "There was a King Lived in the North Country", and unlike many of the early 1900s versions, it has most of the common historical elements: two sisters, a gay gold ring (albeit given by their father), a drowning younger sister... it even has the line "Oh father, oh father, there swims a swan" which we'll see later picked up by Grateful Dead and Crooked Still and others.
So whether Bow Down is The Two Sisters, or whether it has another lost root, it's clear that the two songs are closely related.
In any case, shortly thereafter, we start to see elements emerge that represent the quintessence of the Appalachian (and later, Grateful Dead) lineage of this song. For example, a 1933 version cut by Richard Hayward as "Binnorie" and distributed by Decca Records has the iconic line "a sound that melted hearts of stone", sung with a cadence and rhythm that almost exactly matches Jerry Garcia's delivery of this line 50 years later. My best guess is that Jerry Garcia had access to this recording, or to a version of the song which was influenced by it.
After this, studio versions become more sparse as the song takes on more Appalachian qualities, but also becomes more obscure. Almost all recordings of the next few decades are made by archivists or folklorists, and not by touring musicians.
Of personal interest to me as a newly minted Florida musician, there is a noted version recorded by an archivist visiting a home in High Springs, Florida, in June 1937. It is sung by a Mrs. C. S. MacClellan (Library of Congress entry: https://www.loc.gov/item/afc9999005.5234). This version has a simlar structure to the "Binnorie" versions above (though it says "By Nowling"), but it also tells the story much more in keeping with both the oldest versions and the "Wind and Rain" versions ahead.
A decade later, the Samuel Preston Bayard Papers at Penn State University inclues another "By Nowling" version recorded on August 12, 1948 by Charles S. Brink.
In the second half of the 20th century, surviving versions continue the descent into an underground folk vibration. In fact, it's not clear to me that this song was ever played to a large crowd from about 1860 to 1980. But a few prominent and lovely versions exist from this period of the song's fade to temporary obscurity.
The first is the beautiful 1958 Peggy Seeger version - the last published pre-Dead version that sounds like a distinctly different branch - it is mostly a Bow Down rather than a Wind and Rain. In fact, this might be the last version in this lineage which does not contain the phrase "wind and rain". And although this version sounds quite different (and selects very different verses) from Garcia's, the banjo part is somewhat reminiscent of Jerry's banjo style as we hear it in slower songs; I wonder if he might have heard this version and focused more on the banjo than the structure and verses.
After this point, as the Appalachian approaches the psychedelic, we see a rapid and near-universal resurrection of many of the very old verses, and particularly those which refer to the making of the instrument from the girl's body, after nearly half a century in which most versions omit them.
In the 1960's, Smithsonian Folkways Recordings has in its collection two versions: in 1961, Chicago musicians George and Gerry Armstrong cut a version, and a twangy, floaty 1969 version by an autoharpist named Kilby Snow. These versions appear to be the first that have the clear versewise refrain, "Oh the wind and rain" / "Oh, the dreadful wind and rain".
Two versions in the 1970s are, to my ear, most dispositive on the evolution of the song into the Grateful Dead version which has become the variant of note for many people listening in this era:
In 1975 we have the Red Clay Ramblers version which, I'll note as an aside for later, describes the younger sister's "long curly hair" (rather than fine or yellow hair).
Jody Stecher's version cut in 1977 is the very obvious precusror to the Jerry Garcia version, with nearly identical verses and melody. In fact, Stecher probably deserves credit as the arranger of note for what is today the canonical version of the song, and his version is marvelous. I can't help but notice that my good friends East Nash Grass play the song almost identically to this record.
However, as best I can tell, this was for Stecher largely a labor of love and not part of a touring effort or public exposition of traditional music. I don't know how often he played the song live in this era, but I suspect it never had the light shone on it that it was about to.
The Dead
(Note: here, and in most commentary, when I refer to The Grateful Dead, I mean all of its outgrowths and not strictly the band of that name.)
As with many traditionals which find themselves refreshed at campfires and festivals, the stewardship of The Grateful Dead has been defining for this song.
Jerry Garcia and David Grisman made The Two Sisters (usually as "Oh, the Wind and Rain") a staple of their live shows in the early 1990s. In fact - don't tell anyone - they played it three nights in a row at The Warfield Theater in San Francisco in 1991 - December 7, December 8, and December 9.
If you've read this far, and if you love this song, I encourage you to take a moment to set aside whatever else you're doing and listen to this version by Jerry Garcia and David Grisman from 1992 at Warfield Theater - and consider that this song traveled 334 years as documented above to land in this room, with these instruments, and this packed-in crowd. There's decidedly nothing original here; this is a pure traditional. Taste the acid on the tongue of the fiddle, as Joe Craven enters the scene on cue. This is what rebirth looks like:
Understand that this room is something of a mecca for deadheads - in 1980, The Grateful Dead played a run of 15 nights there, celebrating their 15th anniversary. For The Two Sisters to land there after three and a half centuries is an exercise in genuine human time and space travel.
The DeadDisc entry for The Two Sisters notes a stunning 151 versions of the song leading up to, and then influenced by, the moment captured here.
(As an aside: DeadDisc is an incredible repository of information the appalachian traditionals played by The Grateful Dead, Jerry Garcia, and Old and In The Way, and a quintessential example of the bridge between traditional music and cypherpunk culture. It's a great next step if you're doing research on the appalachian tradition of this song.)
Given that The Two Sisters is well-documented for centuries, and appears in several prominent research repositories and collections, it was never at risk of being forgotten.
However, Jerry Garcia's discovery and development of the song was certainly instrumental in recovering it from relative obscurity.
Around the same time as the Warfield performance above, Garcia and Grisman recorded what I surmise is the most widely heard version among our fanbase, for their Shady Grove record. The record was released in 1996, posthumously in the case of Garica, who died in 1995.
Some things I want to note about the Shady Grove version:
There is one other, earlier, notable version by Garcia: the 1987 recording with The Jerry Garcia Acoustic Band. Notable in this version is the shortened lines in first verse.
Internet Age
Nearly every version after this point has one of two qualities:
There is one version that has both: a 2009 cut by Julie Fowlis, a Scottish Gaelic rivival icon and multi-instrumentalist powerhouse. Fowlis' version has the structure and verse of the Garcia version, but with every other verse sung in Scottish Gaelic, with new-agey harmonies.
Another compelling version with a harmonic focus is the one by Gillian Welch and David Rawlings for the Soundcatcher sountrack. This version includes a dose of Welch and Rawlings' commitment to haunting dissonance, and also includes some lyrics which are quite old, and others which appear, as far as I can tell, for the first time. For example, in this version, the sisters are from County Clare, Ireland. And while I can find references in reviews to the sisters being from County Clare, unless I'm missing something, this 2001 version is the first time this detail appears in the song.
In 2005, the Irish folk band Altan used the Welch/Rawlings lyrics in their version.
The version that squarely put this song on my radar, and which remains one of my favorites, is the 2015 Billy Strings version with Greensky Bluegrass. Strings' version is a direct homage to the Garcia/Grisman version, but with a longing and lonesome sound. As might be expected from Billy Strings, the guitar solos are kept very close to the Garcia version, but with a thick and acute bluegrass picking style.
Another "sad live" version - perhaps the best in class for that category - is the 2019 version by Crooked Still with Lula Wiles. Aoife O'Donovan's voice is a certainly a match for the song, and the band's unusual arrangement (with a Cello-driven intro) is a perfect blend of the old and the new. Lyrically, this version stays close to the Garcia version, including the line "Father, oh Father there swims a swan", borrowed all the way from the 1770 "King Lived in the North Country" version.
Another notable Billy Strings performance of this song for me personally is from November 12, 2022 at The Met in Philadelphia. I, my son Fibonacci, and my parents were in the audience (it was my parents' first Billy Strings show). Toward the end of the first set, Billy Strings said, "Why don't we do... mandolin player's choice!"
At this point, mandolin player Jarrod Walker (who happens to be the younger brother of our frequent studio collaborator and one-of-a-kind banjo phenom Cory Walker) said something into the talkback mic, to which Billy looked surprised and replied, "that's not what I expected of you! I was expecting like Ashes of Love or something like that, but I think this is what Jarrod Walker wants to hear."
And sure enough, Jarrod's choice had been The Two Sisters (which they typically write on setlists as "Oh The Wind And Rain"). Jarrod knows that I like and play this song, and also knew I was going to be in the audience that night, but whether my presence had any effect I don't know; almost certainly coincidence I reckon. Either way, I was thankful because it was a lovely version, and it was a great moment for me and my family.
As an aside: after that show, the band's bassist Royal Masat came over to my AirBnb, where we shared a bottle of wine and talked about music late into the night. He was very encouraging about my music and my career, and it was a very validating moment, just when I needed a reason to keep going. Thanks Royal.
As cryptograss
Although I don't closely keep track of this statistic, I suspect that The Two Sisters is the traditional song which we are most often requested to play at various cypherpunk events. I have noted and described the incredible thirst for traditional music among the cypherpunk community at many shows and podcast appearances, and the age and many threads of this one seems to be captivating to the crypto native mind.
We played this song at six show with our dear friend and kickass fiddler Kuba Hejhal, who left this world in September 2024. Kuba was a special cryptograsser; he took very seriously the task of doing justice to this song.
As a band, we've played it at 12 shows so far, and it's always interesting to see how y'all respond. For whatever reason(s), it seems to be counter-intuitive to many people to suppose that these old songs might hit so hard, and be such a quintessential piece of the cypherpunk puzzle. But this song has survived and surged for three and a half centuries with no need for copyrights or centralized authorities to protect it. It serves as an easy and obvious example of what we hope decentralized society to be.
Lyrical variations, and choosing which verses to sing
What instrument? Made from what body part(s)?
I have always sung "he made a little fiddle", from her breastbone. This seems to me to be the oldest, most common, and most compelling story to tell to an audience, especially when we're playing with a great fiddler who can bring the song home.
The Scots Language Centre claims this song as "One of Scotland's oldest ballads", and notes that the song ends with a "wonderful magical ending, in which a harp or fiddle (or even a piano) is made from the sisters white bones and yellow hair, and taken to play at the wedding, where it sings the story of the murder."
However, I can find no version where a piano is made. I also find it slightly unlikely that any version which is contemporaneous with the establishment of the three themes I've outlined above might explicitly mention a piano, since the word "piano" doesn't appear to have been coined until after 1800 (and its predesseor, "pianoforte", from the 1760s), well after these themes had been established in dozens of versions or more.
The sound and song of the fiddle / harp
One of the most frissionate moments of the versions we've come to love, from Jerry Garcia to Crooked Still to Billy Strings, is the line, "with a sound that melted hearts of stone."
This line is old, going back to at least an 1802 version (see Child C in the Child balads link above), and very likely even further - nearly all "Binnorie" variants have a line nearly exactly matching this one:
In most variants, the only song that the fiddle can play is the story of the younger sister's death, and in many, it is a self-reference to the very song being played. So for example, in the "Wind and Rain" versions, we almost always have a line resembling:
With that in mind, I think it's interesting that the 1658 version has the following lines:
"Mall syms" here is a reference to a popular dance of the time, an extant notation of which exists from 1626 - is it possible that The Two Sisters was once set to the music of this dance, long before the events described in this document? This is one of the possible avenues by which this song might be discovered to be even older than 1658 - if some enterprising reader wants to dig into 'Mall Syms', I'm happy to hear what y'all find. :-)
Are the sisters rich? Or poor?
In some versions, the sisters are unambiguously heirs to a fortune. In the 1658 version, for example, they appear to be daughters of the king, and the pretense that the elder sister gives for bringing the younger one to the water is to watch their father's fleet of ships come in.
Child writes (using his alphabetical notation here for different variations):
A 1931 version has the sisters going to the water to sse their father's fishing boat, which seems to support the idea that the story varied with regard to their wealth.
In many "Bow Down" versions, including the 1928 version linked above, the sisters are decidedly not royalty, and instead we have something like:
(which seems strange on its own, and possibly another indication of an independent origin of "Bow Down" - if it's "The Two Sisters", why do these variants describe three of four daughters?)
In many versions, but almost every "poor sisters" version, we see the suitor court one or both of them with a "gay gold ring" or "gay gold pin", which makes sense - a young poor girl might presumably marvel at a gold ring.
However, we also see the ring in older versions where it's not clear whether the sisters are royalty or not, which suggests to me that there may be lost "poor sisters" versions predating the 1900s "Wind and Rain" versions.
The barefoot younger sister
I have often included in our versions a reference to the younger sister being barefoot.
An 1828 book, "Ballads of the North of Scotland" II, 128, lists verses for a version written as "The Bonny Bows O' London". This version mentions the younger sister not wearing shoes:
A 1931 version, written as "Bilnorie", also seems to suggest that she was barefoot, referring to her "lily feet" in verse 14.
A very different 1932 version, as "The Twa Sisters", also mentions a barefoot younger sister:"
An 1860 version written as 'Benorie', documented in the Scottish Highlands, also mentions the younger sister being barefoot. amd the miller pulling her out of the water by her bare feet.
Some other versions mention her having silver slippers (in some caes, which fall off during her float toward the miller's dam/pond) or leather shoes.
I sometimes sing as follows:
"There were two sisters of County Luce" (keeping the reference from the Welch/Rawlings version of their country of origin, but changing it to a county in Michigan, near my own roots, and Billy Strings' as well)
"They were only able to afford the older one's shoes" (which nods to the many "poor sisters" and "barefoot younger sister" variants)
Her hair: what is it used for? Is it fine or dready?
In most versions, the younger sister's hair is used to make either the strings of the instrument or to sring the the bow. In some versions, it is used to make the fiddle itself.
The most common quality of her hair we find is its color: yellow, gold, or "gowd", though, as I mentioned above, this quality only gains quasi-ubiquity once the songs becomes ppopular in Scandinavia.
Several modern versions, including both the Jerry Garcia versions and the Gillian Welch version, say, "he strung his fiddle bow with her long yella' hair."
As I mentioned above, some versions expressly describe her hair as "curly", while others say it is "fine".
And then, several of the oldest versions have the fiddler/harpist taking a "tate" or "tait" or her hair. What is a tate? It appears to me to be an old Scottish word for a dreadlock. (https://www.dsl.ac.uk/entry/snd/tait)
And since it doesn't make much sense to make a fiddle string out of a dreadlock - and because I love dreadlocks - I generally sing "He made a fiddle case of her long nappy/dready hair". However, I must admit that I can find no historical version mentioning that the fiddler makes a case for the fiddle.
So yeah, barefoot and dreadlocked. What can I say? I like stories about hippies.