>> From the Library of Congress in Washington, D.C.
[ Silence ]   >> Nancy Groce: Good afternoon.

Hi, hello, I'm Nancy Groce from the American Folklife Center,
and on behalf of the Center, I'd like to welcome you
to a special presentation
in our Benjamin Botkin Folklife Lecture Series.
The Botkin Series provides a platform
for professional folklorists and ethnographers working both
in the academy and the public sector, to present findings
from their ongoing research.
This series allows us to interact with cutting edge scholars
in folklife and cultural heritage today,
while at the same time building the collections here
at the Library for tomorrow.
The lectures are an important acquisition project for the Center.
These presentations are being videotaped, and they'll become part
of the permanent collection of the Center, and will be developed
into webcasts and eventually appear on the LC website.
Which reminds me, if you have anything that rings or buzzes
that you don't want to be immortalized for the ages,
now would be an excellent time to turn it off.
The Botkin Lecture Series in total,
helps the American Folklife Center fulfill its mandate to preserve
and present the Folklife scholarship that is informing
and impacting our lives today.
I have the honor today of introducing Steve Roud, a folklorist
and an indefatigable researcher from the town of Maresfield (is
that right?), in East Sussex, England.
He was a local studies librarian for the borough of Croydon in London
for many years, and has served on the committee
and as the honorary librarian for the English Folklore Society.
About 40 years ago, Mr. Roud started his Folk Song Index
and Broadside Index as a relatively small scale project,
to find English language folksongs in published sources.
Over the years, his lists have grown in size and complexity.
Today the online Roud Folk Song Index has more than 175,000 entries,
and the Roud Broadside Index comprises more
than 145,000 references, and both have become major go-to resources
for song researchers across the globe.
In addition, in his "spare time," Mr. Roud is also an expert
on numerous other aspects of British traditional culture.
His publications include "London Lore: The Legends and Traditions
of the World's Most Vibrant City," "The Lore of the Playground:
One Hundred Years of Children's Games, Rhymes,
and Traditions," "The English Year."
And is this your most recent "Monday's Child is Fair of Face"?
>> Steve Roud: No, the "Playground Lore" is.
>> Nancy Groce: The Playground is.
But he also wrote "Monday's Child is Fair of Face"
and other traditional beliefs about babies and motherhood.
We're delighted that Mr. Roud has made time to come
to the Library during his recent trip to America,
and I ask you to join me in welcoming him for this presentation
of "Chorus and Verse: The Challenges
of Designing the Roud Folk Song Index."
Steve?   >> Steve Roud: Thank you.

Thank you.   Thank you very much.

As I said this morning to one of your colleagues, it's very nice
to be here, because indexers don't get out much [laughter],

so it's particularly gratifying
to see all you people really interested in indexing.
Yeah, I'm glad you're here.
I should explain that the Folk Song Index started by accident.
If I was planning it, if I was going to do it again, I wouldn't do it.
[laughter] So I was born in 1949, which doesn't mean much to some
of you, but that meant that I was 13, 14 when the Beatles hit us.
So I was that generation.
And for some unknown reason, I don't know why -- well, no, I do know why;
everybody does this -- I wanted to catalog my own record collection,
as one does -- everybody.
Well, I don't know if you still do it with downloads.
But, in the old days with LPs and 45s and things, 78s,
you indexed your own record collection.
And, because I had a few books, I indexed my books as well.
My friends didn't quite understand what I was doing, but,
you know, that's what you do.
And I did that for a while, and then I started buying the sort
of music press, you know, the newspapers that specialized
in pop music, and I thought I could index them as well.
So that's what I did for a while.
And then I discovered folk music in about 1970.
And I thought, oh, this is much more interesting than pop music.
So I threw away my old index, just like that, just overnight.
It was on 5x3 index cards -- you remember those --
in shoe boxes under the bed.   My mum used to buy me 5x3 cards.

Once a week she'd buy me a packet of cards.
Anyway, enough of that.
[laughter] I then got interested in folk music, and I thought, well,
I could start indexing that.   So the same 5x3 cards in shoe boxes.

But the reason for all this story is that, you know,
to tell you that that's how the index started.
It started in my bedroom, because I wanted to get control,
as librarians now call it, control of the data
on my records and in my books.
And I soon realized that folksongs have different titles.
That's the key problem to all of this,
is that folksongs don't have an official title.
The same song can have different titles.
The same title can refer to different songs.
So in my index I had to work out a system for coping with that.
And I was quite happy in my bedroom, you know, indexing these things.
Nobody else needed to care.   And then I went to library school.

In my 30s I suddenly decided, oh, if I'm doing this I might
as well be a proper librarian, I got a degree in librarianship.
And at the same time, the same day in fact,
that I started there they invented computers.
[laughter] And I said, oh, this would be handy, a computer.
I could put my cards onto the computer.
And then they started giving away computers
with cereal packets and I got one at home.
[laughter] And then it was called a database,
and then people started wanting to use my database.
They heard about my database, and it had 5,000 records
in at that time, I think.   It was all on floppy disks.

So I just started giving them to people,
and then they started paying me occasionally for the postage.
But then the trouble started, because as all of you know
who have done any indexing at all,
and you can do your own index quite happily.
You can be idiosyncratic, have your own abbreviations --
you had to have abbreviations in those days
because computers didn't have much memory space.
So all the titles of my books were reduced to sort
of ABFS meant American Book of Folk Songs.
And I understood that, but when other people started using it,
then they didn't understand it,
and that's when my real problems started,
is when I went public, as it were.
And people started saying, well, we don't like this bit,
or why don't you do so-and-so, you know.
And then, then I had a user group.
I had people who were using my index, and they had a say
in what I did, and they started, you know,
some of them gently suggested things.
Some people complained, and so on.   That's life in the library.

Anyway, that's how it was started.
The point being that it started by accident,
it started as an amateur way of dealing with the practical problem.
It had no theory.
It still has no theory, and that's the problem
that I'm constantly trying to catch up with.
If I planned it in the first place, as I say,
if I'd known about being a librarian when I started,
I wouldn't have done it in this way.   I would have done different things.

So, two things I want to do today as we've got time is to take you
through the index so that you will then be expert users.
And you won't complain,
because you'll know why something happens or doesn't happen.
[laughter] But I'd also, if I have time at the end,
like to point out some of the things that are happening in England,
from a folksong point of view, some very exciting things,
digitization projects and making things available around the world.
Even in America, we'll be able to use the British sources more easily.
The indexes, my Folk Song Index and my Broadside Index, are now housed
by the EFDSS, which is the English Folk Dance
and Song Society, or "FDis" for short.
As this is being videotaped,
I'll say this is a wonderful organization,
because they're going to be checking on me.
But EFDSS, again, it's an amateur organization.
It's not government funded.
We don't have anything like your wonderful Folklore Center
in England, because in England we take no notice of our folklore.
It's sort of, you know, "we don't have folklore."
The Scots and the Irish and the Welsh and foreigners,
they have folklore, but the English don't.
[laughter] We don't have any folklore.
So there is no real -- there's no government department
to look after English folklore.
There is hardly any university anymore in England
that does folklore as a course.
So everything is done by people like myself, amateurs and volunteers,
and EFDSS, which covers folk dance and song,
and sometimes the dance people talk
to the song people, sometimes they don't.
And it's a sort of arts and scholarship organization all in one.
It does a good job at the moment, hasn't done in the past, maybe,
but at the moment it's great -- and "hello" to my friends in EFDSS.
They host the index for me, and this is how it looks on the web.
I apologize for the color.
I didn't have anything to do with the red,
a bit garish for an English site, but never mind.
And if anybody's interested in the sites that I'm talking about or some
of the others, I've got a handout with the addresses on,
so you don't have to write down all the web addresses
if you don't want to.   I can give you those later on.

But this is -- if you go onto the EFDSS website,
this is "Search the Roud Folk Song Index."
And when people organizing this lecture asked me,
is the Folk Song Index spelled as folk songs,
two words or as one word?   I didn't know.

I never really thought about it.
It varies, depending on who's writing it down, but apparently
up there, it's just one word, folksong.
So let's -- this just happens to be the first entry in the index.
It's not the first one I did.
But this is one of the problems that I encountered is
that I've used different programs over the years,
because they've invented new programs.
If you're working on your own, you don't have support and so on,
so you just buy the next program that comes and you hope
that stuff will transfer from one,
will go from one program to the other.
So this program, I was using at this time,
put things in alphabetical order of the title, which was quite handy.
So "Apprentice Boy" comes first.
Now, I've actually done a search here under the title "Foggy,"
because I was trying to find the song called "The Foggy Dew,"
which some of you might know.
But of course, I forgot that there's a book called "The Foggy Dew,"
so if anybody should have remembered that, I should have done.
So this isn't a version of "The Foggy Dew";
it's in a book called "The Foggy Dew."
But I'll take you through the different fields --
not all of them, because they're not all particularly interesting.
But song title: in folksong studies, the song title is a major pain.
Because, as I say, there should be a law,
I'm sure Congress could do it [laughter],
that all song titles should be, as they are in my index,
[laughter] you know, there should be a master title
for the whole world, but there isn't.
Not only does the same song appear under different guises,
people use dialect spelling.
Some people use dialect spelling because it's a good thing --
for instance, the Scots "Ould Lang Syne" --
O-U-L- D. If I translated that into Old Long Since and put it out there,
then the Scots would all complain, quite rightly.
But if you think about it, there's at least eight different ways
of spelling the word old -- O-L-D, O-U-L-D, A-U-L-D, O-L-E,
O-L-apostrophe, and so on.
Now, some of those are good spellings,
as I say in Scottish terms.
Sometimes they are people trying to do an African American accent,
for instance, you know, in some of the spirituals.
So it's not the way they've chosen to spell it,
it's the way the collector has chosen to spell it, and so on.
So spelling is a real problem in that first --
so in the first field we come across, the "song title,"
is already major problems.
In Britain, we've flirted with the idea of master titles,
and there is a field which you don't see,
because I'm not using it, called "master title."
But then we quickly discovered that we could just
about do a master title for England that some of us agree to,
but same song would have a different master title in America,
and a different one in Australia,
and certainly a different one in Scotland.
So we would need at least five different master titles.
So you end up -- we won't do master titles, ok.
And just the last thing I think, as a librarian,
I really feel that I have to reflect how it is spelled in the book
or in the record that I'm indexing, because somebody working
from that book, they don't know how else I have spelt it.
So if that was there as "'Prentice Boy," which it often was,
apostrophe "'Prentice Boy," that's what I should put in the index,
because somebody working from this particular book looking
for other versions will look up "'Prentice Boy."
Do you see what I mean?
So even if the book is wrong, you have to reflect what's
in the book or on the record.
I should have said at the beginning, that the index now tries
to cover published books, published recordings,
unpublished manuscript collections, and unpublished sound recordings.
In the English-speaking world, every single folksong --
any English spoken word in English is going to be in my index one day.
You know, you've got to think big, don't you?
But anyway, this happens to be an English book.
So "tune" is there for the second field, is when there's a named tune.
For instance, broadsides are often to the tune of "The Foggy Dew."
So if it's a named tune, because some people are looking for that.
The third field, "first line," is --
and I said, a lot of people don't bother with that.
Now, I'm quite pleased that I did think of putting in first lines,
because with folksongs, if you just got the title of a song,
if it's a bit better than as AI Roved Out,
or something like that -- but normally with a folksong,
if you got the title, you can be 50 percent sure what song it is,
you know, roughly.
But if you've got the title and the first line,
it goes up to about 75 percent.
It doesn't always work, but the first line is very time-consuming,
and a lot of other indexes don't do that.
But I do recommend it, because it is very handy.
The next field, "type."
So that just tells you that this is a book rather
than a sound recording, because you might be just searching,
as often people do, for an audio version.
You might just want audio; you might just want printed.
So that field "type" will say "book, sound recording, broadside,
manuscript," that sort of thing Asource.
@ This is a book by Frank Purslow called "The Foggy Dew,"
and under it, "bibliography ref(s)," theoretically, I click on that,
then I do -- then you get the bibliographic reference
or the address of the archives
that that manuscript is in, that sort of thing.
So the source field in is shortened,
but you can find quite easily where it's from.
[ Pause, background noise ]
>> "Previous source," again, this is important in folksong scholarship,
because people are often using things collected by other people.
So this fellow, Frank Purslow, has used things
from G.B. Gardiner manuscripts, so the idea being
that you can trace back where the thing has come from.
I also use that field.
That's sort of backwards in time, if you like -- I also use that field,
I'll show you in a minute, where if it's now available on the website,
the website address goes in that field,
so you can actually click on it and go to it.
That's the theory, anyway.
Then you got things like the "performer's name,"
the "place name," the "date," and the "collector."
Now, in folksong studies, as most of you know,
these are very important things.   At least 50 percent of the people

who are using my index want a particular song
from a particular place.
If you're doing "Barbara Allen," the most popular of the Child ballads,
there's a thousand versions in here.
That's no good to anybody, so you want to narrow it down;
I only want the American versions, or only want the Canadian versions.
So all of these fields are searchable
and sortable, so you can do that.
And just recently, my friend, the website man,
has given me this rather nifty thing.
Look at that!
Now, I couldn't do that with my old shoe boxes, that's for sure.
I mean, a lot of these places I've never even heard of,
but that looks pretty accurate to me.
And he's also done it for the American ones as well,
so you can do that with the American place names.
So the date the song was collected, the collector.
"Contents," again, this idea that some people are only looking
for versions that have the music.
Some people are looking for versions that have the text.
So if it's a fragmentary text, then it will put it there
and it will say text in brackets, "one verse," which often happens
in the collections, or it says music, or it would say audio
if it's a sound recording.
Then, skip down a bit to subjects, where it says Aapprentice,
loves young woman, >Shady Grove,
dark and rolling eye, master disapproves.
@ If I was starting again, I would do myself a thesaurus
of subject terms, which some of you here are doing.
Not as easy as it sounds.
You could say, I'll do a thesaurus of subject terms.
It's quite easy to say, but it's very difficult to do.
And in the absence of a subject index, I can't find a subject index
that really helps with folksong.
There are lots of subject indexes out there,
but they are other subjects, or they're so general
that they don't help, or they are culture-specific.
You know, because a subject index
in America is not necessarily going to work in Britain.
Some of it will, some of it won't.
So when I hit a song that I haven't met before, I do a just sort
of a brief thing like that, which gives you some idea
of what it's about, so that when somebody asks me, oh,
there's that song that mentions shady grove;
but it's not AShady Grove,@ it=s some other song,
then at least I've got some chance of sticking
in "shady" and finding it.
Because a lot of people who are looking
for songs don't quite know what they're looking for.
They've heard it; oh, it's that song my Mum used to sing
and it had something about a shady grove in the chorus or whatever.
Theoretically, what we need is a proper index,
and all the songs indexed and all the motifs indexed.
Now, until that happens, you're stuck with this sort of thing,
but you'll be surprised how often that works.
And just down the bottom that's "notes," "one verse from singer,
the rest of text from broadside."
In other words, this book has cobbled together
that version from different versions.
But the most important invention of my life
and of this indexing project, is the Roud number,
which we'll now talk about, which is there, 1671.
Going back to my cards, I soon found that if I had three titles
for a song, I had to write all three titles on all three cards.
And then if I got another song with a different title,
I had to raise another card and go back and write
that title on all four cards.
Now, I soon discovered that this was not a good way of spending my time,
and I didn't have any staff, remember, to do it for me.
Otherwise that would have been easier, just say,
"Go away and write those cards."   So I invented a Roud number,

didn't call it a Roud number; it was just a number.
And this was before I heard about Child.
I didn't know anything about Child or Laws
or any other numbering system.   Again, if I was starting again,

the first 305 of my numbers would be the same as the Child numbers.
But I didn't know anything about Child, so I started with number 1,
which happened to be AThe Dark-Eyed Gypsy,
AGypsy Countess,@ what you call it?
AGypsy Davie@: Child number 200, Roud number 1.
And then the next one happened to be AThe Unfortunate Rake,
AThe Streets of Laredo@ -- that sort of corpus -- number two.
AThe Seeds of Love,@ which should have been number one really,
because that's the first song collected by Cecil Sharp,
that's number three, and so on -- up to 22,000 now, with these numbers.
But the idea being that if you want all the versions of this song,
AThe Apprentice Boy,@ under all the different titles that I've come
across it, you just do a search
on Roud number 1671, which I hope will work.
[ Pause, background noise ]   >> Right, that was quick, yes.

So I've got 28 versions of this particular song.
[ Pause, background noise ]
>> There we go, Canadian version, in Helen Creighton's book.
Simply, really, isn't it?
You just give things a number, search for the number.
But, that then brings into the idea of what is "the same song."
How do you identify ones -- You've got five texts, all about,
boy meets girl, boy murders girl, boy gets hanged.
[laughter] That's the pretty standard folk song,
murdered sweetheart.
If I had a Thesaurus I'd call it the murdered sweetheart motif,
but it's the same thing.
But some people will want to bring together all the murdered sweetheart
songs and they don't care which ones they are.
Other people want a specific murdered sweetheart song
about a particular murder
or a particular mythical murder, as it were.
So should the Roud number refer to things
that are definitely the same song because they've got the same words
in there, or some of the same words?
Or should it be sort of more generic?
Now, I didn't know anything about this when I started,
so I started off with a sort of generic idea.
So there are lots of alphabet songs, for instance.
I don't know if you know them.
There's a lumberman's alphabet, A for axes and B for something,
but there's the same song, sailor's alphabet, which is A for anchors
and B for boatswain, and so on, goes through the alphabet.
And there are miners' versions.
There should be a librarian's version, but I don't think there is.
But when I started, I gave all of those the same number,
because they're all alphabet songs.
And you could argue that the text is the same, because they all go, A, B,
C. But the bits that the As and the Bs go with aren't the same.
But then people started saying, well,
actually I just want the mining ones.
I think the mining ones are different, too, the lumbering ones.
And they are different, too, the sailor's ones.
And so I had to renumber my alphabet songs.
So if you want alphabet songs,
you now have to search for different numbers.
But if you want the lumberman's alphabet song,
you can go straight to it.
And this is the crux of the problem of indexing like this,
is that by making it easier for one group of users,
the ones who want the specific song, I've made it harder for those
who want the generic ones.
So there's this constant problem, now that the index is public
and I can't just do what I want, of juggling different peoples' needs,
different users, because as a librarian,
my concern is to serve my users.
The whole thing should be designed so that it can be used.
It's no bloody good at all if nobody can use it.
So constantly I have to think, okay,
who's using my index, and what are they after?
And this has ramifications for the whole thing;
it's not just the Roud number.
But if, as I know, that the index is now used by proper researchers,
real people with Ph.D.s and that sort of thing,
to do real index research, but it's also used by complete beginners,
by students, by people who have just stumbled in and,
you know, want to find a song.
So we have this continuum of users from if you like,
we'll call them the lower end but maybe they're the higher end,
but the lower end of people who are not serious researchers,
people who are performers, people who just want a song,
just want to find the words to a song, at one end.
And at the other end you've got the people who really do need
to know exactly when a song was collected and why it was collected
and who collected, and so on.
And in between, you've got all sorts of other users.
And of course, the same users can slide up and down the scale.
I can be a researcher one day,
but the next day I quickly want the words to something,
because I'm going to sing it tonight or something.
So there's this constant juggling for who I'm trying to serve,
and of course I'm a particularly nice sort of chap;
I want to serve everybody.   That's the purpose of it.

That's not always possible.
If you go back to the source field, the --
Helen Creighton's book, that's fine.   Some people would argue

that I should be indexing just the authoritative major sources.
So for instance, the Carter family.   You all know the Carter family.

Yes, I love the Carter family.
There is a box set of CDs available in Britain, probably available here
by the Baer family, of 5 CDs or something --
every song that the Carter family ever recorded.
But it costs 150 quid.
So I could index that and say, "I'm not going
to index any more Carter family CDs,
because I have done the authoritative version."
But nobody, none of my -- a few of my users, put it that way --
will have 150 pounds' worth of Carter family CDs.
But what they will have are the cheap reprints
that you can buy in your average shop.
You know, they will have access
to the popular rather than to the scholarly.
So I, therefore, have to index the lot.
Now, there's something like Carter family
or Woody Guthrie or Leadbelly.
There's a new Woody Guthrie CD out every week.
And some of them are going to last, some of them aren't.
Some of them are going to disappear in five minutes.
But I feel that I should really index those,
because those are the things we've got access to now.
It's no good indexing the original 78s alone,
because nobody else has access to those.
So you end up, as I say, trying to cater for so many different groups,
that sometimes you don't please anybody;
sometimes you can hit a happy medium.
Then there are interest groups within the folksong world.
For those of you who are outside the folksong world might thing
that we're all the same, but we're not.
And as soon as the index went up online and I proudly said, oh,
it has American stuff on there, the first e-mail I got was, "Well,
you haven't got the hillbilly records."
Oh, okay. We better put some hillbilly records on there.
Or, "You haven't got the shantys," or,
"You haven't got this," "You haven't got that."
You think, "Give me a break; I'm doing my best here!"
[laughter] But it's true.   It's true.

If it's going to be comprehensive and usable by everybody,
as I claim it will be one day, then those groups have just as much right
to demand that their particular brand of folk music is on there.
And why shouldn't it be?
So every now and again, I sort of have a brainstorm.
I think, right, we're going to do hillbilly records,
and so for a few weeks I'm beavering away, doing hillbilly records.
But then I get an e-mail from the shanty people,
and I have to do the shantys.   Now, I don't get on with shantys.

Yeah, shantys are a problem.
[laughter] Getting back to the Roud numbers, Roud numbers only work
if you've got a relatively conservative corpus
of material to work from.
If somebody 100 years ago sang a song,
and the next person sang a song and the next person sang that song,
the same song, if they changed things willy-nilly,
just decided to change things overnight, the song,
by the time it had come through three or four generations,
would be a completely different song.
There'd be hardly any contact between them.
But certainly in the British tradition,
and I think to a large extent, in American say,
in the white American tradition,
folksong singers are basically conservative, in that they do,
more or less, sing what they learned.
They do fiddle around with things, and there are good singers
and there are bad singers.
People forget things, people add things,
but not as much as we sometimes think.
Most people, I would say, if you asked them, they would say, Aoh,
yes, I sing it the same as Granny did,@ and they do more or less.
But shantys doesn't fit that bill at all.
Shantys are much more fluid.   And they really did make up things

and they really did combine songs and so on.
So the system of numbering for this sort of Roud system works as far
as I'm concerned, for what's to me the main corpus,
because that's the Ang --
if you like the Anglo American white tradition.
Doesn't work too well with shantys.
And it's pretty hopeless for blues and that sort of thing.
Because again, the songs are too fluid for me to say,
"this is a version of this."
Now, again, there should be a law that tidies it up.
[laughter] But until there is, really,
at the moment I sort of stay away from blues.
There are some blues things in there, but they tend to be the sort
of blues ballads, the ones that ever -
tell a story that I can actually identify and so on.
So as I say, every field has a problem.
[laughs] There's nothing here that is simple,
that I thought it was simple when I started.
And the real crux of it is -- and somebody at the end will ask me this
if I don't say it now -- how do you define a folksong?
And the answer is, I don't.
I leave it to the collectors to define the folksongs.
Because really, all I'm doing is coming along after the collectors.
If a song hasn't been written down or recorded
or in some way perpetuated in stone, then I can't index it.
If I walk out here and hear somebody singing a song, as I walk past,
I can't index that performance really,
because nobody can go back to it.
I can only index what's been documented.
So I don't have to choose what's a folksong.
I just have to choose which collections I trust.
So Helen Creighton, who did her songs and ballads from Nova Scotia,
if she included music hall songs that she had collected from singers,
then that to me is good enough.   That's a folksong.

Now, if I come across a book and they don't tell me
where they got their songs from, then I'm in the dark.
I don't know if they're traditional.   I don't know if they're folk or not.

They might have just been made up the day before.
So in effect, I don't index them in this index.
They go into the Broadside Index.
And I'll just explain what the Broadside Index does.
It's exactly the same format, so that it's interchangeable
with the Folk Song Index.
But that's where I put songs that appeared in --
do you know what broadsides are?   Everybody know broadsides?

Yes? Chapbooks, songsters, music hall ditties, and early recordings,
and that sort of stuff -- in other words, songs,
the material that the folk song scholars need access
to to do the history of folksong.
Because one of our problems in folksong scholarship is
when suddenly, somebody pops up with a song, and we say, oh,
that's very nice, but where did it come from?
When was the first version?   Is it the same as this other one?

That sort of thing.   That's what we do.

We're unsure of ourselves if we've got no history.
So the Broadside Index --
I could throw anything in the Broadside Index
because it's not saying that it's an accredited folksong
or anything like that.
So if you're looking for a particular song,
especially if it's a 20th century song
or late 19th century music hall-type song, then if you look
in the Broadside Index, you'll find references to broadsides, chapbooks,
books, recordings and that sort of stuff.
I mean, I don't put all pop music in there, obviously it tends
to sort of peter out about 1920.
Things went downhill after about 1920.
But that's there as a sort of mirror of this,
and it works in exactly the same way, although fields like
"place collected" and so on aren't there, but other fields, like,
the "author," author of the song and that sort
of thing becomes more important there.
Right. That's all of what I'm got to tell you about the index.
I just want you to know that I've been working very hard on this
for your benefit all these years, [laughter] and I'd
like a little credit for this.   [ Laughter ]

>> Right. A few of the things that's happening in England.
Now, I'm not talking about Scotland, Ireland, and Wales here;
I'm just talking specifically about England.
The English Folk Dance and Song Society, as I say,
is an amateur organization, but that doesn't mean that they do things
in an amateur way, especially nowadays.
And the library is called the Vaughan Williams Memorial Library.
Most of you know that Vaughan Williams was one
of our early collectors, also wrote a few symphonies and things.
But we like him because he collected songs -- 1903 and onwards.
Also, Cecil Sharp, one of our heroes,
although he gets bad press these days, which I might come
onto in a moment, but Cecil -- There.

There he is in all his glory.
Now, and some of you will know that Cecil Sharp actually came over here
and collected -- it's one of your important collections, really,
isn't it, in the Appalachians, 1915 to 1918.
He came over for another reason.
He came over to do the music for Midsummer Night's Dream that was
on in Broadway, and he was the music director.
And I forgot the name of the Lady.   What's the name?

Dame Olive Campbell, sought him out and said, oh,
I found these rather interesting songs down there in the Appalachians
and Cecil fell through the floor and thought, "My goodness,
this is better than we've got at home
because it's more complete versions, and it's more real."
So he then came back later anyway and collected.
One of the important things he did, as you know, he collected the tunes,
because he was more interested in the tunes,
while your guys were more interested in the words at the time.
Broadly speaking, they came from literary backgrounds;
he came from a musical background.
But one thing that they've now done, at Cecil Sharp House is
to put Cecil's Appalachian diaries online,


>> Stephen Winick: There's a link right at the top.
>> Nancy Groce: There's a link.
>> Stephen Winick: Those diaries can now be-
>> Steve Roud: Where is it?   There it is.

That's it; thank you.   Here we go.

"I Greeted the New Year in Bed."   Is that the right page?

>> Female voice: Yes.
>> Yes. APandemonium in the streets of New York.
@ I've just come from New York; it's very similar.
[laughter] So here he is in New York.

But -- If you can't cope with his handwriting --
I think this is rather nifty -- you can --
you have a transcription and you also have --
the links telling you a bit about the people and the places.
Now, that's something -- this is a work in progress,
like all of these things.
People are still being identified and indexed and so on.
So this is one of the things, one of the small things
that EFDSS has managed to do, and as I say,
they're an amateur organization but they don't do things
in an amateur way these days.
They had another project, called Take Six.
Which one's Take Six?   Oh, go to the arrow.

Right, now this -- I should say that in the realms of folk music
and digitization, you guys were there long before us
with your American Memory Project.   Your websites are excellent.

They really are, and they're still better than we're doing now,
and they've been up there for ten years, something like that.
So I have to hand it to you that your stuff --
But we could do with some more, please.
[laugher] That was a good start, but we want some more.
This was a project, takes six of the major collections --
you may well notice that there are seven of them there [laughter] -
but we'll worry about that in a minute.
And to digitize the manuscripts, so this is working
on manuscripts rather than sound recordings.
So the second one, the debonair chap with the pipe.
That's George Butterworth; you might know him.
He was a classical composer, but was killed in the First World War,
so didn't get very far in his career.
But he was one of Cecil Sharp's Morris dancers
and collectors and so on.   And there he is.

We search the catalogs.   So again, you can search for songs.

So if you know your Roud numbers, which some of them I do,
I search for 558, which is "The Foggy Dew,"
which was the one I was trying to find in the first place.
Ok. So there are two copies,
two versions in the Janet Blunt collection,
several in the George Gardiner collection, and so on.
And choose one at random.
That gives you the details, and then you get the original field
notebooks, and music, and so on.   So this is for that particular song.

Hammond apparently did not note the text from this singer -- fine.
But what he did write was, "usual words."
[laughter] That's what a lot --
a lot of the collectors did that -- fair enough.
The poor chaps were out there with pen and pencil,
and their bicycles, and things.
[laughter] But this is rather neat, in that although any archivist
in the audience might disagree with me here, we took the decision
to bring together things from different parts of the collection.
George Gardiner is a good example in that he went out with his notebook,
but he wasn't competent enough to note the music,
so he had a musician with him.
So he noted the words, and then he went back home and he wrote them
out in his best handwriting.
The musician he had with him did the same,
did the music, came back, wrote it out.
So you've already got four pieces of paper referring
to that one singer singing that one song.
Sometimes the musician went back, or George went back
to check the words or the music.
Sometimes the repeat visit was with a different musician.
So you can get eight pieces of paper.
[laughs] And then after that, somebody came
and transcribes the whole lot and typed out the words.
You can get ten pieces of paper.
Now, if I was a proper archivist, which I'm not --
I'm a librarian, so I can do what I like --
archivists tend to like to keep their collection in the right order,
so George's notebooks are there, and then his texts are there,
and then somebody else's words are there, and so on.
So they -- if we just digitize them, they all appear in different places.
So if you are looking for that song collected from that singer
on that day, you either have to ferret around all over the place,
or we have to do all these links.
And what we decided to do is to bring them all together.
So you search for a particular song, and as in that case there,
you had all three bits of paper for that song collected
from that person on that day, altogether.
Now, I think that's a sensible thing to do, because we know
that 90 percent, at least, of people using this website are looking
for specific songs.
They're not looking for, how did George Gardiner go about collecting,
or what was he doing on a particular day;
they're looking for a particular song.
So my job as a librarian is to cater for that need
and pull the stuff together.
So that's the way we did it, but of course,
the first complaints we got was, "you can't browse the collection."
People have said, "What I want to do, I don't want to search
for a song; I want to start from page 1 and work your way through."
And we say, well, actually we don't really want that.
If you could see page 1 and page 2 and page 3,
they are almost unintelligible.   You'd soon give up.

You'd get up to page 10 and you'd give up.
But that's what some people want; some people want to browse.
Now, we design the -- I mean, I wasn't the designer here.
Malcolm Taylor was the designer, but I helped.
We designed things for searching rather than browsing,
so when the next update of this project, which I'll tell you
about in a moment, happens, then we will have another,
browsers go this way; searchers go that way,
and hopefully we'll sort everybody out.
Now, Take Six had six collections.
Since then, one project down in Devon has done the same thing
with the Baring Gould -- you may not have heard of --
Sabine, no, Sabine Baring Gould--
was a parson and a squire down in Devon in the --
he was one of our earliest collectors.
He was also a novelist, and he wrote
"Onward Christian Soldiers" - you know, the hymn, yes?
Anyway, quite a sort of -- he had 13 children, I think;
couldn't quite remember their names, but who can?
But with Baring Gould, because it was done by different people,
though to our template, they actually did the things separately.
So you got "The Foggy Dew" by Baring Gould there B,
and "The Foggy Dew" tune only and "The Foggy Dew - A",
and another "Foggy Dew - A" because he copied it out of another book.
Now, he probably only collected it once, or maybe twice,
but you've got all these different headings for,
in effect, the same thing.
So again, that's good for the browsers; they can go through page
by page, but it's not good for the searchers.
But the really interesting -- well, really exciting thing from our point
of view, is that EFDSS has just got a grant
from the Heritage Lottery Fund.   Do you have such things over here?

You have lotteries, right.
Now, the lottery is -- it's a commercial lottery,
but when it was set up fairly recently, some of the money
from the lottery goes into good causes.
So it's not actually a government grant;
the lottery does its own decision-making,
and often it's 5 million pounds to buy a Rubins
who would be leaving the country or to do up a castle or something--
it is England, after all.
But Malcolm Taylor at EFDSS has worked very hard
to get half a million pounds
to digitize all the other major collections, folk song collections,
in England - and as I say, not Scotland, Ireland, or Wales.
And as from later this year, that project will be going ahead,
including Cecil Sharp and his, and therefore his American,
his Appalachian collecting, all his pages of songs,
the original manuscripts will be there.
They'll be indexed, they'll be Roud numbered, you know,
all of the paraphernalia that goes with it.
And also, I forgot to put a mark here, but you can go to their site.
You can see the photographs that Sharp took of his singers.
There's some wonderful pictures
from the Appalachians, in 1916 of the singers.
They're already up there, in fact, the photos.

The future, in a way - - in a way, from folk song studies in England,
we're entering a sort of golden era, because each collection's,
they've been available before, but if you wanted
to see the Cecil Sharp collection you'd go to Cambridge.
If you wanted to see Frank Kitson's collection, it's in Glasgow.
In other words, you have to be a gentleman scholar like myself,
with all the time in the world and all the money
in the world, to actually do it.
[laughter] But for the first time ever, all these collections will be
in one place and available to everybody.
One other sort of strands to what we're doing, is in broadside,
broadside research -- we mention broadsides
because if you're doing the history of folksong,
you bump into broadsides in England, especially, very quickly.
They're the only way you can date most songs,
is if the thing was in print at any time.
So if Cecil Sharp collected the song in 1903, that's fine,
but you don't know anything about it unless you've got some
printed source.
So broadsides are the important thing.
And there's lots of collections.
Now, the Bodleian Library has pioneered, if you like,


So again, you can do--   [Pause]

>> Right, so again, we have now access to the 30,000 Broadsides
in the Bodleian, and wonderful resource it is.
Now, actually I don't have any say at the Bodleian,
so they didn't put Roud numbers on this index,
but it works the other way around, that if I found a song
in the Bodleian in my index, you'd just click on field
and it would take you direct to this site.
And really, that's the future for my index, is that--
Some people say that once everything's digitized and out
in the way, we don't need indexes like mine.
And it's true in a way.
You can search on Google for Barbara Allen.
You might get 3 million hits, which is not very much good to you,
and you won't know which ones
of those you're interested in and so on.
My point of view is that the more stuff is available,
the more you need people to do the indexing
to bring it together, to give you shortcuts.
So if you're interested in folksong, then if you go through my index,
you may still get 1,000 hits, but that's better than 3 million,
and you'll know that that thousand is in some way connected
to being traditional, or collected,
or on broadside -- you see what I mean.
I said, of all that information, now I'm going to index this little bit.
If you're interested in this little bit,
come via my index rather than via Google.
But indexes are really rather old-fashioned.
My index, you'll find that it's in Creighton's book.
You then have to go and find Creighton's book.
You then have to go and find Creighton's book.
Now, I've got on my shelf, because I collect that sort
of thing, but most of you may not.   Most of you don't work here.

In other words, it's a sort of two- or three-step information thing.
Now, if you think Google's got Creighton's book,
you might find it on the web.
What needs to happen is that my index has to be more interactive
and more interconnected with the digital world, in the same way
as I'm doing with the broadside ballads, for instance,
because that's there and I have access to it.
If I find it on there, I can just stick a link in,
and I know that the Bodleian site's going to be there next week.
One of the problems is if you don't know it's going to be there.
So that's the future when I get 'round to it,
is to throw away the 5x3 cards in the shoe boxes
and get into the 21st century.
And there's no reason -- it's not that it's down to copyright.
If somebody is willing to host my index with the images,
there's no reason why I can't scan the image as well as indexing it.
Okay, it's another few minutes, but it saves the user a lot of time,
and that's what index is for, to save the user a lot of time.
So that's it.   It's dead on 1 o'clock.

[ Applause ]
>> Nancy Groce: Is there time for a few questions?
>> Steve Roud: Time for questions.
>> Nancy Groce: And I'll hand out your handout while you're answering.
>> Jennifer Cutting: I think there are only twenty copies.
>> Steve Roud: There's two pages there.
That just gives the websites I talked about
and some others I didn't have time to cover.
If you want them electronically, you can email me
and I can send you the list if you don't want to type
in all the squiggles and things.   Yes, Steve?

>> Stephen Winick: I 'm wondering whether the index deals with,
and this is something I've never noticed this in looking through it,
deals with unpublished archival resources at all?
>> Steve Roud: Yes.   >> Stephen Winick: It does.

For example the Creighton you showed, of Ben Henneberry singing
"The Foggy Dew" - I am reasonably sure we have a copy
of the sound recording on which that was based in our archive
and there would be another copy in Nova Scotia.
>> Steve Roud: Yes.
>> Stephen Winick: Are those in the index?
>> Steve Roud: They -- Some of them are.
The situation now is that when this project with EFDSS has happened,
and if I index at the same time as their catalog.
I will be doing the cataloging to some extent.
When that is done, say, eighteen month's time, or two years time,
virtually all the English manuscripts, and books,
and field recordings that aren't in private hands, will be indexed.
So, England done!
[laughs] If you look at the American holdings on my index,
if it's in a published book, it's more or less likely to be in there.
I have done most of the state collections,
the Helen Creighton's in the USA and Canada.
But for the unpublished collections, if I have been able
to persuade the librarians to send me a Xeroxed copy,
for a small collection, for instance,
then the whole thing's in there.   Beck, E.C. Beck's collector?

His stuff is there, because I can get access to it at home.
In other words, if I can get access to it, I can do it.
One of the things, once I get all the English stuff out of the way,
I would like to come back to do more American stuff.
Because most of my users are American, probably.
And would like to make sure
that I've got have your stuff here on there.
Because there is no point in not doing it.
>> Jennifer Cutting: Are you doing the James Madison Carpenter
Collection right now?   >> Steve Roud: Yes.

Forgot to mention him.
James Madison Carpenter Collection which is here,
and it does have American stuff in it, but the most important material
in there, from our point of view,
was that he collected in Britian in the 1930s.
And he made recordings in the 1930s, which nobody else was doing.
So there are thousands of songs in that collection
which I am gradually adding.
Now there is this index up so that I check things.
Within a year or so the Madison Collection will be there as well.
But certainly -- as a researcher,
obvioiusly the manuscript is more important than the secondary book,
but most of my users want to get to the books first.
Yes?
>> Female voice: I have two questions.
First off, as an independent I researcher I find
that when I approach librarians with a brilliant idea to do an index,
they are kind of like, "Uh, oh!"
How did you convince people that this was worthy
and get the sort of cooperation you got?
>> Steve Roud: Just by trial and error, I think,
just by it being there and people starting to say, "Oh,
this is neat, this is useful."
To give them their due at the Vaughn Williams Library, Malcom Taylor,
who has been there for thirty, over thirty years now --
when I took the index in to Malcom Taylor on floppy disks
and he loaded it up, he immediately saw that it was useful.
He said, "This is just what we need, because people come in here asking
for a song and you've indexed things.
We don't have to do any more indexing."
He said. So he encouraged me to develop it
and to keep at it, as it were.
And he -- In England he is a key person, if you have convinced Malcom
of something, then you're half way there.
>> Female voice: Secondly, [inaudible] tunes like the
"Star of the County Down?"   >> Ah, right, yes.

No, is the answer.
I am not a musician anyway, so I am not qualified to do the tunes.
But I don't think that you can do tunes
in the same way that you do texts.
You know you can get two texts together and say "Well,
it's 50 percent the same words," or "It is 60 percent the same words,
whatever, say, I am going to have a cut-off at 60 percent,
so you can say that is the same song, you know what I mean.
You can't do that with tunes, really.
There are too many variables there.
A text means something, read it and everybody understands what it means.
You can argue about it and say what it means,
but a tune doesn't mean something.   A tune is a tune.

I am not saying in the right words, but you get what I mean.
If someone can come up with a proper system for identifying, comparing,
and labeling tunes, then I 'd more than happily put in a reference
in my index, but I can't at the moment.
The only thing I can do is if there is a named tune, like "Greensleeves"
or something, and the broadside says, "Sung to the tune
of Greensleeves," then I can put in "Greensleeves,"
but that's the best I can do.
>>Jennifer Cutting: There is an Irish tune motif index,
I think, put out by [inaudible].
They exist, but not for English folksongs.
>> Steve Roud: Not for English.
I am not sufficiently musical to say whether they work or not.
But I am told that none of the systems that exist
at the moment are very good.
>> Nancy Groce: Maybe one more question?
>> Female voice: I was thinking of what you were saying
about linking more to the web.
Creighton's book is 1932, so probably you can't get
to the full text in Google Books or [inaudible].
There are copyright issues.
But any of the Child collections you probably can get to.
We make great use of unpaid labor here, students and interns.
And it seems that that is a manageable project
to have some student go through and find the links
to which books have been digitized make them available.
That seems like a concrete, easy project.
>> Steve Roud: Yes, you are right.   And as part of the new project

that EFDSS is doing called the Full English Project -
I don't know if you know, but "full English" means,
a full English breakfast.
If you go to an hotel and have "the full English," then you have,
you know, the whole thing.
So it is a neat name for the project.
As part of that, I am hoping
that I can scrounge some help to do exactly that.
Putting links to material that's there, as long as it is stable.
Because I've come unstuck in the past spending lots of time linking
to things only to have them disappear.
And there is nothing more annoying than links that don't work.
So, nowadays things are more stable than they used to be.
But as I said right at the beginning, one of the problems
with a long project like this is that you always have your history
on your back, as it were.
The baggage of old programs, and, you know, the way you do this is
because twenty years ago you couldn't do it otherwise.
Whereas nowadays you can do more
or less what you want, in database terms.
But not so long ago, you had to worry about the length of a field,
and you couldn't put letters into a number field, and all that sort
of stuff was a major constraint.
Sorry I wandered off the topic there.
But yes, in answer to your question is yes, I hope so.
>> Nancy Groce: So please join me
in thanking Steve Roud for coming today?
[applause] Our next Botkin Lecture will be May 5th.
We are having Professor Don Yoder come to talk
about Pennsylvania Dutch culture.
We will look for this to go up on the web
in the not-too-distant future.   Thank you all for coming.