Oh Susanna Artigo

"Oh!
Susanna": Oral Transmission and Tune Transformation

Author(s): John Spitzer
Source: Journal of the American Musicological Society, Vol. 47, No. 1 (Spring, 1994), pp. 90-
136
Published by: University of California Press on behalf of the American Musicological Society
Stable URL: http://www.jstor.org/stable/3128837 .
Accessed: 16/06/2014 05:57
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
University of California Press and American Musicological Society are collaborating with JSTOR to digitize,
preserve and extend access to Journal of the American Musicological Society.
http://www.jstor.org
This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

All use subject to JSTOR Terms and Conditions
"Oh! Susanna": Oral Transmission and
Tune Transformation*
BYJOHN SPITZER
Versions
of "Susanna"
Journaldated i i September
A NOTICEin the PittsburghDaily Commercial
1847 announcedthe performanceof "SUSANNA-A new song, never
beforegiven to the public,"upcomingat Andrews'Eagle Ice Cream
Saloon.'The singersat the EagleSaloonwere Nelson Kneassand his
blackfacetroupe,who subsequentlytookthe song on tour in the Ohio
Valleyduringthe fallof 1847.2 Otherminstreltroupesseemto havesung
"Susanna"as well, among them the Sable Harmonists,with whom
Kneasshad previouslybeen associatedand who visited Pittsburghin
SeptemberI847.3The ChristyMinstrelsin New YorkCityalsoseemto
havehad"Susanna" in theirrepertory,forthe firsteditionof "Susanna,"
publishedby Holt in New YorkCity, proclaimson the firstpage:"Oh!
Susanna.Sung by G. N. Christy, of the Christy Minstrels."Holt
registeredthe songforcopyrightat the DistrictCourtof SouthernNew
Yorkon 25 FebruaryI848.4 The earliestprintto attribute"Susanna" to
Stephen Foster was publishedby W. C. Peters in Louisville,Kentucky,
*This article grew out of a paper given at the Fifty-sixth Annual Meeting of the
American Musicological Society, Oakland, California, November i99o. I would like
to thank William Austin, Andre Barbera,Jae Kim, Deane Root, and Steve Saunders
for their ideas and suggestions.
' John Tasker Howard, StepbenFoster,America'sTroubadour (New York: Thomas
Crowell Co., 1934), 127.
2
Ernst C. Krohn, "Nelson Kneass: Minstrel Singer and Composer," Interamerican
MusicalResearchYearbook7 (197 ): 23.
3 Robert W. Gordon, "Preliminary Report on Stephen C. Foster's 'Oh!
Su-
sanna,' " August 193I, typescript in the Foster Hall Collection of the Stephen Foster
Memorial, University of Pittsburgh, Pittsburgh, Pa., p. 4. Gordon also notes a
performance of "Susanna" in Pittsburgh in July 1848 by the "Tyrolean Vocalists,"
who included Nelson Kneass, M. Campbell, and J. Farrell.
4 Gordon, "Preliminary Report," 5.

"OH! SUSANNA" 91
and depositedfor copyrighton 30 December1848. A facsimileof this

firstPeterseditionappearsas FigureI.s
"Susanna"was an instant success. It was picked up by several
more minstrelcompanies,and it spreadacrossthe United Statesand
overseas.Alreadyin I849BayardTaylor,on his way to the gold fields
in California,reported that a boatmanwho took him across the
Isthmus of Panamapaddledto the tune of "Oh! Susanna,"which,
Taylor wrote, "he sangto a most ludicrousimitationof the words."6
Between 1848 and I850, "Susanna" was printed and reprinted,
arrangedand rearrangedby at least sixteen different publishers.
AppendixA lists early printsof "Susanna"-seventeenfor voice and
piano, four for voice and guitar,nine for piano solo.7Althoughthey
were printedwithin a two-yearperiod, the editions differmarkedly
from one another,giving differentversionsof the text, the tune, the
harmonies,and the accompaniment.They advertisethe song under
several titles and attributeit to variouscomposers,arrangers,and
performers.Yet all are clearly versionsof the same song. Figure 2
collatesseven of these early "Susanna"prints, showingwords, tune,
and salientharmoniesfor the first verse and the chorus.s
The differences between versions are substantial. In the first
measure,for example,the Petersversion(top line) has an eighth-note
pickup,the otherversionsa pairof sixteenths.The cadentialformulas
in measures7-8, 15-16, and 23-24 differmarkedlyfrom edition to
edition. Peters approachesthe tonic from scale-step^; Holt, Willig,
and Atwill from i; Ferrett, Millet, and Mitchell from . At the
s Figure i reproduces the earlier of at least two versions of the first Peters edition.
Peters seems to have had a new set of plates engraved, and errors were introduced-
for example, the initial alla breve time signature lacks a slash in the later version, and
"night" in the first line of the third verse is misspelled as "nigh." Both versions bear
the plate number 1228.
Bayard Taylor, Eldorado,or Adventuresin thePath of Empire,vol. I (New York:
Putnam, 1850), 13. By 1853 Taylor could report hearing "Susanna" sung by a
"Hindoo minstrel" in Delhi (A Visit to India, China, and
Japan in the Year 1853 [New
York: Putnam, i855], See also William Austin, "Susanna,""Jeanie,"and "TheOld
155).
Folksat Home":TheSongsof StephenC. Fosterfrom His Time to Ours, 2d ed. (Urbana:
University of Illinois Press, 1987), 28.
7 The prints in Appendix A are in the Foster Hall Collection,
Pittsburgh; the
Lester Levy Collection in the Milton S. Eisenhower
Library at The Johns Hopkins
University, Baltimore; and the collections of the Library of Congress and the British
Library.
8 Figure 2
omits piano accompaniments and does not give the words after the first
verse and chorus. To make comparisons easier, all versions are
given in the key of G.
Rhythmic errors in the original prints have been retained. The Peters version, which
was notated in the original in cut time, has been rewritten in
Figure 2 as 2/4.

92 JOURNALOF THE AMERICANMUSICOLOGICAL
SOCIETY
a ug y TiNNt HA.
MN of tie Sable Written and Cornposed by s.V.Foster.

as suiog by eTi.ibtor Harmottist.
L
0
-inati1
1/
.L"W.-;, , .,
I NON0 - -No
Sir
I romle frlom
A4l
.a -bapa with my Banljo on my knee~ r'
J d
ed r I of
AosaAf ter ith my Ba rjo mtc kn ee_ Se
gwine to Lo -
..a -
My true for to ee. It rain'd all night de day I
.. .1j.
. . .. .. ..W
I_ l.ft,
gwine to LoOsi n true I t s It rad all nig de day Ileft,
'De
T I
C• My- 1 i fr,, v.d- iA ' "I
- ?
i tt 1
4"
- --L
wedder
wldder it
it was dry;
was dry; The
The sun
sun so
so but
hot II froze
froze to
to def-
deL" Su__sannadont
dust you
you cry.
cry.
S"s"nna_
I f
_ ILI
)ner
avrL At
u?e_ o .-a-- -- -11-
h
er147b lieo ititCt~ rK,
'(.~trIilteCeks
Figure i. The Peterseditionof "Oh!Susanna"(1848).FosterHall Collectionof the

Universityof Pittsburgh.
StephenFosterMemorial,
beginningof thechorus(m. 17)therearethreedistinctversionsof the

tune, one beginningon i, two on ^4.
Thewords,too,varyconsiderably. In someversionsthetextbegins
"Icome"(Peters),in others"Icame"(Holt,Willig),in stillothers"Ise

"OH! SUSANNA" 93
4
Cho rus.
Soprano
and At U
Alto.
Oh! do not cry for mei
sanu....na,
u...._
Tenore.
Oh! Su- san do not cry for me;

... ...nma
Basso. I
Oh! Su- do not cry for mei

....san....na,
Piano.
*r I
come from Wid my on my knee.

a__.ba_ Ban._jo
AlI....... _ma
come from Al -_ba _ ma, Wid my on my knee.

..__a.. Ban._jo
,.8va..4.,
•",,r .I a."rJ.:-A .. . . .... ... _'

•
-- r__•
•
•-•
. ...
r_.
.. '1•
-h •
'1• [! -• -loom • .. - . . ._ ...
Figure i (continued)
come" (Millet, Ferrett). Most versions give "It rain'dall night the day I
left" (m. 9), but Millet's edition reads "it raind all day the night I left."
Differences in dialect spellings abound: "gwine," "gwan," "gwyne";
"wedder," "weather";"froze," "frose." Harmonizations also differ be-

SOCIETY
I jumsp'd aboard the tel e.graph and down de ribber, De

trabl•ed
Piano.
lectrick tlu id magnified and illd five hundredNigga. De bul ine bust and de hoss ran off,I
Chorus.
reallythoughtI'ddiei I shut Iyeyesto holdmybref u saa dot youcry.

Chorus.
_-I am N I f. 01 , J -1
o
Ak
--i
I had a dreama de udder night, when ebry ting was still;
I thought I saw Susanna dear, a cominng down de bill,
De buckweat cake was in her mouf, de tear was in her eye,

I says, Ise coming fromn de souf*, -Susanna dont you cry.
Oh! Susanna &c.
1i29 . susanna. 4.
Figure I(continued)
tween versions. In measure 14, for example, the Peters edition harmo-
nizes "def/deff/death"with a submediantchord, otherswith a diminished
seventh chord, still others with the tonic triad.
How did the differences between the versions in Figure 2 arise,
and what do they signify? Which version, if any, represents

Animato V4/FV
Peters V
o rig.
I come from Al- "
a- ba ma with my Ban- jo on my
(IFF.r
Allegretto V4/V
Holt
IV r
I came from Al- a-f'" 'Errma wid
ba- my ban jo on my
Allegretto V4/V
Willig
I came from Al- a- ba- ma, Wid my .r jo
ban- on my
I
A tw ill
A
orig.
" -
-,R
I cum from Al- a- bama, Wid my banjo on my
AllegrettoA I
Ferrett
Ise come from Al- a- ba- ma wid my banjo on my
Millet
Ise come from Al- a- ba- ma wid my ban jo on my

Allegretto V4/V
Mitchell
I came from Al- a- ba- ma wid my Ban- jo on my
5
V I I
Peters N
knee -
I'se gwme to Lou' si- a- na My true lub for to
V IV 16
knee, I'm gwan to Lou- si- an- na, My true love for to
V I I
Holtknee, I'm Igwan to Lou-IV*1
si- a- na, My I i
true love for to
V 16
I
Atwill /
knee, I'm gwin to Lou- si- anna, Dat my true lub I may
V I I
knee,
knee, I'm
Ise gwan to
gwan to Lou-
Lou- si-
si- a-
a- My
na, My true love
true love for
for to
to
na,
v TE
F.6
Mitchelli
Figure 2. Collation of "Oh! Susanna" versions

Peters N. IP,
-vr - ; de day I left, De it was
see. It rain'd all night wedder
Holt1t" . -
I I , r r
see, It raind all night the day I left The weather it was
V4/V
WilligO1OI/! O
see; It rain'd all night the day I left, the weather it was
I
Atwill
ff -
see; It rain'd all night de day I leff, De wedder it was
I
Ferrett0O
see; It rain'd all night the day I left, De wedder it was

I
Millet
see It rain'd all day the night I left De wedder it was

r v
Mitchell .
see. It rain'd all night de day I left, De wedder it was
v vi I6
Peters,. dry; The sun so hot I froze to def - Su- sanna, dont you
V viio7/V I1
"" -A I -"
Holt- dry, The sun so hot frose to -
death Su- sanna dont you
V vii7/V 16
Wiltig
dry, The sun so hot I froze to death, Su- san-na, don't you
I
VI
F9 1El
Atwill
,
dry, De sun so hot I frose to deff, Su- sanna dont you
v I4
De u s h h,
det
Ferrettdry. II•l
i".
dry De sun so hot I froze to death, Su- san-na, don't you
V I I
Millet
dry De sun so hot I froze to death Sus- anna dont you
Figure 2 (continued)

Peters 0 i I
.
cry. Oh! Su- sanna, do not cry for me; I
IV
Holt !tt1I1
cry. Oh! Su- sanna Oh! dont you cry for me I've
IV
Willig Pt
cry. Oh! Su- sanna Oh! don't you cry for me, I've
IV
Atwill _ i
cry. Oh, Su- san na Dont you cry for me, I'm
14 IVL
Ferrett . 11
cry. Oh Su- sanna dont you cry for me Ise
IV
Mil et
S IsoAt I
cry. Oh Sus- anna dont you cry for me Ise
IV
Mitchell I a R *
cry. Oh! Su- " na
san- dont you cry for me, I've
lvi I
Peters n•
comefrom Al- a- ba-ma, Wid my Ban- jo on my knee.
I
Holt
comefrom Al- a- ba- ma wid mi banjo on my knee.
Willig X-1,I
come from A- la- ba- ma Wid my banjo on my knee.

I
Atwill
?r F w r
cum from Al- a- bama, Wid my banjo on my knee.
I
Ferrett & L i
- come
r
R.. ?Al-
from a- ba- ma wid
9...
de banjo on my knee.
Millet
Come from Al- a- ba- ma wid de banjo on my knee.
Mitchell
come from Al- a- ba- ma Wid my banjo on my knee.


SOCIETY
"Susanna"as Foster wrote it? These questionscan be addressedin

two ways:(i) link one or moreof the printsto the composer-that is,
authenticateone or more versions of "Susanna";(2) construct a
stemma-a family tree of "Susanna"sources, showing the lineageof
derivationfrom one sourceto another.
A study of the sourcesof "Susanna"was undertakenby RobertW.
Gordon, the founderof the Archiveof AmericanFolk Song, in an
essay entitled "PreliminaryReport on Stephen C. Foster's 'Oh!
Susanna.'" Gordon'spaper, written in 1931, was never published,
but it is preservedin typescript in the Foster Hall Collection in
Pittsburghand at the Libraryof Congress.9Gordon made progress
both toward authenticationand toward constructinga stemma of
"Susanna."He locatedearlyeditionsof "Susanna"andcheckedcourt
recordsfor the dates on which these editionshad been depositedfor
copyright.Basedon the copyrightdates,Gordonestablishedthe Holt
print as the first edition of "Susanna."Gordonperceived,however,
that the Peters print must be the authenticedition, on the basis of a
letter dated 25 May I849 from Stephen Foster to the New York
publisherWilliamMillet in which Fosterstated:"I gave manuscript
copies of each of the songs 'Lou'sianaBelle'--'Uncle Ned'-& 'Oh,
Susanna'to several persons before I gave them to Mr. Peters for
publication. . . . Mr. Peters has my receipt for each of the songs."'0
The letter not only shows that Peters received "Susanna"directly
from Foster; it also suggests that several authenticmanuscriptsof
"Susanna"circulated during I847 and 1848, presumably among
minstrel-showperformers.An announcementby Peters in the Lou-
isvilleDailyDemocrat of 6 September1848advertising"SusannaDon't
You Cry" as "recentlypublished"suggestedto Gordon that Peters
the
9 Gordon undertook the report as a commission from Josiah Kirby Lilly,
Indiana industrialist and collector of Foster memorabilia. See Debora Kodish, Good
Friendsand Bad Enemies:RobertWinslow Gordonand the Study of AmericanFolksong
(Urbana: University of Illinois Press, 1986), I86. John Tasker Howard evidently had
a copy of Gordon's paper, and he relied on it heavily for his discussion of "Susanna"
in his Foster biography (America'sTroubadour,I41ff.).
So Harold Vincent Milligan, Stephen Collins Foster: A Biographyof America's
evidence
Folk-SongComposer(New York: Schirmer, I920), 45-46. Another piece of
that Peters received "Susanna" directly from Foster is Morrison Foster's complaint
that Stephen had "made a present" of "Old Uncle Ned" and "Oh, Susanna"to Peters
and that Peters had "made ten thousand dollars out of them" (Morrison Foster, My
BrotherStephen[1896; privately reprinted in Indianapolis, 1932], 35). Robert P. Nevin,
in a memoir based on personal acquaintance with Foster, reports that Peters paid
Foster one hundred dollars for "Susanna"("Stephen C. Foster and Negro Minstrel-
sy," Atlantic Monthly20 [1867]: 614)-

"OH! SUSANNA" 99
might have issued his edition as early as July or August, even though
he did not deposit it for copyright until December.I"
Gordon was also able to deduce some relationships of the early
prints to one another. Analyzing details of typography and layout as
well as more general similarities and differences in words and tunes,
Gordon established, for example, that of the two editions by Ditson,
the one credited to G. N. Christy was derived from the Holt edition,
while Ditson's "Ethiopian Serenaders" edition was related to the
Benteen print. The Ferrett and Millet editions, he noted, were related
to one another; Willig's first edition and Benteen formed another
doublet. The second Willig edition, on the other hand, was related to
the Peters edition.
Extending Gordon's work in their critical edition of Foster's
works, Steven Saunders and Deane Root divide "Susanna" sources
into four "traditions":(i) a "Christy Minstrel" tradition, consisting of
Holt's first edition plus related prints by Hall, Howe, and Ditson; (2)
a "Sable Harmonist" tradition, including prints by Peters and Willig;
(3) a "Baltimore" tradition (Benteen and Willig); and (4) a "mixed"
tradition (Millet and Ferrett).I" Although they arrange sources in
groups, Saunders and Root do not attempt to construct a stemma for
"Susanna," stating in their introduction that for early Foster songs
"the complexity of the source situation . .. and the sizable differences
between various arrangements [make] it impossible to establish com-
plete stemmata."'3
Given the character of the differences between "Susanna" ver-
sions, it is indeed difficult to establish a stemma for "Susanna."How,
for example, can we account for the readings in measure 5? Why
would a copyist or an engraver working from a manuscript that gave
the reading in Example ia write out the version shown in Example
i b? What about cases where the words are identical or
nearly identical
but the tunes vary considerably? Example 2 shows divergence of this
sort between the Peters, Holt, and Ferrett editions. So long as the
underlying model is that of written transmission-that is, copying
"
" Gordon, "Preliminary Report," 12.
Steven Saunders and Deane L. Root, The Music of StephenC. Foster, vol. i,
I844--I855 (Washington, D.C.: Smithsonian Institution Press, 1990), 461. Saun-
ders and Root's designation of the Peters/Willig prints as a "Sable Harmonist"
tradition is confusing, because sources in the Millet/Ferrett
group (Saunders and
Root's "mixed" tradition) also bear the designation "As sung
by the Sable Harmon-
ists.
'3 Saunders and Root, TheMusicof Fosteri:xxii. Saunders attempted a provisional
stemma of "Susanna," which he graciously shared with me. It is similar in its broad
outline to the stemma I present in Figure 3 below.

I00 JOURNALOF THE AMERICANMUSICOLOGICAL
SOCIETY
Examplei
"Susanna,"mm. 5-6
(a) Petersedition
I'se gwine to Lou' si- a- na
(b) Holt edition
I'm gwan to Lou- si- an- na,
Example2
"Susanna,"mm. 7-8, Peters,Holt, and Ferretteditions
Peters I
.
My true lub for to see.
Holt I
My true love for to see,
Ferrett I
.vtru ov
My true love for to see;
written-out words and tune from a manuscript or print into another

manuscript or print-then there is no way to explain this sort of
differential variation. No number of hypothetical intermediaries can
explain why the words should be transmitted faithfully while the tune
is changed arbitrarily.
The musicological panacea in cases like this is to invoke oral
transmission. Instead of a scribe copying from one written source into
another, the model proposes a singer (or an instrumentalist) and a
listener who sings (or plays) the song at a later time. With oral
transmission, extensive and seemingly arbitrarychanges in the words
or the tune become more plausible. Variants are no longer seen as
errors but as alternate readings; the notion of a stemma of filiated

"OH! SUSANNA" IOI
sourcesis replacedby the looserconceptof tunefamilies.14 Gordon,

in his studyof "Susanna," acknowledges the possibilityof "changes
introduced fromsinging."I5His model,however,clearlyremainsone
of transmission fromone printedsourceto another.Saundersand
Rootmovefurthertowardanoralmodel,suggestingthatmostof the
earlyeditionsof "Susanna" were"filtered
throughminstrelperform-
ers beforethey reacheda publisher.""6 The hypothesisthat differ-
ences betweenearly versionsof "Susanna" are the resultof oral
transmission is reasonable.
"Susanna" was widelyperformed on the
minstrelstagebeforeit was printed.A letterto FosterfromFirth,
Pondand Co., his principalpublisherfrom 1850on, impliesthat
publishersoften receivedsongs not fromthe composerbut from
performers:
Fromyouracquaintance with the proprietorsor managersof the different
bandsof "minstrels,"& fromyourknownreputation,you can undoubt-
edly arrangewith themto sing [yoursongs]& thus introducethemto the
public in that way, but in orderto securethe copyrightexclusivelyfor
our house, it is safe to hand such personsprintedcopies only, of the
pieces, for if manuscriptcopiesareissuedparticularlyby the author,the
marketwill be floodedwith spuriousissues in a shorttime.17
The hypothesis that differences between versions of "Susanna"are

the result of oral transmission cannot explain, however, why the
words to the song in the early editions are so much more similar to one
another than the tunes are. In verse i and the chorus, only "Dat my
true lub I may see" in Atwill and "rain'dall day the night I left" in
Millet present significant deviations from Foster's original text. Other
differences are merely in spellings of dialect, and in many cases even
the spellings are maintained from version to version. In
subsequent
verses the differences in words are similarly modest. Peters lacks a
verse that most others have; Atwill adds a verse missing in the others.
If "Susanna" were being transmitted orally, from performer to
'4 For a recent example of tune family analysis, see John M. Ward, "The Buffons
Family of Tune Families: Variations on a Theme of Otto Gombosi's," in Themesand
Variations:Writings on Music in Honor of Rulan Chao Pian (Harvard
University:
Department of Music, forthcoming).
5sGordon, "Preliminary Report," 2.
x6 Saunders and Root, TheMusic Foster
of 1:461.
'7 Quoted in Howard, America'sTroubadour,I52. It is striking that the letter
recognizes the danger of unauthorized written transmission but not of oral transmis-
sion. Neither this letter nor Foster's letter to Millet seems to the danger
of a performer learning "Susanna"from a performance and acknowledge
taking that version to a
publisher.

SOCIETY
performer,onewouldexpectthewordsto be alteredandvariedalong
withthetunes,as is oftenthecasewiththetransmissionof American
balladsandblues.The consistencies fromversionto versionseemto
pointto rather
written, than oral,transmission.
of MixedTransmission
TheStemmatics
This seeming contradiction disappears when it is recognized that

oral transmission and written transmission are not mutually exclusive.
In some cases a song may be transmitted both orally and in writing,
a phenomenon that might be called mixedtransmission.The transmis-
sion of a song can be "mixed"in two senses: (i) the song's stemma may
involve some sources that are entirely oral, others that are entirely
written; and (2) a given source may be partly oral, partly written.'8
The possibility of mixed transmission has been recognized by
scholars of medieval chant. Leo Treitler, for example, has argued in
several studies that the extant sources of chant often represent the
residue of both oral and written transmission, overlapping and
interacting in complex ways.'9 "Reading, remembering, and extem-
porizing," according to Treitler, "were continuous acts; they were
mutually supportive and interdependent."2' Any attempt to construct
a stemma that shows the relation between oral and written transmis-
sion of a given chant is additionally problematic, because in most cases
an original version of the chant cannot be identified."'
In nineteenth-century America, oral and written transmission
sometimes overlapped in ways that were similar to Treitler's model
for medieval chant. Ballad singers, dance fiddlers, and minstrel-stage
performers orally transmitted popular songs and tunes alongside of
and interacting with written transmission. Indeed Treitler cites Le
Disir: Favorite Waltz by Beethoven,published in Boston in 1842, as an
example of a piece that changed considerably as it moved from written
's The term mixedtransmissionwas coined by Charles Seeger in his article "Oral
Tradition in Music" in Funk and Wagnall'sStandardDictionaryof Folklore,ed. Maria
Leach (New York: Funk and Wagnall, 1950), 825-29. Seeger employs the term in the
first of the senses discussed here but not the second.
'9 Leo Treitler, "Oral, Written, and Literate Process in the Transmission of
Medieval Music," Speculum56 (198 ): 471-9'; "Transmission and the Study of Music
History," InternationalMusicologicalSociety--Reportof the Twelfth Congress,Berkeley,
1977 (Kassel: BArenreiter, 1981), 202- i; "The 'Unwritten' and 'Written Transmis-
sion' of Medieval Chant and the Start-up of Musical Notation," Journal of Musicology
io (1992): I31-91." "
20
Treitler, 'Unwritten' and 'Written Transmission,' i35-
2
Treitler, "Transmission and the Study of Music History," 208.

"OH!SUSANNA" 103
to oral and back to written transmission." But the enterprise of

unraveling oral and written transmissionis more feasible for the nine-
teenth-century repertories than it is for medieval chant, because the
dominant mode of transmissionin the nineteenth century was not the
manuscriptbut printedmusic. Printsdocumentmany versionsof a song,
and in the case of a song as popular as "Susanna,"they document the
versions almost instantaneously. And compared to medieval manu-
scripts, nineteenth-century American (and English) prints have been
extremely well preserved. Thus, there is some hope of constructing a
stemma for "Susanna."It would be desirablein additionfor that stemma
to show where transmissionwas oral and where it was written.
To construct a stemma when mixed transmission is suspected, the
piece in question can be thought of as a bundle of "aspects." Aspects
are conceptually distinguishable features of a piece that may be altered
or changed independently and that may be transmitted together or in
various combinations.23 When the transmission of a piece is mixed,
then some sources probably transmit all aspects in writing; others may
transmit all aspects orally; still others may transmit some aspects in
writing and other aspects orally. In the case of a song like "Susanna,"
we can identify the following aspects:
a. words
b. tune
c. key
d. meter
e. partwritingin the chorus
f. harmonization
g. accompaniment
h. pianointroductionand coda24
22
Treitler, "Oral, Written, and Literate Process," 47 1. LeDhsir,Treitler notes, is
actually the second of Schubert's opus 9 waltzes. Treitler argues that some of the
differences between versions of Le Disir result from copying errors, others from
editorial emendations, still others from remembering and recomposition.
23
Although aspects may be altered independently, a change in one aspect often
entails a change in another. For example, if the words to a song are
changed, the tune
is often adjusted to fit-and vice versa.
24 "Words" should be understood to include
spelling and punctuation as well as
the words as grammatical entities. "Harmonization"refers to the chords and their
inversions, regardless of how they are realized in the piano or vocal parts. "Accom-
paniment" refers to the details of the piano part. This list of aspects could be
expanded as necessary. For "Susanna"it might also include the fingerings in the piano
part. In other repertories it could be extended to include instrumentation, ornamen-
tation, dynamics, tempo, and so on.
The attribution of a piece to a composer, performer,
arranger, and/or lyricist can

SOCIETY
Aspectsat the headof thelist aremostlikelyto be transmitted in

writing,aspectsat the bottom least likely.For example, if a source
transmitsthe tune in writing,then the accompaniments may be
transmittedeitherorallyorin writing.If thetuneis transmitted orally
in the source,however,then the accompaniments will almostcer-
tainlybe transmittedorallytoo, sincewrittenaccompaniments areof
little use withouta writtenmelody. If both tune and wordsareorally
transmittedin a source,then harmonieswill be oralas well. Thus the
order of aspectsin the list representsa logical or at least a practical
hierarchyof written versusoral aspectsin mixed sources.2s
In the manuscriptthat Foster sent to Peters for publication,all
aspectsof "Susanna"were transmittedin writing.In the performance
that BayardTaylor heardon his way to California,all aspectswere
transmittedorally.It is easyto imagineintermediatescenariosin which
some aspectsare transmittedin writing, others orally. One minstrel
performermightlearnthe songfromanother(allaspectsoral),jotdown
the wordsso as not to forgetthem (wordswritten,otheraspectsoral),
pass the song on to anotherperformerby writing out words and
symbolsfor the chords(words,harmonization, and key written,other
aspectsoral), then takethe song to a who
publisher hiresan arrangerto
add a piano accompaniment(all aspectswritten). Processes like these
couldin principleaccountforthe thirty"Susanna" sourcesandall their
variantreadings.Clearly,it is impossibleto reconstructsuch detailed
scenarioson the basisof extantprintedsources.It is possible,though,
to constructa stemmathat shows the filiationof extantsources,posits
hypotheticalsources, and characterizesthese hypotheticalsourcesas
written,oral, or mixedin variouscombinations.
TheStemmaof "Susanna"
The best place to begin is with the words. Differencesbetween

words in the various versions of "Susanna"are not as strikingas
alsobe consideredas an aspect.Fora discussionof the stemmaticsof attribution,see

John Spitzer, "Authorshipand Attributionin WesternArt Music"(Ph.D. diss.,
CornellUniversity, 1983).
ratherthan oral,
25 It is preferableto think of oral, written, or mixed sources
Sourcesare consideredhere both as vehiclesof and,
written, or mixed transmissions.
in the case of extant sources, as evidence of transmission.Instead of imagining
hypotheticaltransmissionssuch as a singerlearningby ear or an engraverwriting
down what a singer dictates,it is more practicalto think of hypotheticalsources:
manuscriptsor performancesor mixturesof the two. In this way oral and written
aspectscan be stipulatedunequivocallywith a minimumof speculation.

"OH! SUSANNA" 105
differencesin tunes, but becausethere are four verses, there is more

materialto work with. In addition,the words are more likely than
otheraspectsto be transmittedin writing. 6 To establisha filiationof
sources,I have identifieddistinctiveword-readings.27In some casesa
distinctive reading is clearly an error;in other cases readings are
plausible alternativesto one another. Based on these distinctive
readingsof the words, I havepostulatedthe stemmashown in Figure
3. Extantprintsaredenotedin Figure3 by publisher'sname,keyedto
the list in AppendixA. Hypotheticalsourcesare denotedby lower-
case letters in brackets.A hypotheticalsource might have been a
manuscript(now lost); it might have been a performance;it might
have been mixed, along the lines of the sourceshypothesizedin the
previous section.28
The relationsof extant sources to one anotherin Figure 3 are
relativelystraightforward.Peters-guitar,for example,is printedfrom
Lee and Walker'splates;the publisher'sname has been erasedand
replaced,but the copyrightline still reads"Lee& Walker... Eastern
District of Penn." Both Ditson prints seem to be copies of earlier
editions by other publishers:Ditson-Christyof Holt-i, Ditson-E.S.
of Benteen.29Peters-Baltimore,as Gordondeduced,was printedby
or in cooperationwith Willig;the platenumberfits into Willig'sseries
ratherthan Peters's.3 Otherextantsources,it can be shown, are not
derived from one another but from a common source, no longer
extant. Millet-i and Ferrett, for example, each contain distinctive
readingsthatdo not turnup in the other.They must, then, havebeen
derived independentlyfrom source [f].3' Holt-i and Benteen must
26 See the list of aspects in the

previous section.
27 The term distinctivereadinghas been pressed into service as a
compromise
between "conjunctive"and "disjunctive"variants. The distinctive
word-readings on
which the stemma in Figure 3 is based are listed in Appendix B.
28
Given the excellent preservation of nineteenth-century American sheet music,
it can be assumed that none of the hypothetical sources were
prints.
29 Gordon thought that Ditson-E. S. was
printed from the same plates as Benteen,
and indeed they have the same plate number (1427). Slight differences in detail show
that these are not the same plates, however. It is possible that Benteen had another set
of plates engraved with the same number and lent these to Ditson.
3o See Gordon, "Preliminary Report," 17-18.
3' The keyboard introductions and accompaniments in the Ferrett and Millet
prints differ extensively from one another, but the prints otherwise resemble each
other so closely that it seems as though one might have been
copied directly from the
other, with the piano part altered to avoid making the piracy too obvious. Examina-
tion of details shows that this cannot have been the case, however. The distinctive
reading "Lectrie"in verse 2, which Millet shares with Holt on another branch of the
stemma, does not occur in Ferrett and cannot have been introduced indepen-

FOSTER
[a] autograph [b] autograph

all aspects written all aspects
(3 verses) (4 verses)
Peters-Louisville
[C] all aspects
Peters-Baltimore
[d] words written
other aspects oral
[e] words written [f] words written [g] al

tune written tune written
chorus parts oral chorus parts written
introduction oral introduction oral
harmoniesoral harmoniesoral
accompanimentoral accompanimentoral
Holt-i Benteen Ferrett Millet-l Atwill
Holt- Willig Johnston Millet-2
Howe Ditson-Christy Benteen-guitar Ditson-E.S.
Holt-5 Lee and Walker
Peters-guitar
Figure 3. Proposed stemma for "Oh! Susanna"

"OH! SUSANNA" IO7
likewisebe derivedfrom a commonsource, the hypothetical[e].

What was the characterof these hypotheticalsources?Which
aspectsdid they transmitorally, which in writing?To answerthese
questions,I broadenedthe examinationfromthe words to all aspects
of "Susanna"-tune, harmonies,and the rest-and comparedextant
sourcesthat accordingto the analysisof the wordswere derivedfrom
the same hypotheticalsource. If the readingsfor a given aspectwere
identical, or if the differencesin readings were interpretableas
copyingerrorsor emendations,then thataspectwas probablywritten
in the hypotheticalsource. If, on the other hand, the readingsfor an
aspectdifferedin ways thatareunlikelyto haveresultedfromerroror
emendation,then that aspect was probablyoral in the hypothetical
source.32
A comparisonof the FerrettandMilleteditionsprovidesexamples
of this reasoning.33The words in the two printsagreesubstantially;
all differencescan be attributedto erroror emendation.Source [f],
therefore, from which Ferrett and Millet are derived, probably
transmittedthe wordsto "Susanna"in writing.The tunes in the two
sourcesarenot identical,as Figure2 shows, but they areso similarin
detailsof pitch and rhythmas to suggestthat both were copied from
a writtensource.Again,differencesbetweenthemcan be explainedas
errors or editorialemendation.34The impressionthat Ferrett and
Millet are both derivedfromthe samewrittenmodelis borneout by
the partwritingin the chorus,which is identicalin all voices, except
for a pitch error in measure 23 of the top voice in Millet and a
rhythmicerrorin the same measurein Ferrett(see Fig. 2). On the
dently. Therefore Millet cannot have been copied from Ferrett. That Ferrett did not
copy from Millet is shown by the reversal of "day" and "night" in verse i, which
occurs in Millet but not in Ferrett. If Ferrett had been copying from an
exemplar that
read "Rain'd all day the night I left," he could not have corrected the
reading by
conjecture. Other distinctive readings reinforce the point that Ferrett and Millet
cannot have been copied from each other but must be derived from a third source. It
is possible that the two prints were copied from two closely related
manuscripts, but
there is no need in the stemma to multiply hypothetical sources.
32 The method
requires the exercise of a certain amount of judgment. Errors in
copying are easy to identify, but it is sometimes hard to tell whether a given difference
between two related sources represents an editor's emendation or a
change introduced
by oral transmission.
33 There are actually two Millet editions of "Susanna,"designated in Appendix A
as "Millet-i" and "Millet-2." Millet-2 corrects some errors of the earlier edition but
leaves several others unchanged. The readings in Figure 2 are from Millet-i.
34 For example, in measure Io Millet sets "left, De" with even eighths, while
Ferrett uses a dotted eighth and sixteenth. In measure
at the end of the measure in all voices-obviously an error.17 Millet adds an eighth rest

I08 JOURNALOF THE AMERICANMUSICOLOGICAL
SOCIETY
other hand, Ferrett's and Millet's piano introductions differ massively

from one another, as do the piano codas at the end of the song. The
piano accompaniments to the song itself are similar in approach but
different enough in detail that they cannot have been copied from the
same exemplar.3s Source [f], therefore, seems to have been a written-
out vocal score to "Susanna," containing the words, the tune, and a
four-part harmonization of the chorus. The piano accompaniments,
however, along with the introductions and codas, must have been
supplied by a performer or an arranger. Finally, the Millet print
transmits not one but several attributions in writing: "As sung by the
Sable Harmonists ... Music by S. C. Foster ... Arranged by W. J.
Wetmore." Ferrett, on the other hand, does not give an attribution to
composer, performer, or arranger. Source [f] may have contained a
written attribution, which was picked up by Millet and dropped by
Ferrett. It seems more likely, however, that source [f] was unattrib-
uted and that Millet received the attribution orally, perhaps from a
member of the Sable Harmonists troupe.36
I applied similar reasoning to the other hypothetical sources to
arrive at the characterizationsshown next to each source in Figure 3-
Two hypothetical sources [a] and [b] are Foster autographs, the
manuscript copies that Foster "gave to several persons" before he
delivered the song to Peters. I assume them to have been written out
in all aspects and to have resembled the Peters prints in most respects,
except that [b] may have contained a fourth verse that [a] lacked.
Hypothetical source [c] is an oral source from which all versions
except Peters are descended. The words in source [c] and its descen-
dants are separated from the Peters tradition by several distinctive
readings, among them: "don't you cry" in place of "do not cry" in the
chorus, the omission of "dear"in verse 3, and the rearrangementof "I
says I'se coming" to "says I, I'se coming," again in verse 3. It is
unlikely that these changes (as well as many more details of spelling
and punctuation) would have been introduced as editorial emenda-
tions of a written manuscript. Thus [c] probably transmitted the
words to "Susanna" orally. If the words were unwritten, then it is
likely that tune, harmonies, and accompaniments were also transmit-
ted orally. And indeed tune, harmonies, accompaniments, and meter
3s Forexample,in measure13Ferrettcontinuesthe AlbertibasswhileMilletgoes

backto chords.At the beginningof the chorusin Ferrett,the pianoplaysin unison
with the voices;in Millet the pianoplays a chordalaccompaniment.
36 Foster'sletter to Millet, quotedabove, seems to be a responseto an inquiry
from the publisheras to whetherFosterwas indeed the authorof "Susanna"and
"UncleNed" and whetheran authorizededitionhad been published.

"OH!SUSANNA" 109
also differ significantly from the Peters tradition in all sources

descended from [c]. With all aspects oral, source [c] can best be
thoughtof as a performanceor a groupof performancesby a minstrel
performeror a troupein I847 or early in I848.
Source [d], on the other hand, did transmitthe words to "Su-
sanna"in writing, as shown by a distinctive error, "Lectrie"for
"lectric,"which turns up in verse 2 in both Holt-i and Millet-i.37
However, the tunes in the Holt/Benteenversus the Millet/Ferrett
traditionsdiffer significantlyenough from one another to make it
probable that source [d] transmittedno more than the words in
writing. Tune, harmonies, accompaniments,and key were most
likely oral. Source [e] did transmita written tune along with the
words, as shown by the similaritiesin rhythmicand melodic detail
between Holt-I and Benteen. In the chorus, however, [e] seems to
have writtenout only the melody, because,althoughthe top voice is
nearly identicalin Holt-i and Benteen, there are substantialdiffer-
ences in the lower three voices. Sources[g] and [h] are entirelyoral;
both are derivedfrom [c], but they differsignificantlyboth from [d]
and from one another. Source [h] is the hypotheticalsource from
which all early English prints of "Susanna"are descended. These
Englisheditionssharedistinctiveword-readingsthatcannotbe linked
to the U.S. prints by a written source.38Thus [h] must have been
entirely oral, probably a version of "Susanna"that an American
minstreltroupetook to Englandin 1848 or 1849.
Figure 3 shows that the early transmissionof Foster's"Susanna"
was "mixed"in both of the sensesdefinedabove. The stemmaposits
some hypothetical sources that are entirely oral, others that are
entirelywritten,andseveralthatarebothoralandwritten.Indeedthe
stemmaproposesmixed sourceswith threedifferentcombinationsof
oraland writtenaspects.In source[d], the wordsare written, but all
otheraspectsareoral;in source[e], wordsand tune arewritten,other
37 Gordon noticed the erroneous "e" in the Holt and Millet prints. He believed,
however, that it meant Millet was copying from the Holt print ("Preliminary
Report," 2i).
The hypothesis that source [d] transmitted the words in
writing requires us to
assume that differences in words between the Holt/Benteen and the Millet/Ferrett
traditions were the result of editorial emendation of a written text. The
only cases in
which this seems problematic are the addition of "Oh" ("Oh! dont
you cry") in the
chorus of Holt/Benteen and the substitution of "floated"for "trabbled"in Millet/Fer-
rett verse 2.
38 Distinctive readings in the English prints include: "And I'm gwan" for "I'm
gwan" in verse i, "Susy Ann" for "Susanna"in verses 3 and 4, and "to de South" for
"from the South" in verse 3. See Appendix B.

IIO JOURNALOF THE AMERICANMUSICOLOGICALSOCIETY
aspectsareoral;in source[f], words,tune, and the partsin the chorus

are written, while the piano accompanimentand the harmonies
(expectfor the chorus)are oral.
How plausibleare the mixturesof oral and written transmission
shown in Figure 3? Have sourcesbeen preservedfrom the world of
nineteenth-centuryAmericanminstrelsythat resemblethe sources
hypothesized in the "Susanna"stemma? Several Stephen Foster
autographshave survived,of which three offersome insightinto the
characterof sources[a]and [b].39The firstis a holographmanuscript
of "Lou'sianaBelle," one of Foster's early songs and one that he
specificallyacknowledgedin his letter to Millet having "given to
several persons"before he sent it to Peters for publication.4' The
manuscriptdiffersfromthe Petersprintof "Lou'sianaBelle"in several
respects.It is writtenin commontime insteadof 2/4. It lacksa piano
introduction, although it does have a "symphony"to be played
betweenthe verses.4'Manuscriptand printoccasionallygive slightly
differentnotes, both in the vocalpartsandin the accompaniment,and
there are many differencesin the treatmentof dotted rhythms. The
words to the song are substantiallythe same, but there are small
differences in diction and many differences in dialect spellings.
Curiously,the manuscript"Lou'sianaBelle,"althoughit is in Foster's
hand, is not attributedto Stephen Foster. The Peters print is also
anonymous. Finally, the manuscriptcontains six verses in all, the
print only four verses.42
The otherFostersourcesof interestarea pairof autographcopies
of "Oh! Boys CarryMe 'Long.''43One seems to be the manuscript
that Fostersent to Firthand Pondfor publication,the othera copy he
39 For a nearly complete list of Foster autographs, see Calvin Elliker, Stephen
CollinsFoster:A Guideto Research(New York: Garland, 1988), 8 1-82. The autographs
are described in further detail in the Critical Reports to Saunders and Root, TheMusic
of Foster.
40 The first page of the "Lou'siana
Belle" autograph is reproduced in Saunders
and Root, TheMusicofFoster 1:457. Saunders and Root believe that the manuscript of
"LouisianaBelle" does not represent a copy that Foster gave to a minstrel performer
but a presentation copy for a friend (TheMusicof Foster 1:455).
4, Both the manuscript and the print contain this "symphony." The symphony in
the manuscript, however, is scored for an ensemble of first and second violins, flute,
guitar, and piano. The print's symphony is for piano only.
42 It is
striking that of the two verses missing from the print, one contains the only
lines that might be construed as a criticism of slavery: "My masa took my lub one day,
/ He put her up to sell. / I thought I'd pine my life away, / For de Lousiana belle."
43 Both manuscripts
are in the Huntington Library in San Marino, California.
They are described in Saunders and Root, TheMusicof Foster 1:469. I thank Deane
Root for photocopies.

"OH! SUSANNA" III
sent to the Christy troupe for performance.44 In both manuscripts

Foster writes the song out in exhaustive detail, with the piano
accompanimentfully notatedandthe voiceswrittenon separatestaves
in the chorus. Words, notes, spelling, and punctuationare nearly
identicalin the two manuscripts,andthey correspondto the readings
in the FirthandPondedition.Bothcopiestransmitpianointroduction
and coda;both are attributedto "StephenC. Foster."
The "Lou'sianaBelle"and "Oh! Boys"manuscriptsimply some-
what differentviews about the characterof hypotheticalsources[a]
and [b]. The "Oh! Boys" manuscriptsseem to show that Foster
notated a song carefully and in all aspects, not only when he
submittedit to a publisherbut alsowhen he gaveit to a performerlike
Christy.They seemto imply as well that Fostermaintaineddetailsof
spellingandnotationfromone manuscriptto the next andthus that[a]
and [b] ought to resembleone anotherclosely. The "Lou'sianaBelle"
manuscript,on the other hand, which was writtencloserto the time
of "Susanna,"suggeststhat [a]and [b] mightdiffersubstantiallyfrom
one another. The absenceof the piano introduction,of two entire
verses, and of the attributionto Fosterin "Lou'sianaBelle"raisesthe
possibility that majorfeaturesof "Susanna"--likethe attributionor
the fourthverse--could be presentin [b] but lackingin [a].
Manuscriptswith characteristicssimilar to those proposed for
otherhypotheticalsourcesin the stemmahavenot survivedfor Foster,
but they do exist for the songs of Dan Emmett,a minstrelperformer
best knownas the composerof "Dixie."45s The Emmettsources-most
44
In a letter to E. P. Christy dated i2 June 185 I, Foster explains that he has sent
a manuscript of "Oh! Boys" to Firth and Pond, and for a price of ten dollars he offers
to send Christy a manuscript for prepublication performance. In a second letter (20
June I85 i) Foster acknowledges receipt of ten dollars and states that he is enclosing
a manuscript of "Oh! Boys" for Christy to use. The letters are transcribed in
Howard, America'sTroubadour,
186-87.
4s Emmett's authorship has been questioned recently by Howard L. Sacks and
Judith Rose Sacks in Way Up North in Dixie: A BlackFamily'sClaim to the Confederate
Anthem(Washington, D.C., and London: Smithsonian Institution Press, I993). They
contend that "Dixie" was composed by members of the Snowden family, black
musicians who lived, like Emmett, in Mt. Vernon, Ohio. The principal evidence for
their view consists of an oral tradition among black families in the Mt. Vernon area
and an inscription on a memorial to Ben and Lew Snowden, which reads:
"They
taught 'Dixie' to Dan Emmett." In I859, when Emmett introduced "Dixie" in New
York City, Ben Snowden was nineteen and Lew eleven years old; thus it is
unlikely
that the Snowden brothers themselves wrote the song or
taught it to Emmett. The
Sackses speculate that perhaps "Dixie" came from the
repertory of Ben and Lew's
parents, Thomas and Ellen Snowden (pp. I6o-62). Way Up North in Dixie contains

SOCIETY
of themnow at the OhioHistoricalSociety-are of two types.46In

the firsttype Emmettwroteout the wordsonly, omittingbothtune
andaccompaniment. Thesemanuscripts arein a hastyhandandseem
to be copies that Emmettmadeduringhis days as a performer,
perhapsto keeptrackof his repertoryor to teachhis songsto his
fellowperformers.47 The secondtypegivesbothwordsandtunebut
no accompaniments. These manuscripts were preparedwith more
careand are probablycleancopiesthat Emmettmadelaterin life,
perhapsto sendto a publisher.An earlyexampleof thissecondtype
of Emmettmanuscript, a copyof "Dixie"datedI859,containsa violin
four
part: introductory chords in trebleclefprecedingthesongandan
eight-measure "dance" for violin the
following song.48
TheEmmettmanuscripts in thefirstgroupcorrespond to source[d]
in the "Susanna" stemma.They suggestthat a commonway for
minstrelperformers to storetheirrepertory wasto writeoutthewords
butto commit thetunes to memory.49 The Emmett manuscripts in the
secondgroupcorrespond to source[e]. These sourcessuggestthat
minstrelperformers couldanddid writeout "leadsheets"withwords
plustuneandperhapsalsoan instrumental solo. Leadsheetsdo not
seemto havebeenusedto storerepertory, buttheymight,onoccasion,
haveprovideda meansfora performer to deliversongsto a publisher
or perhapsto otherperformers. Sourceswith the character of hypo-
theticalsources[c], [g], and[h](allaspectsoral)havenot beenpreserved
fromthe nineteenthcentury-at leastnot beforethe adventof record-
ing. Their existence is intuitively certain, however. Examples of
sources with the configurationproposedfor [f] (words, tune, and
choruspartswritten)have not been located.
fascinating material on the daily life of a black family in the nineteenth century, but
it presents little evidence that members of the Snowden family rather than Dan
Emmett were the authors of "Dixie."
46 Hans Nathan, Dan Emmett and the Rise of Early Negro Minstrelsy(Norman:
University of Oklahoma Press, 1962), 290; see also Gilbert D. Schneider, "Daniel
Emmett's Negro Sermons and Hymns: An Inventory," OhioHistory85 (1976): 67-83-
47 Personal communication from Katherine Reed-Maxfield, who has examined
the Emmett manuscripts.
48 This "Dixie" autograph has been published
in facsimile as a pamphlet (Library
of Congress call number ML 96.5 .E65) and also in Nathan, Dan Emmett, 249. In a
letter accompanying the manuscript, Emmett asserted that this was the original copy
of "Dixie," written out in New York City in 1859. This claim is suspect, but Nathan
argues that the 1859 date is probably accurate. For a discussion of versions of "Dixie,"
see below.
49 The majority of printed songsters containing nineteenth-century minstrel
songs likewise give texts only, without tunes or accompaniments.

"OH!SUSANNA" I 13
Of all the sources for "Susanna," only the Peters prints represent
a written transmission of all aspects from a Foster autograph. All
other versions were filtered through a source-source [c--in which all
aspects were oral. That source [c] was entirely oral does not mean,
however, that sources descended from it are less "authentic" in all
respects than sources descended from [a]. The song's fourth verse ("I
soon will be in New Orleans," and so on) may have been written by
Foster, even though it is lacking in the Peters prints. Clearly the verse
was present in source [c], because it is transmitted in almost all prints
descended from [c]. Was the fourth verse present in [b], the Foster
manuscript from which [c] was derived? The sources do not provide
any way to decide this question; however, the omission of verses from
Peters's "Lou'siana Belle" print suggests that the fourth "Susanna"
verse may be authentic.
Other variants in [c] and the sources descended from [c] are not
authentic. In particular, the features of the tune that distinguish
sources descended from [c] from sources in the Peters tradition-the
two-note pickups, the cadences from 2 or 3 instead of , the chorus
beginning on --cannot have been present in [b] and thus cannot have
come from Foster. These features are transformations introduced
during the process of oral transmission.
But why should oral transmission have induced a change from
eighth-note pickups to sixteenths and exchanged one cadential for-
mula for another? It is tempting to think that these changes are
manifestations of broader principles, that the differences between
Foster's original and the other versions reflect general tendencies of
oral transmission.
TheTendencies
of OralTransmission
Based on a comparison of Foster's original version of "Susanna,"as
transmitted in the Peters edition, with the versions of other publish-
ers, four tendencies of oral transmission can be postulated.
The first is a tendency to alter rhythms in order to clarify the beat.
Foster's original version begins with three undifferentiated eighth
notes, and it is possible (though in retrospect not easy) to hear the
initial eighth as the downbeat rather than the pickup, as shown by
Example 3. In all the versions that passed through oral transmission,
the pickup eighth has been changed into two sixteenths, creating an
agogic accent on "come" and making the downbeat unmistakable (see
Fig. 2 above). This two-note pickup does not set two syllables of text
either in the first verse or in subsequent verses; its purpose is

114 JOURNALOF THE AMERICANMUSICOLOGICALSOCIETY
Example3
m. i, rhythmdisplaced
"Susanna,"
I come from Al- - a- ba- ma
exclusivelymetrical.The same tendencyis seen in reversein Figure

2, measure3, whereFosterwrotean agogicaccenton the secondbeat
of the measure,as shown by the Petersedition(top line). The dotted
eighth sets the text naturally,but it is metricallyproblematicbecause
it stressesa weak beat. All the orally transmittedversionseliminate
the dotted note, movingthe accentto the bar line. A few clarifythe
meterfurtherby addingemphasisto the precedingdownbeat,Atwill
with a Scotch snap, Ferrett and Millet with a dotted eighth and
sixteenth.
The second postulatedtendencyof oral transmissionis to penta-
tonicizethe melody. At all his cadences,Fosterapproachesthe tonic
^
by half step from scale-step (see Fig. 2, ^top line). The orally
transmittedversions,on the otherhand, avoid in almostevery case,
cadencing from above by whole step (Fig. 2, mm. 7-8, I5-I6, 23-24).
Sometimesthe cadenceis embellishedwith an upperneighbor,as in
the Millet and Mitchell editions. The only exceptionis the Ferrett
print:it cadencesfromabovein measures8 and I6 but retains Foster's
^ from the
leading tone at the end of the chorus.5s By eliminating
cadences, the orally transmittedversions render the verse of "Su-
sanna"entirelypentatonic.
At the beginning of the chorus, on the other hand, the orally
influencedversionssuddenlyleap up to ^, whereasFosterbegins the
choruson i andtouches42only brieflyon the way to 6 (Fig. 2, m. i7).
^
This changefromi to at the chorussuggestsa thirdtendencyof oral
transmission:the tendencyfor a salientharmonyto drawthe melody
to the rootof the chord. Foster'ssubdominantchordat the beginning
of the chorus of "Susanna"is strikingand unusual:the chorusesof
most earlyminstrelsongsbeginon the tonic. Movingthe melodyto ^4
the new harmony.Althoughearlyversionsof "Susanna"
emphasizes
cadencemeanthatthe Ferrettprintretains
so See Figure2. Does this leading-tone
Foster'soriginalcadence?But only the last cadence(m. 2i) presentsthis reading;all
previouscadencesin Ferrettcomefromabove.An independentderivationof Ferrett
froma Fosterautographis contradictedby the analysisof the wordsto the song. It
is easierto explainthe readingin measure2 as an editorialalteration.

"OH! SUSANNA" 115
exhibitseveraldifferentmelodicshapesatthebeginning of thechorus,
allversionsthatweretransmitted throughanoralintermediary begin
the choruson 4.
Thefourthpostulated tendencyof oraltransmission is to eliminate
differencesbetweenparallelpassages.In the original"Susanna," the
openingphrasetakesa slightlydifferentformeachtime it appears
(see Ex. 4). In the orallytransmittedversions,these passagesbe-
comenearlyidenticalin everyappearance, differingonlyin rhythmic
detail.Forexample,the Holt editionin the corresponding measures
givesthe readingsin Example5. Anotheradjustment forparallelism
occursin measuresI8-I9, whereseveralversionsadd a pickup,so
that the passagecorresponds to measures2-3, 6-7, and so on. To
accommodate thepickup,anextrasyllableof textis added(seeEx. 6).
Example4
"Susanna,"mm. 1, 9, and 13 collated,Petersedition(original?)
I come from Al- a- ba ma with
It rain'd all night de day I left,

13
The sun so hot I froze to def-
Example 5
"Susanna," mm. I, 9, and 13 collated, Holt edition
I came from Al- a- ba- ma wid
It raind all night the day I left

13 so hot I frose to death
The sun so hot I frose to death

I 16 JOURNALOF THE AMERICANMUSICOLOGICAL
SOCIETY
Example6
"Susanna,"mm. 3 and 19 collated,Holt edition
my ban jo on my knee,
19
Oh! dont you cry for me
The four tendencies proposed as consequences of oral transmission

can also be seen in a functionalistframeworkas aids to that transmission.
A minstrel performer who learned "Susanna"in one or two hearings
might well remember the cadences as moving to the tonic from above
ratherthan below, since the leading-tonecadencewas comparativelyrare
in the repertory of dance tunes and minstrel songs, while the cadence
^
from was common. One could also imagine that a performer, having
learned "Susanna" as Foster wrote it, might then pentatonicize it
graduallyas he maintainedit in his repertory.Altering the melody so as
to clarify the beat-the first tendency above--makes it easier for a
listener to grasp from the outset the rhythmic structureof the song. The
subdominantchord at the beginningof the chorus is a particularlysalient
feature, and a performer who learned the song orally might well
rememberthe chord more clearly than the pitch. Moving the melody to
the chord root also allows the melody and the harmony to reinforceone
another so that the harmony can be more emphaticallyconveyed by the
tune. Finally, since a song is heard and learnedfrom the first bar to the
last, it is reasonableto imagine that parallelpassages will be heard and
remembered as reiterations of previous passages. These and similar
explanations do not prove that the four postulated tendencies operate
inevitably in oral transmission,but they suggest mechanisms that make
the tendencies seem plausible.
Most oral transformationsof the tune seem to have been introduced
early in the transmission of "Susanna"-in [c], the hypothetical oral
source from which all versions except the Peters prints are derived.
Readings for [c] can be established by analyzing similaritiesbetween its
extant descendants. Source [c], it seems, already clarified the opening
^
rhythm with a two-note pickup; it eliminated from cadences; it began
the chorus on ^; and it made measures 6 and i parallelto measure 2.
In measures 20-21, on the other hand, source [c] must have retained the
original one-note pickup, because that reading turns up in hypothetical
source [f] and all sources derived from [f]. Although source [c]

"OH! SUSANNA" 117
eliminated 7 from cadences, it did not give cadences a consistent or

definitive form. Oral transformationscontinued to alter the shape of the
cadences in [d], [e], [f], and other sources derived from [c], as demon-
strated by differences between prints by Holt, Atwill, Ferrett, Millet,
and Mitchell (see Fig. 2). Melodic shape at the beginning of the chorus
also continued to change. Source [c] and its descendantsall begin on 4,
but Atwill retains the shape of Foster's original by continuing the
melodic motion up to g"(see Fig. 2). The variabilityof sources below [c]
on the stemma suggests that the four tendencies of oral transmission
continued independently and parallel to one another as the tune to
"Susanna"was transmittedthrough sources [d], [e], [f], and beyond.
OtherMinstrelSongs
The analysis of "Susanna" suggests that, as a composed and

notated song moves into oral transmission, it undergoes predictable
changes: its rhythms are altered to clarify the beat; it tends toward
pentatonicism; its melody is altered to reflect salient harmonies;
pitches and rhythms are adjusted to conform to parallel passages.
These tendencies of oral transmission constitute a set of hypotheses.
They can be tested on other songs for which both written originals
and orally transmitted versions are available.
Stephen Foster's "Uncle Ned" is shown in Figure 4, a facsimile of
the 1848 Peters edition, which once again represents Foster's original
version.5' Again Foster passed out manuscript copies to minstrel
performers before publication;52again there are many early prints,
whose discrepancies suggest oral and mixed transmission.53When the
s' Figure4 reproducesone of at least two versionsof the Petersedition, again

printedfromdifferentplates.The laterversionhas the platenumber2159-
s5 See Foster'sletterto Milletquotedabovein the text at note lo.
s3 I examinedthe followingeditionsof "UncleNed": "UncleNed, Writtenand
Composedby S. C. FosterEsq"(Louisville:W. C. PetersandCo., 1848),plate 122o;
"OldUncle Ned, Written&ComposedForWm. RoarkOf the SableHarmonists.By
S. C. Fosterof Cincinnati"(New York:W. E. Millet, 1848);"OldUncle Ned-An
EthiopianMelody, Arrangedwith Symphoniesand Accompaniments for the Piano
Forteby R. O. Wilson"(Baltimore:GeorgeWilligJr., n.d.), plate2159; "OldUncle
Ned-An EthiopianMelody, Arrangedwith Symphoniesand Accompaniments for
the Voice and Pianoby R. O. Wilson"(Baltimore:F. D. Benteen,1848),plate
I391;
"MyUnkle Ned, Rewritten& Arrangedby WilliamClifton"(New York:JosephF.
Atwill, 1848);"Old Uncle Ned" (coverplate:"JimCrowJubilee--A Collectionof
Negro Melodiesas sungby A. F. Winnemore&his bandof Serenaders, for
the PianoForteby AugustusClapp")(Boston:Geo. P. Reed, 1847);"MyArranged
UnkleNed"
(Boston:Elias Howe, I849 ["Bypermissionof W. C. Peters,proprietor"]); "Uncle
Ned" (pianoonly) (Baltimore:W. C. Peters,
I851).

I 18 OFTHEAMERICAN
JOURNAL MUSICOLOGICAL
SOCIETY
I'N U l, I. N EI).
- ti- a i l o,,r d b,y S . r e

o-l tt hl Ul•
Mw
l lU
' ,t r ':Y
.•. .
z ".......
":7"....
"1 * ?O??+ ?
?
--.---_-
--.
t. -••-
. .
-
-
. 1? 11?
.
drr
Vi..~
B.
-----------
Ut.t 1 .
Nr FHes U
I t
" ti Ot(. g.atf
D
,. n batal .t
r o, IieD -our
4L -*-L O P
12240 IfI
t ifthv
toActCongro*,
tticordin
Ent#trwd VP1427 Inb*
Cerk*
Ofice
W.C.PterafCugrt
thDitric
oKy
Figure4. The Peterseditionof "UncleNed" (1848). LesterS. Levy Collectionof

Sheet Music, MiltonS. EisenhowerLibrary,The Johns HopkinsUniversity.
versionsof "UncleNed" are comparedto one another,all four

showup strongly.Example7 shows
tendenciesof oraltransmission
how, in two Foster's
versions, pickupis subdi-
originalquarter-note

"OH! SUSANNA" I 19
I * t ~. 1
?4 No A. t I, a aa&Itv c it be"myt , a. alstv to.
CIUOaeL. e' atketn ba a• ! ibould aIs
vi
pV&
If*
So.pr14n
ur All* j:M 9
t 'ia w: tiwr
IIAt,g tie
tidd' a dr8
U.ill't" letd
wor
.
! ttg "'• ftddir' bd de bow- Not aors- litl-of w itfor
PtJ 4
i7. 4 -
1 ...
..........ii.
.-
I9
s .
.ts
"IO j.•~.
~~NLA . ?
tollr Old Ned WN e whar do pond

iW* Nigga.
NI 0 Micre hurd work ir

pfor Ol Ned-Rels usw war& god Niges W
1Irt N Ltk: haid p

.o
,d .
to : ..•e- ild - --I~ ... -r . .. ..
? ? 1 -= ? i " f --4
u-
NIf w_---I.
.? .A LJ_--_ __L O,' _. •
---
lie" tuo hardworkror poor Old NdHt d
-. nri
f-A .
rcA
IOW hardLI 4~ls -
4F (?lw'w""~~
_ .. .A .
lg Z2
vided in orderto clarifythe downbeatat the beginningof the song.54

Eventhe Peters-pianoversion,publishedin i85i, reflectsthis change,
although Peters's arrangerpresumably had access to the Foster
54ThePeters-piano as4/4inthisand
in 2/4,hasbeenrewritten
version,originally
thefollowingexamples.

SOCIETY
: : ::':? ???' : - :
. . . ......B
.. . ?. ?. " r
'"
.•:'..??: .:~ ?., . .. .. : . . . . . .. .. . . .
It, i'
.
.V
ox
T
- ol
m I ' .. . . . . ..
.:nd t
'4 J-.: .;ah]
'04...-.
,.
sfm,.d:w-
,.
..i:: , . .at.VcV4,,.
. ,:: •::•
. :b...in. ,,L-;;:
:. -. -'
1A
.:~c ':""
. ...l -r~j
.. .Is ..d . L..
i:io
r.. ?
7. ,
.- . ".' .'.
k?-
?n. .. :t
o i
?:,,•_•-:.. " /
..... _._iL.. . .... : ;...".:::- • . , " .' ? .
.p.m....
.. .....,. ,? ' ....
l ?~~: ::~: ~:j :~h f~ aa dj~W~iut l~.k

..x..~ 4*t* .d A.
:uio " - j
"
"
.
Alw" o, , i . . .- . . .
,-" . . . .4 , ..
4..
original.Willigand Millet, on the otherhand, retainthe quarter-note

pickup.Willigchangesthe openinginterval,however,reinforcingthe
downbeatmelodicallyratherthan rhythmically.
The tendencyto pentatonicizecadencesis shownin Example8. In
"UncleNed" as in "Susanna,"Fosterapproachesthe firstfull cadence

"OH! SUSANNA" 12 I
Example7
"UncleNed," mm. 1-2, Peters,Willig, Atwill, Peters-piano,and Millet editions
Peters F V
R OI
,
Dere was an old Nig- ga, dey call'd him Un cle Ned
Willig
There was an old Niggar and his name was Un cle Ned,
Atwill
orig.G 4
p r*
Dar was an ole nigger and dey call him Unkle Ned,
orig.
Peters-piano,, -
-V
Millet
orig. Gaa
Dere was an old nigga, dey calld him Un cle Ned,
Example8
"UncleNed," mm. 7-8, Peters,Willig, Atwill, Peters-piano,and Milleteditions
Peters I W
De place whar de wool ought to grow.
Willig
In de place where de wool ought to grow.
Atwill P
orig. G
On de spot where de wool ought to grow;
Peters-piano
orig.
Millet
orig. G
De place wha de wool ought to grow.
from the leading tone. This leading-tone cadence is altered in three of

^
the versions, but each one avoids with a different formula, suggest-
ing once again that the tendencies of oral transmission worked
independently in the different versions. The leading-tone cadence is
retained in the Peters keyboard edition.

SOCIETY
The tendencyof the harmonyto draw the melody to the root of

the chordis seen in Example9. Here a cadencein the key of F pulls
the a' on the firstbeatof measurei i down to f' in the Willigversion.
Example 9
"Uncle Ned," mm. Io-12, Peters, Willig, Atwill, Peters-piano, and Millet editions
Peters a
', ,
hoe.... Hang up de fiddle and de bow:
WilligPO
hoe.... Hang up de fid die and de bow;
Atwill
orig. G
Ho o o Hang up de fiddle and de bow;
Peters-piano
orig.
Millet
orig. G
hoe...o And hang up de fid- die and de bow.
....o
The otherversionskeepthe melodyon a', althoughtwo of themmake

it easierfor the singerto findthe thirdof the chord:Atwill, by halting
the bass descentat A; Millet, by doublingthe g' in the uppervoice.
The final postulatedtendency-that oral transmissioneliminates
differencesbetween parallelpassages-turns up severaltimes in the
"Uncle Ned" versions. Example io shows how the Atwill and
Peters-piano versions add a pickup to measure 13, so that it corre-
sponds to measure i. In measure 14 the Willig and Peters-piano
versionsmove the high d"backto the secondbeatin orderto createa
parallelto measure2.ss
The analysisof "UncleNed" seems to providestrongsupportfor
the conclusion that the four tendencies postulated above are not
peculiar to "Susanna";rather, they are general tendencies of oral
transmission.They can be found in other songs from the minstrel
ss See Example7. The alterationof the cadencescan also be seen as an example

of increasingthe similarityof parallelpassages. In the originalversion, Foster
cadencesfromthe leadingtone in the verse, and fromi in the chorus.In the orally
transmittedversions,all cadencesareidentical.In this case, though,the alterationis
retrospective--anearlierpassageis changedto correspondto a laterone.

"OH! SUSANNA" 123
repertory when early prints are compared to modem versions of the

same songs. An early edition of "Blue Tail Fly"'56undermines the
rhythm of the chorus with a syncopation and ends with a cadence
from the leading tone (see Ex. I1). Many later versions make the first
two measures of the chorus parallel to one another. A few also
eliminate the leading tone and instead cadence from above, as shown
in Example 12.57 "Blue Tail Fly" is less persuasive than "Susanna"or
Example To
"Uncle Ned," mm. 13-14, Peters, Willig, Atwill, Peters-piano, and Millet editions
I *
Peters a
No more hard work for poor Old Ned - He's gone
Willig kP A
No more hard work for poor Old Ned He is gone
Atwill
orig. G Q
'I
Dar's no more hard work for my Unkle Ned, He is gone
Peters-piano
orig.f 6==Jv
•-O-t
-/
Millet rPO
orig.G
No more hard work for poor old Ned, He's gone
Example I
"Blue Tail Fly," mm. 13-i6, Benteen edition
Jim crack
r cor
rdon't care, Ole Massa gone a- way.
Jim crack corn I don't care, Ole Massa gone a- way.
56 "The Virginia Minstrels, 'Jim Crack Corn' or the Blue tail

Fly, Composed for
the Piano Forte" (Baltimore:F. D. Benteen, 1846). A yet earlier version of "Blue Tail
Fly," published in i844, may be found in Eric Lott, Love and Theft-Blackface
Minstrelsyand theAmericanWorkingClass(New York and Oxford: Oxford University
Press, 1993), I8i. Its text is similar to the Virginia Minstrels version, but its tune is
entirely different.
s7 Richard Johnston, ed., Folk SongsNorthAmericaSings(Toronto: E. C. Kerby,
1984), 304. The version of "Blue Tail Fly" in TheFiresideBookofFolk Songsclarifies the

SOCIETY
Example 12
"Blue Tail Fly," mm. I13-16,Folk SongsNorthAmericaSings
II
j~W I . ~ i
Jim-my crack corn and I don't care, My mas- ter's gone a- way.
"Uncle Ned," however, because it is not known whether this early

print represents the original version, and because a stemma cannot be
established on the basis of extant sources.
A better example is Henry Clay Work's "Kingdom Coming"
(1862), as arranged for piano by Mrs. Joe Person, a medicine-show
performer of the late nineteenth century. Chris Goertzen, in his study
of Mrs. Person and her repertory, argues that Mrs. Person's piano
arrangements, which she published in 1889, were based on oral
tradition and recollection more than on printed sources.s5 Her version
of "Kingdom Coming" shows two of the four tendencies at work. The
original "Kingdom Coming" begins, like "Susanna," with an eighth-
note pickup followed by a string of even eighths. Mrs. Person clarifies
the rhythm by subdividing the opening pickup just as the oral
versions of "Susanna" do (Ex. 13). Later in the verse Mrs. Person
Example 13
"Kingdom Coming," mm. 1-4, original versus "De Year of Jubilo," Mrs. Person's
version, based on Goertzen, "Mrs. Joe Person"
"Kingdom
Coming"P
Say dar-kies hab you seen de mas-sa, wid da muff-stashon his face?
"De Year of Jubilo"

Ig ,0101o
emphasizesa IV chord by movingthe melody to the chord root, as

shown in Example14. On the otherhand, Mrs. Person'sarrangement
adds a leadingtone to severalcadencesthat Work approachedfrom
above (Ex. i5). This leading tone, however, falls on a weak beat
rhythm similarly in the chorus but retains the leading-tone cadence (Margaret
Bradford Boni, ed., The FiresideBookof Folksongs[New York: Simon and Schuster,
'9471, 73)-
58 Chris Goertzen, "Mrs. Joe Person's Popular Airs: Early Blackface Minstrel
Tunes in Oral Tradition," Ethnomusicology35 (1991): 31-53-

"OH! SUSANNA" 12 5
Example 14
"Kingdom Coming," mm. 17-18, original versus Mrs. Person's version, based on
Goertzen
IV I
"KingdomComing" 0B
--
De massa run? ha, ha!
"DeYear
ofJubilo"
IV I
Example 15
"Kingdom Coming," mm. 5-8, original versus Mrs. Person's version, based on
Goertzen
"KingdomComing" j
k .
"
Go long de road some time dis mor-nin',like he gwine to leab de place?
"De Year of Jubilo"

*.
following ^, and it might be explained as the result of a pianist's

aversion to repeating notes.
For "Dixie," by Dan Emmett, an original version has been
transmitted in an autographmanuscript,59and a large number of early
prints transmit variant versions. The history of "Dixie" is similar to
that of "Susanna." Emmett composed "Dixie's Land-Walk 'Round"
in 1859 for Bryant's Minstrels in New York City, and the song spread
rapidly among minstrel performers. The first print seems to have been
that of Werlein in New Orleans in 1860.60 Other editions were
59 See above, note 48.

6 "I Wish I Was in Dixie, Words by Walter H. Peters, Music by J. C. Viereck"
(New Orleans:P. P. Werlein, i860). The words to "Dixie"without the music had been
published earlierin a songster:Bryant'sPowerof Music(New York: De Witt, 1859). See
James J. Fuld, TheBookof World-Famous Music(New York: Crown, 1966),
I67-69.
Other early editions of "Dixie" include: "Away Down in Dixie's Land, As Sung
&
by Hooley Campbell's Minstrels, words by 'JerryBlossom,' Music by 'Dixie, Jr.' "
(Louisville: D. P. Faulds, I86o); "I Wish I was in Dixie, Words by J. Newcomb,
Music by J. C. Viereck, Sung by Mrs. John Wood" (New Orleans: Werlein, I86o);
"I Wish I was in Dixie's Land," Written and Composed expressly for Bryants
Minstrels by Dan. D. Emmett, Arranged for the Piano Forte by W. L. Hobbs (New
York: Firth, Pond and Co., I86o); "Dixie Unionized... by A. W. Muzzy, Author

SOCIETY
published during I86o in New York, Philadelphia, Cincinnati, and

Louisville, and many more during the Civil War--often rewritten to
suit the Northern or Southern cause. The prints differ significantly
from each other, both in words and in tune, once again very likely the
result of oral transmission. The Firth and Pond edition is probably an
authorized version: it is attributed to Emmett, and its tune is identical
to the tune in the autograph manuscript.6'
The orally transmitted versions of "Dixie" show the same tenden-
cies as "Susanna," but not nearly as consistently. Example 16 collates
six versions of the opening measures. The "Dixie Unionized" and
Faulds versions subdivide the last beat of measure 2, thus adding
weight to the downbeat of measure 3. The Blackmar version, on the
other hand, eliminates the song's opening pickup entirely, thereby
confusing the rhythm. Werlein, Faulds, and Blackmar eliminate the
dot on the first beat of measure 4, again weakening a downbeat that
was stronger in the original. The tendency to pentatonicize is
^
glimpsed in measure 4 of the "Unionized" version, which avoids in
a scale of sixteenth notes. The Werlein and Faulds versions, on the
other hand, add a leading tone to the cadence at the end of the chorus
(see Ex. i7). The tendency for the harmony to draw the melody to the
root of the chord may be seen in Example 18, where Werlein and
Faulds give the chord root instead of the third as a pickup to measure
25. A similar change is seen in measure 24, where both sources
eliminate an appoggiatura. The tendency toward eliminating differ-
ences between parallel passages is scarcely seen in the "Dixie"
versions. Parallel passages are already so similar in Emmett's original
manuscript that there is hardly any opportunity for the postulated
tendency to operate.
This last observation suggests why the tendencies of oral trans-
mission operate so inconsistently in the case of "Dixie." In Chris
Goertzen's felicitous phrase, "The composer 'got it right' the first
time."62 The rhythm of the melody clarifies the beat with the
two-note opening pickup and throughout the song. The original
of the Words, Melody by Dan D. Emmett" (New York: Firth, Pond and Co., i861);
"The Original Dixey's Land ... Arranged for Piano by Augustus Cull" (New York:
Horace Waters, i86o); "Dixie--War Song, Written by H. S. Stanton, Esq.,
Arranged by A. Noir" (Augusta: Blackmar and Bro., 1861).
It is unlikely, however, that the Emmett manuscript served as a written model
for the Firth and Pond edition. The words in Firth and Pond are significantly
different from those in the Emmett manuscript, and the song has been transposed
from D to C. Where the print has a piano accompaniment, the manuscript has a violin
part.
62
Goertzen, "Mrs. Joe Person," 42 n. 8.

"OH! SUSANNA" 127
Example16
"Dixie," mm. 1-5, Emmett manuscript,and Firth and Pond, Werlein, Faulds,
and"Unionized"
Blackmar, editions
Emmett
orig.
D.. I wish I was in de land ob cot- ton,
Firth and Pondk
I wish I was in de land ob cot- ton,
Werlein k.
I wish I was in the land of cot- ton,
,
Faulds
Oh Dix- ie am de par- i- dise Wharde
Blackmar
orig. D
4"
Hear ye not the sounds of bat- tie,
Unionized I 1
" --
0 - O--
O! I'm glad I live in a land of free-dom,Wherewe
Emmett
orig. D
1-"
Cim- mon seed an san- dy bottom, Look a- way-
Firth and Pond
old times dar am not for- got- ten; Look a- way!
Werlein 9
cinna- mon seed and san- dy bot-tom look a- way,
Faulds
raise de cot- ton and de rice, Come a- way
Blackmar
orig D
Sab- res' clash and mus- kets' rat- tie? To Arms!
Unionized
have no slaves nor do we needthem,Lok a- way,

SOCIETY
Example 17
"Dixie," mm. 31-32, Emmett manuscript, and Werlein and Faulds editions
Emme
tt"
a- way downsouth in Dix- ie.
Werlein
a- way downsouth in Dix- ey.
Faulds
We'll soon be off for Dix- ie.
Example I8
"Dixie," mm. 23-26, Emmett manuscript, and Werlein and Faulds editions
Emmett a
orig.D - r
-
To lib an die in., Dix-
.to ie. A- way, a- way.
Werlein
and live and die in Dix- ey a- way a- way
Faulds
When we get down in Dix- ie. A- way, a- way,
"Dixie"always cadencesfromabove, never from the leadingtone; 7

occurs only on weak beats in scalar motion. The melody already
includes most of the chord roots, and it places them on prominent
beats. Finally, parallelpassagescorrespondclosely to one another.
Emmett'soriginal"Dixie,"as representedby the autographmanu-
scriptandthe FirthandPondedition,alreadyembodiesthe tendencies
of oral transmission.It cannot be transformed,or at least it will be
transformedonly inconsistently.63
To an extent, the differencesin this respectbetween"Dixie"and
"Susanna"can be explainedby the differencesin the situationsof the
composers. Emmett, when he wrote "Dixie," was an experienced
63 Goertzen notes that Mrs. Person's version of "Dixie" matches Emmett's

original almost exactly (ibid., 42). If the Sackses' hypothesis in Way Up North is
correct, and if "Dixie" was indeed part of the Snowden family repertory, eventually
transmitted to Emmett, this would help explain why the earliest versions of "Dixie"
already embody the tendencies of oral transmission.

"OH! SUSANNA" 129
performer, andhe wasworkingcloselywithBryant'sMinstrels.It is

possiblethat the "original"
versionof "Dixie"alreadyincorporates
revisionsfromrehearsals andperformances duringI859. Foster,on
the otherhand,at the time he wrote"Susanna," was a youngand
inexperienced songwriter;he was not a minstrelperformer; andhe
movedmainlyin a circleof amateurs.64 Foster,at this stagein his
career,wasnot as likelyas Emmettto get it rightthe firsttime.
Tune Transformation
The four tendencies of oral transmission proposed here are not

intended to constitute an exhaustive inventory. Indeed, the analysis of
"Susanna" suggests the possibility of at least one more tendency of
oral transmission: a tendency to replace settings that slur syllables
over two or more notes with strictly syllabic settings. This tendency
did not appear in the comparison of the Peters version with the others,
since Foster's original settings are syllabic throughout. It is seen,
however, in the fate of the two-note pickup added in the early stages
of oral transmission. In the early editions this pickup is a sung to a
single syllable (see Fig. 2). Many later versions, however, add a
syllable to accommodate the extra note. This is the case, for example,
in James Taylor's rendition, recorded in the 197os and shown in
Example 19. The tendency was at work already in the nineteenth
Example 19
"Susanna," mm. 1-2, James Taylor, Warner Brothers record I843 (original key: C)
Well I come from Al- a- ba- ma
century, for the Mitchell edition adds a syllable in a parallel passage

(Ex. 20). It may be seen as well in the transmission of "Dixie" at the
beginning of the "Dixie Unionized" version (Ex. i6). Further tenden-
cies might be postulated based on the analysis of other songs and their
transmission.
Although this study has emphasized mixed transmission, the
relation between oral and written transmission is yet more complex
than indicated here. For example, some of the differences between
editions of "Susanna"may be changes introduced not by oral trans-
64 See Austin, "Susanna,""Jeanie,"3-12.

SOCIETY
Example 20
"Susanna," mm. 5-6, Mitchell edition
And I'm gwan to Lou- si- a- na
mission but rather by editorialemendation.The details of dotted

eighths and sixteeenths,cited aboveas examplesof an oral tendency
towardparallelism,mightalso be accountedfor as the workof editors
and arrangerssimplifyingthe song for the public. The possibilitiesof
contaminationbecomeparticularlytroublesomein a mixed transmis-
sion. In an exclusivelywritten transmission,a scribe needs to have
two sourcesat handfor contaminationto occur;in a mixedtransmis-
sion, however,a versionthatis "inthe air"cancontaminatea tradition
without beingpresentin writing.The Holt editionof "Susanna"was
reprintedmany times and widely copied;it may have influencedoral
or written versionsdistantfrom it on the stemma.For example,the
English editions are not directly descendedfrom Holt, but some of
them (Addisonand Leoni)harmonize"death"in measure14 with a
diminished seventh chord, a hallmarkof the Holt edition and its
descendants.This couldbe the resultof a singeror an arrangerhaving
hearda renditionof "Susanna"basedon the Holt tradition.65
The four tendencies postulatedhere could be tested on other
repertorieswhere composedand notatedsongs subsequentlymoved
into an oral tradition.Englishand AmericanChristmascarolsmight
provide examplesfor such a study. The oral transmissionof songs
from eighteenth-centuryFrenchvaudevilleswould be anotherpossi-
bility;yet anothermightbe songsby Belliniand Verdi thathavebeen
passed down as popular airs alongside their written versions.66
Studiesof these and similarrepertoriescouldprovidea test of the four
65 Charles Seeger finds,

in his survey of versions of "Barbara Allen," that a
successful printed edition like Holt often influences the words to a song but seems to
have less influence on the tunes of orally transmitted versions ("Versions and Variants
of the Tunes of 'BarbaraAllen,' " in his Studiesin Musicology,1935-1975 [Berkeley:
University of California Press, 1977], 313).
6
The oral transmission of songs by Donizetti and Verdi is discussed by Marcello
Sorce-Keller in " 'Gesunkenes Kulturgut'and Neapolitan Songs: Verdi, Donizetti, and
the Folk and Popular Traditions," in Atti del XIV Congresso dellaSociettInternazionale
di Musicologia-Bologna, 1987 (Torino: Edizioni di Torino, 1987), 3:401-5. His
example of an oral version of "Questa e dunque l'iniqua mercede" from Verdi's I due
Foscaridemonstrates the tendency for the harmony to draw the melody to the chord
root. On the other hand, his examples do not display any tendency to pentatonicize
the melody.

"OH! SUSANNA" 131I
tendencies proposed here and might generate further hypotheses. A

rigorous investigation is only possible, though, if the original version
of the song can be identified and if a stemma or something close to it
can be established, independent of the changes being investigated.
Folklorists of the late nineteenth century noticed that many "folk
songs" seemed to be descended from composed songs of earlier times.
They turned this observation into a theory of gesunkenesKulturgut,
according to which music composed and notated as high art and
associated with the upper classes "sank down" to a lower cultural
stratum, where it was transmitted in altered and trivialized versions.67
The example of "Susanna"shows that the gesunkenes Kulturguttheory,
despite its elitist overtones, is correct in important respects: composed
songs did move from written to oral transmission, and they were
significantly altered in the process.
But are the changes introduced by oral transmission inevitably or
even usually inferiorto the originalreadings?Sometimes oral transmis-
sion does seem to efface the subtleties of a composed original. For
example the Brahms"Lullaby,"which in its originalversion begins as in
Example 21, is often heard in music boxes and cradle gyms with the
opening measures made parallel, as shown in Example 22. Here the
Example 21
Johannes Brahms, "Wiegenlied," op. 49, no. 4
Gu- ten A- bend, gut Nacht,
Example 22
Brahms "Lullaby," cradle gym version
parallelism,introducedpresumablyby oral transmission,does sound

trivialcomparedto the original.In the caseof "Susanna,"
on the other
hand, the tendenciesof oraltransmissionhaveevidentlyimprovedthe
tune. They makethe rhythmclearer;they improvethe voice leading;
67 For an early formulation of the gesunkenesKulturgut theory, see Hans Nau-

mann, GrundziigederdeutschenVolkskunde, 2d ed. (Leipzig: Quelle und Meyer, 1929),
I 14-17. Naumann discusses the transmission and trivialization of
song texts; he does
not discuss tunes.

SOCIETY
they make the tune easier to sing and easier to remember.When

composersfail to get it rightthe firsttime, perhapssingersand players
have something additional and valuable to contribute to the composi-
tional process.
PeabodyConservatory,The Johns HopkinsUniversity
APPENDIXA
Early "Susanna"Prints (1850 and Before)68
Voiceand keyboard
"Oh! Susanna, The Popular Song. Sung by All the Ethiopian Serenaders, Arranged
expressly for these celebrated singers by G. Herbert Rodwell." London: R.
Addison and Co., [1849]. [Addison]
"Susanna Dont You Cry for Me, Rewritten and Arranged by William Clifton." New
York: Joseph F. Atwill, 1848. Copyright deposited 28 Aug. 1848. [Atwill]
"Oh! Susanna, as Sung by The Ethiopian Serenaders, written by Wells, Arranged
for the Piano Forte." Baltimore: F. D. Benteen, n.d. Plate 1427. [Benteen]
"The Popular Song of Oh! Susanna, Arranged with an accompaniment for the Piano
Forte ... Sung by G. N. Christy of the Christy Minstrels." Boston: Oliver Ditson,
n.d. [Ditson-Christy]
"Oh! Susanna, as Sung by the Ethiopian Serenaders, Arranged for the Piano Forte."
Boston: Oliver Ditson and Co., n.d. Plate 1427. [Ditson-E.S.]
"Oh Susanna." Philadelphia: E. Ferrett and Co., n.d. [Ferrett]
"Oh! Susanna, Sung by G. N. Christy of the Christy Minstrels." New York: C. Holt
Jr., 1848. Copyright deposited 25 Feb. 1848. [Holt-I]
"Oh! Susanna, Sung by G. N. Christy of the Christy Minstrels (Fourth Edition)."
New York: Chas. Holt Jr., 1848. [Holt-4]
"Oh! Susanna, Sung by G. N. Christy, of the Christy Minstrels (Fifth Edition)."
New York: C. Holt Jr., 1848. [Holt-51
"Oh! Susanna." In The EthiopianGleeBook: Containingthe SongsSung by the Christy
Minstrels . . . arrangedfor Quartett Clubs. Boston: Elias Howe, 1848. Page 88.
Copyright deposited 12 Sept. 1848. [Howe]
"Oh! Susanna Dont You Cry." London: Leoni Lee and Coxhead, [ca. I85o]. [Leoni]
"Oh Susanna. As sung by the Sable Harmonists, Music by S. C. Foster, Arranged
by W. J. Wetmore." New York: W. E. Millet, 1848. One-line copyright
announcement. [Millet-i]
"Oh Susanna. As sung by the Sable Harmonists, Music by S. C. Foster, Arranged by
W. J. Wetmore."New York:W. E. Millet, i848. Two-line copyright announcement.
[Millet-2]
"Oh! Susanna, as sung by Dumbolton's Serenaders,in their popular Ethiopian Enter-
tainment, at the St. James's Theatre, Arrangedexpressly for these performances,by
F. Howard, and Nightly encored." London:John Mitchell, [I849]. [Mitchell]
"Oh! Susannah; and Goin Ober de Mountain." London: Musical Bouquet, n.d. No.
5 I. [Musical Bouquet]
68 Copyright deposit dates are taken from Gordon, "Preliminary Report." Dates
for English issues are taken from Catalogueof PrintedMusicin theBritishLibraryto 1980
(London: British Library, 1981-87).

"OH! SUSANNA" 133
"Susanna. as sung by Mr. Tichnor of the Sable Harmonists. Written and Composed
by S. C. Foster." [Louisville]: W. C. Peters, 1848. Plate 1228. Copyright deposited
30 Dec. 1848. [Peters-Louisville]
"Susanna. as sung by Mr. Tichnor of the Sable Harmonists. Written and Composed
by S. C. Foster." Kentucky: W. C. Peters, 1848. Plate 2157. [Peters-Baltimore]
"Oh! Susanna, as Sung by The Ethiopian Serenaders, Written by Wells, Arranged
for the Piano Forte." Baltimore: Geo. Willig Jr., n.d. [Willig]
Voiceandguitar
"Oh Susanna, As Sung by the Ethiopian Serenaders, Arranged for Guitar by

Thomas J. Martin." Baltimore: F. D. Benteen, 1849. Plate 1477. [Benteen-guitar]
"Oh! Susanna-Ethiopian Song, Arranged for the Guitar by Edward Pique."
Philadelphia: E. R. Johnston, 1849. [Johnston]
"Oh Susanna-A Popular Ethiopian Song, Arranged for the Guitar by F. Weiland."
Philadelphia: Lee and Walker, 1849. Plate 617. Copyright deposited 24 April 1849.
[Lee and Walker]
"Oh Susana, A Popular Ethiopian Song, Arranged for the Guitar by F. Weiland."
Baltimore: W. C. Peters, 1849. Plate 617. [Peters-guitar]
Keyboardonly
"Oh! Susanna, an Ethiopian Melody, Composed with Variations for the Piano Forte
by J. Hunter." Baltimore: Benteen, 1848. Plate 1414. Copyright deposited 2 x Nov.
1848.
"Susanna
Polkafor the PianoForte,As Playedwith unbounded applauseby the
SteyermarkMusicalCompany,Composedby FrancisRziha,Leaderof the
Company."Baltimore: Benteen,1849.Copyright
depositedI I Jan. 1849.
"Oh!SusannahQuickStep, In Whichare Introduced the FavoriteAirs of Oh!
Susannah& Oh!Carryme back,&c.,Arranged by EdwardL. White."Boston:
Oliver Ditson, 1848. Plate 1627. Copyright deposited 22 Nov. 1848.
"Oh Susanna! With Easy Variations, Composed for the Piano Forte by Edward L.
White." Boston: Oliver Ditson, 1848. Plate 1631. Copyright deposited 22 Nov.
1848.
"OhSusannaQuickstep,Arrangedfor the PianoForte."Philadelphia:
E. Ferrettand
Co., n.d.
"Oh! Susanna, With Variations, C. Grobe. Op: 124." Philadelphia: Lee and Walker,
1849. Copyright deposited 30 June 1849-
"Oh! Susanna, Introduction and Brilliant Variations, Composed and
Arranged for
the Piano Forte by James G. Maeder." Boston: W. H. Oakes,
1849.
"Oh! Susanna, Arranged by William Clifton." New York: Vanderbeek, n.d.
"Oh Susanna-Julius Quadrilles, Arranged from the Celebrated
Christy Melodies by
John Casper Scherpf." Second set, no. 2. New York: Vanderbeek, 1848. Copy-
right deposited 4 Nov. 1848.
Appendix B
"Oh! Susanna" Texts: Distinctive Readings
Readingsin the Peterstradition

I. I come from Al-a-bama
I'se gwine to Lou'si-a-na

SOCIETY
Chorus do not cry for me

2. trabbled down de ribber
De lectrick fluid magnified
De bulgine bust and de hoss ran off
I really thought I'd die
I shut my eyes to hold my bref
3. I thought I saw Susanna dear,
a coming down de hill
I says, I'se coming from de souf,
4. [Fourth verse lacking]
amongsourcesdescended
Readingswidespread from [c] (commonoral source)
I. I come from Al-a-ba-ma
I'se gwine to Lou-si-a-na
Chorus don't you cry for me
2. trabbled down de rib-ber
De lectric fluid magnified
De bulgine bust, de horse run off
I really thought I'd die
I shut my eyes to hold my breath
3. I thought I saw Susanna,
A coming down de hill
Says I, I'se coming from de South
4. I soon will be in New Orleans
look all round
I'll fall upon de ground
Dis darkie sure will die
Readingsfound in boththe Holt/Benteenand the MilletlFerretttraditions
i. I come from Al-a-ba-ma

I'se gwine to Lou-si-a-na
Chorus don't you cry for me
2. trabbled down de rib-ber
De Lectrie fluid magnified
De bulgine bust, de horse run off
I realy thought I'd die
3. I thought I saw Susanna,
look all round
Dis darkie sure will die
Readingsin the Holt/Benteentradition

. I came from Al-a-ba-ma
I'm gwan to Lou-si-a-na
Chorus Oh don't you cry for me
2. trabelled down de riber

"OH!SUSANNA" 135
De bullginebust,de horserunoff
I realythoughtI'ddie
I shutmy eyesto holdmy breath
3- I thoughtI sawSusanna,
A coming down de hill,
Says I, I'm comingfromde South
look all round
I'll fall upon the ground
But if I do not find her
Dis darkie'lsurelydie
Readingsin the Millet/Ferretttradition
I. Ise come from Al-a-ba-ma

Ise gwan to Lou-si-a-na
Chorus dont you cry for me
2. floated down de riber
De bulgine bust de horse run off
I realy thought I'd die
3. I thought I saw Susanna
4. I soon shall be in New Orleans
look all round
I fall right on de ground
Dis nigger sure will die
Readingsuniqueto Atwill
i. I come from Al-a-ba-ma

I'm gwine
Chorus Don't you cry for me
2. steamed it down de riber,
De Lectrick fluid den multiplied
De bulljine burst, de horse run off,
Dis nigger almost die
I shut my eyes to hold my bref,
3. I drempt I se'd Susanna
A rolling down de hill
Says I, watt makes de tears run so,
look all round
Dis darky sure will die
Readingsin the Englishtradition

I. I came from A-la-ba-ma
And I'm g'wan
Chorus dont you cry for me
2. trabelld down de riber

SOCIETY
De lectricfluidmagnified
De bulgineburstde horserun off
I reallythoughtI'd die
3. I thoughtI saw my Susy Ann
A rollingdown de hill
Says I, I'm comingto de South
4. I soon will go to New Orleans
look around
if I find my Susy Ann
Dis darkywill surelydie
ABSTRACT
Early prints of "Oh! Susanna"by Stephen Foster transmit versions of the

tune that differ strikingly from one another. It is likely that these variants
arose as "Susanna"was orally transmitted among minstrel-show performers.
Variant readings are compared in order to establish a stemma that shows not
only the filiation of sources, but also the ways in which oral and written
aspects were mixed in the transmission of "Susanna." The variants in
versions of "Susanna"demonstrate four general tendencies of oral transmis-
sion: (i) a tendency to alter rhythms in order to clarify the beat; (2) a
tendency to pentatonicize the melody; (3) a tendency for a salient harmony to
draw the melody to the chord root; and (4) a tendency to eliminate differences
between parallel passages. Analysis reveals that the four tendencies are also
present in the transmitted versions of other songs from the repertory of
nineteenth-century American minstrelsy.


Oh Susanna Artigo

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Oh Susanna Artigo

Uploaded by

Copyright:

Available Formats

"Oh!

Susanna": Oral Transmission and Tune Transformation

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

and depositedfor copyrighton 30 December1848. A facsimileof this

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

MN of tie Sable Written and Cornposed by s.V.Foster.

gwine to LoOsi n true I t s It rad all nig de day Ileft,

Figure i. The Peterseditionof "Oh!Susanna"(1848).FosterHall Collectionof the

beginningof thechorus(m. 17)therearethreedistinctversionsof the

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

Oh! Su- san do not cry for me;

Oh! Su- do not cry for mei

come from Wid my on my knee.

come from Al -_ba _ ma, Wid my on my knee.

•",,r .I a."rJ.:-A .. . . .... ... _'

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

I jumsp'd aboard the tel e.graph and down de ribber, De

reallythoughtI'ddiei I shut Iyeyesto holdmybref u saa dot youcry.

I had a dreama de udder night, when ebry ting was still;

I thought I saw Susanna dear, a cominng down de bill,

De buckweat cake was in her mouf, de tear was in her eye,

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

Ise come from Al- a- ba- ma wid my ban- jo on my

Ise come from Al- a- ba- ma wid my ban jo on my

I came from Al- a- ba- ma wid my Ban- jo on my

Figure 2. Collation of "Oh! Susanna" versions

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

see; It rain'd all night the day I left, De wedder it was

see It rain'd all day the night I left De wedder it was

see. It rain'd all night de day I left, De wedder it was

dry De sun so hot I froze to death Sus- an- na dont you

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

comefrom Al- a- ba- ma wid mi ban- jo on my knee.

come from A- la- ba- ma Wid my ban- jo on my knee.

Come from Al- a- ba- ma wid de ban- jo on my knee.

come from Al- a- ba- ma Wid my ban- jo on my knee.

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

"Susanna"as Foster wrote it? These questionscan be addressedin

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

I'se gwine to Lou' si- a- na

(b) Holt edition

I'm gwan to Lou- si- an- na,

My true love for to see,

My true love for to see;

written-out words and tune from a manuscript or print into another

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

sourcesis replacedby the looserconceptof tunefamilies.14 Gordon,

The hypothesis that differences between versions of "Susanna"are

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

This seeming contradiction disappears when it is recognized that

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

to oral and back to written transmission." But the enterprise of

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

Aspectsat the headof thelist aremostlikelyto be transmitted in

The best place to begin is with the words. Differencesbetween

alsobe consideredas an aspect.Fora discussionof the stemmaticsof attribution,see

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

differencesin tunes, but becausethere are four verses, there is more

26 See the list of aspects in the

This content downloaded from 62.122.76.60 on Mon, 16 Jun 2014 05:57:51 AM

[a] autograph [b] autograph

[e] words written [f] words written [g] al