The Affective Character of the Major and Minor Modes in Music

Author(s): Kate Hevner

Source: The American Journal of Psychology, Vol. 47, No. 1 (Jan., 1935), pp. 103-118
Published by: University of Illinois Press
Stable URL: http://www.jstor.org/stable/1416710
Accessed: 23-04-2015 18:49 UTC

REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/1416710?seq=1&cid=pdf-reference#references_tab_contents

You may need to log in to JSTOR to access the linked references.

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 Illinois Press is collaborating with JSTOR to digitize, preserve and extend access to The American Journal of
Psychology.

http://www.jstor.org

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
THE AFFECTIVECHARACTEROF THE MAJORAND MINOR
MODESIN MUSIC

of Minnesota
BY KATEHEVNER,University

This experiment is an investigationof the affectivequalityof the major

and minormodesin music.It is an attemptto determine(1) wheier
the historicallyairmed characteristics of the two modesare apparentto
listeners,and (2) if theyare,to whatextentthe recognition of thesechar-
acteristicsdependson trainingin music,on intelligence,or on the talent
measured by the Seashoretests.
In general,it has beenassumedby professional musiciansandby inter^
pretersof musicalmeaningthatthereis a differencein the affectivequali-
ties of the majorandminormodes.The majoris associated with the fol-
lowingcharacteristics: it is dynamic,an upwarddrivingforce;it is deter-
miningand defining,andmorenaturaland fundamental thanthe minor;
it expressesvaryingdegreesof joy andexcitement;it soundsbright,clear,
sweet,hopeful,strong,andhappy.
For the minor,the characteristic qualitiesincludethe following:it is
passive,a downwarddrawingweight; it is determinedand defined;it
expressesgloom,despair,sorrow,grief, mystery,longing,obscurity,rest-
lessness,melancholy; it is mournful,dark,depressing,doleful,dull, plairl-
tive,andsoothing.
Theexplanations givenfor theseeffectsvarywiththe theoryof consonancewhich
is accepted.1They are attributedvariouslyto combination and differencetones
(Helmholtz),or to tonalfusion (Stumpf),or to conformity and noncomformity
to a mentalstandard (Britan).2By thoseexperimenterswhodo notaccepttheeffects
as inherentin the physicalsoundsof the two modes,theyare explainedin terms
of specifictraining,individualdifferenees,and previousexperience( Downey,8
Heinlein).4The necessityof demonstrating of the major
the qualitiescharacteristic
andminormodesseems,to mostprofessional musicians, Theyhabitually
ridiculous.
use the minormodeto achievecertaineffects,althoughtheyareof courseverywell
awarethat thereare otherdevicesby meansof whichthesesameeffectsmaybe
produced.Variationsin tempo,rhythm,harmony,intensity,and in the melodic
$ Acceptedfor publicationJune 15, 1933.
1Fora good briefreviewof the theoriesand explanations see C. P. Heinlein
of themajorandminormodesin music,J. Comp.Psycbol.8,
Theafectivecharacters
1928,101 f, and for a shortreviewof the theoriesof consonancesee R. C. Lar-
son, Studieson Seashore's
Measuresof MusicalTalent,Psyrhol.Monog.)40, 1930
(no- 181), 33 73-
2H. H. Britan,TbePbilosotbyof Music}1911,1-311.
3J.E.Downey,A musicalexperiment, thisJOURNAL,9, 1897,63-69.
4Op. Cit.

103

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
 HEVNER
104
line are equally useful or perhaps more useful
in carryingthe meaning or ex-
pressivenessof the composition.Musicians also
recognize that in producing its
effect in the listener, the mode (major or
minor) is never the sole factor; it is
only a partof the total effect,inseparablefrom
the otherelements,rhythm,harmony,
melody,intensity,and tempo,and is moreoveralways
apprehendednot in an isolated
momentof time but as part of a sequence,
influencedby what has been heard im-
mediatelypreviously.For example,the soothingeffect
of a certaincompositionmay
be due to any one of a numberof different
factors or to a combinationof them;
to the slow tempo, or low intensity,or
smooth and simple harmony or the even,
unaccentedrhythm,or to the contrastwith a preceding
It would be difficultto extricateany one of animatedand excited section.
these elementsfrom the whole texture
of the composition,and to study its effect in
isolationSand this difficultyhas made
it impossiblefor experimentersto achieve
stableand consistentand veriSableresults.
Naturallythe effect on the listenersof a sprightly
minuetis cheerfuland stirring,
even though it is written for the most part in
the minor mode, and the effect of a
funeral march, with ponderous chords in slow
tempo, is heavy and depressing
even though it is written in the major key.
Psychologistswho have experimented
with mood effects by presentingto their
observersa numberof compositions,of
variouskinds, sometimesin one mode, sometimes
in another,have thereforealways
foundgreat inconsistenciesin the results,and they
have been led to the conclusion
thatthe effects are probablymost largely
determinednot by the music itself but by
somethingentirely extraneous,by the nature and
of the listener,or by his specifictrainingin complexityof the organic state
interpretingmusic.
Experimentsin which an attempt has been made to
modality,from all the other elements,have also led isolate the one factor,
to a denial of the historically
affirmedcharacteristicsof the two modes. The
most recent and extensive ex-
perimenton the affective characterof the major
and minor is that of Heinlein.5
Thisexperimenterused 30 Os, whose training
or lack of training in music had
beenpreviouslyascertained.They listened and
reportedthe eSect of a numberof
different series of major and of minor chords,presented
under very carefullycon-
trolledconditions.Each chord was presented
alone, and reactedto as a separate
unit.The intensity,the position, and the pitch
were varied systematicallyS and the
experimenter was able to show that intensity changes affect
the complexionof a
harmonic configuration.He states that "a change in intensityis
completely capableof almost
maskingthe historicallyairmed effect providedby
chords are rarelysoothing either in majoror minor, either mode."6Loud
and soft chords are frequently
soothing. Likewise "changesin pitch modify feeling tone
either mode. Chords,either major or minor, in the reaction to chords of
upper pitch registercall forth
alargerpercentageof major responsesthan
do chords in the lower pitch register.
Reciprocally, chords of either mode in the lower registercall
centage of minor responsesthan do chords of either forth a greaterper-
mode in the upper pitch
register."7Heinlein is also of the opinion that, "the
by assumptionlong entertained
theorists in regardto the supposed intrinsic
charactersof the modes must be
dismissed. It has been shown that reactionto harmonic
both configurationis variablefor
trainedand untrainedOs. The mode in which a
littlerelation to the type of feeling which the compositionis written has
composition may arouse....
5Op. ri.
oOp. rit, 138. XOp. [it., 140.

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
 THE AFFECTIVE CHARACTEROF MUSIC 105

Anyfixityof feelingtonein relationto a givenmodeis dependent upontraining

to reactin a specificmannerto a purelyintellectual discrimination."8
But theseconclusionsare basedon the studyof chordsonly-chordsin both
modes,groupedand heardin fairlyrapidsuccession,and thereare a numberof
seriousobjectionsto the use of suchmaterialas stimulifor an experiment in ex-
pressiveness or moodeffect.(1) In the firstplacethe stimulicannotbe regarded
as "music."The composerswho acclaimthe minormode as dull, melancholy,
plaintive,etc., have referenceof courseto its effectwhen it is used in musical
compositions. (2) The expressiveness, especiallyif it is in the natureof a mood
effect,requiresmorethanone momentof timeto be apprehended by the listener.
Theplaintiveness mayperhapslie onlyin the vccxmxlations of certainsensationsor
perceptions, requiringa relativelylong time for their apprehension. The rapid
shiftingof attituderequiredin listeningto seriesof unrelatedmajorand minor
chordswouldcertainly tendto obliterate
anyintrinsicdifferences whichmaybe found
in either.(3) Certainof theeffectsclaimedforthedifferent modeslie in themelodic
skips,thatis, in the interrelationof the toneswhichmakeup the two scales.Not
all of the effectsattributed
to the minoraredueto the harmony, for someof them
are arousedby the peculiarqualitiesof the melodieswrittenin the minorscales.
Theseeffectsaretotallyabsentwhenlone chordsareusedas stimuli.
In orderto obviatethe difficultiesin both the methodspreviouslyused
for the studyof the qualitativeeffectsof thesetwo modes,the methodof
presenting entirelydifferentmusicalcompositions,andthe methodof using
singlemajoror minorchords,the experimenter deviseda new methodof
procedure.The stimuli employedwere 10 short musicalcompositions,
withtwo versionsof each,one writtenin the majormode,the otherin the
minormode.By this means 10 pairsof stimuliwerepresentedwith every
factorheld constantexceptthe modality.In eachpairof compositions, the
rhythm,tempo,and intensitywere identical,and the only differencesin
melodyor harmonywere thoseoccasionedby the transposition from one
mode to the other.The difference,if any, in the effecton the listener,
mustthereforebe attributedto modality,since everyotherfactorin the
two versionsremainedunchanged.
The compositionsusedfor the experimentwereall relativelyshortbut complete
musicalideas,and theywere quitedifferentin character,includinggavottes,folk
songs,minuets,rigaudons,andantes,the.mes
fromsonatas,etc.,by Bach,Beethoven,
Schumann, Rameau, Gluck,andArensky. Someof themwerewrittenoriginally in the
majorandtransposed to minor,andsomeappeared in the minorandwerechanged
to major.
In considering
the natureof the changesmade,it is necessary to keepin mind
the fundamental differences
betweenthe majorand minormodes.Thereis one
formof the majorscale,and two formsof the minor,(a) the harmonic,with
an augmented secondbetweenthe 6th and 7th degrees,and, (b) the melodic,in
8 Op. rit., 140.

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
 , a X l= . rri:
4 4 s t a ! a r- < 1 n t s "

HEVNER
the
106
in formandloweredin
the ascending
tonesare raised since it is
the6th and 7th
nvhich form for our music,
We have employedthe harmonic
descending.
commonly usedtoday. tendenciesof the
more
much are determined throughthe inherent of the scale.
progressions
Harmonic to other triads
of the scale to move and an
builtupon the tones tonesof the majorscaleare listed below one
triads
whichappear upon the three minor,and
Triads thereare threemajor,
of themwill showthat
examination
triad.
diminished

. &
vi vii° I
iii IV V
ii
and it is to be noted
of the minorscaleappearsas below augmented triad.
triadpattern
The minor,
andone
twodiminished,
are
there two major,two
that

<n to Mo q
b: 9' b4'
tg VI vii° i
o
III+
v iii V
vs
v
in both
iio
two
i
seriesshows thatV and vii° are identical
iv, and VI in
Acomparison of these changeto i,
minor, and that I, IV andvi in the major and leading-tone triads,
and
major
no change in the dominantseventh the character
minor.We find
the and submedianttriads,
tonic,subdominant in themusic
a greatchangein the keyfeelingof ourmusic.All the changes triadar-
but
shalldetermine the maintaining the
which
of
have been madewith a view toward
this experiment
for two scales. in the ex-
as theyappearin the
rangements versionswhichwerecontrasted
of the two different original was written
As an example Martini.The
we quote here froma little Gavotteby
periment,
Allegromoderato
E . . 4
I i__
v ¢vtF , X ! -X-w i-

; 9 . .-s- ;4 S. ,>1
i ebqb -ia

;'..a.-lv t tf<z-3-
(b t 4! l i I'
.
-
-

2
#
-

4
-
s .
)

, f f f
/ br-n

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
 Q mf _ #1 1 i ; ; ; ;0 r;

OF MUSIC
THE AFFECTIVECHARACTER 107

in the major (key of P major), and the harmonyalternatesfrom tonic to dominant

seventh. Since the VTchord in the major and minor are identical, this leaves only
the tonic triad to be changed,but it is the tonic triad with its alteredthird, which
gives the definingtouch to the tonality.

Allegromoderato

( W ;; 1 551 tJ J;

iF hJ 1 i 4 1 t =-S

I ¢;+ G---' ;r -4 ' v-s- -

The other items and their changesare as follows:
Schumann,FolkSongfrom1heAlbumfor 1heYoung(Key of D Major).
Two changedversionsare used in this examplein orderthat each may be svholly
major or minor. The selection in the original is in ternaryform, A B A. A ap-
pears each time in minor, B in major.In the changedminor version B appearsin
minor and A is left unchanged,while the opposite occurs in the changed major
version.Therehavebeenno changesin the scale degreeof tones as they appearin the
context, but changes in pitch are necessitatedby the removal or insertion of ac-
cidentals.
Rameau,Le Tambourin (Key of E Minor), flrst part.
As in the Gavotte,harmonicmotion swings to and fro betweentonic and domi-
nant. Thus the minor version dependslargely on the tonic chord for its characteri-
zation, although the ornamentalchanging tones which are chromaticallyaltered
in the major version emphasizethe key sense. A slight pitch change occursin the
transitionafter the second appearanceof the main theme. In the original the creep-
ing chromaticdescent emphasizesthe minor tendencies,and to check this empha-
sis, we omit the half-steps,G sharp, to G natural,to F sharp, by repeating the
G sharp.
Arensky,ChansonRusse,from Transcriptionsfor the Young, edited by A. Siloti
(Key of E Minor).
The element which gives this example its characteristicminor touch is the use
of the minor scale triad arrangement,which is wholly cancelled in the major
version,where many originallyminor chordsare changedto major.
Bach,Musette,from EnglishSuite in G, edited by A. Wier (Key of G Major).
The original signatureof one sharp has been changedto one of two flats. The
most noticeableeffect of this change is the constantlyrecurringB flat of the tonsc

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
 108 HEVNER

chord which adds much to the minor feeling. In the seventh measureof the sec-
ond section the descendingmelodic scale has been used.
Beethoven,Minget,Op. 49, No. 2 (Key of G Major).
Minor feeling in the changedversion dependsmostly on the tonic chord here
becauseit is used in conjunctionwith the VT,excludingfor the most part all other
chords.In the changedversion there is a slight differencein the transition(meas-
ures 9-12). In the original this transitionappearsin the relative minor key of
E, which, because of its close associationwith the major merely emphasizesthe
formertonality, upon its second appearance.In the minor, the transitionis left in
the original key, the only means of maintainingthe original minor feeling with-
out departingtoo much from the original pitch.
Durand,Chaconne, edited by A. Wier (Key of A Minor).
This example is dependentlargely on the tonic and subdominantminor for
its key feeling. Both of these chords are changed to major in the different ver-
sion.
Gluck,Andante(OrpheoJ(Key of F Major).
The changed version of the Andante achieves its minor feeling from the use
of the triads in their minor scale arrangement.In addition to this the descend-
ing melodicminor scale is used in measurethree of the secondsection. In measures
nine and ten of the same section, it is felt that the harmonicform of the minor
scale is more appropriatewith the underlyingharmony.
Rameau,Rigdon (Key of E Major).
By adheringto the minor triad arrangementand, as in Le Tambourin,
ing notes which involve a tone of the minor insteadof the majorscale, the chang-
minor
feeling is stressedin the changedversion.
Bach, Minget, edited by E. Hughes (Key of G Major).
A very simple harmonicscheme which involves the use of I, IV, v governs
the first section. In the minor version I and IV become minor chords. In the sec-
ond section the relative minor key is used leading us back into the original key.
In the changedversion, substitutionof the subdominantkey of the relative major
has beenused here,and thoughthis is not a closelyrelatedkey it can be used without
changingthe staff degrees. By using an F sharp in the 12th measureof the sec-
ond section, we anticipatethe returnto G. Minor.
These variouscompositionswere presentedto 4 differentclass groups, including
in all 205 students, in the ElementaryLaboratory(Sophomore) courses at the
Universityof Minnesota.They were played by the same pianist, who had taken
especialpains to play both versionsof the same compositionin an identicalmanner}
and who, in the opinion of the E, was successfulin so doing. Each group heard
all of these compositions,5 majorversionsand 5 minorversions.No studentheard
botk the major and the minor version of any one composition.E gave names to
the items, such as "Partof a Sonatain 3/4 time," "Song," "Studyin 4/4 time,"
etc. During and immediatelyafter the playing of each item, the Os checkedfrom a
mimeographed word list all the adjectiveswhich seemedto them descriptiveof the
music. One list of adjectiveswas arrangedin groups of opposites for the Os
convenience,and anotherlist was arrangedwith the words of similar connotation
listed together.O was told to check as few or as many as he pleased, and a new
copyof the list was furnishedfor each composition.The lists are as follows:

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
 THE AFFECTIVE CHARACTEROF MUSIC 109

LIST I
Check all the terms in the following list which seem appropriateto the music.
Checkas manyor as few as you like:
vlgorous tranquil heavy delicate
sad happy depressing exalting
serious light exciting soothing
longing satisfying obstinate yielding
sacred secular frustrated triumphant
restless quiet lyrical dramatic
bright dark gloomy gay
melancholy merry
robust serene sentimental tragic
emphatic leisurely tender plaintive
martial hushed passionate mournful
digniSed dreamy yearning pathetic
majestic vague pleading doleful
ponderous
sober
dramatic weird cheerful clownish
impetuous mysterious exhilarated humorous
agitated awe-inspiring ioyous whimsical
sensational grotesque soaring farcical
stormy gruesome triumphant playful
fanatic boorish
spiritual horrible graceful
lofty fantastic fanciful
solemn quaint
mystical sprightly

After the firstplayingof the composition, followinga brief rest period,the

same10 versionswereplayedagain,andthis timethe Os wereaskedto checkthe
oneadjectivein eachpairof a list of oppositeswhichbestdescribedthemusic.
LISTII
In the following pairs of terms, be sureto checkone in eachpair.
vlgorous tranquil heavy delicate
sad happy depressing exalting
serious light exciting soothing
longing satisfying obstinate yielding
sacred secular frustrated triumphant
restless quiet lyrical dramatic
bright dark gloomy gay
merry melancholy dull spirited
What quality in the music is most importantin giving it its character?
rhythm tempo harmony melody
H.
ow does it appealto you?
like indifferent dislike
Does the piece seem familiar?Do you think you have heardit before?
no
yes
Can you give the title or composer?

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
 110 HEVNER

RESULTS
(1) TablesI and II show the actualnumberof votes for the various
adjectivestabulatedfrom the Srst scheduleblank,on whichthe student
couldcheckas manyor as few adjectivesas he chose.In TableI are all
the adjectiveswhichwerecheckedmoreoften for the majorversionthan
for the minor.Theseadjectivesrepresentthe characteristics
of the major
modeas theywere perceivedby this groupof listeners.Likewisethe list
in TableII representsthe adjectives
whichwereassociated
with the minor.
Besidesshowingthe actualnumberof votesfor eachword,the tablesrank
the adjectives
in orderaccordingto theirimportance
in characterizing
either
TABLE I
SHOWING
NUMBER
oF VOTES
FORMAJORADJECTIVES
PORMAJORANDMINOREIERSIONS
Adjective Votes D Adjective Votes D
Major Minor P.E.diff. Major Minor P.E.diff.
Happy 3I6 IOl I7. l7 Sxring 40 ll 3 . I6*
Light 374 I7I IS 43 Leisurely 98 70 3. II*
Sprightly lSI 8I I4. 9S Serene 9I 68 - 7I*
Cheerful l86 IIS I4. Sl WhimSlCal 8s 6s . S°
Joyous I96 S3 I3 . 94 Delicate I34 I09 - 4
Gay 97 Il7 I3. l8 Robust 3I I8 }- 3I*
Bright z6o I08 Il. 96 Lyrical I46 IlI * 3I*
Merry ll4 89 Il. 4S Vigorous 7S 6I 1. 7S*
Playful l03 7S IO. 8S Spiritual 34 l7 1. l3*
Graceful 43 I44 8. IS Exalting 47 39 I. lI*
Exhilarated I I4 S4 7. l0 Clownish 7 3 I . o8*
Satisfying 86 48 4- 87 Farcical 8 4 I . o8*
Qpaint 93 6I 3. 88* Triumphant 70 6} 1. o7*
Impetuous 4S lS 3 - 8S* Soothmg 97 9° O. 78*
Panciful I I7 8I 3 . 68* Exciting 7° 64 ° ' 63*
Humorous l3 8 3- 33* Secular z3 zz O.;*
Sacred 36 I7 3. I6
* Ratios not statisticallyreliable.

majoror minor.This rankorderfor the adjectivestakesinto accountnot

onlythe sizeof the ratioof majorto minorvotes,butalsothe totalnumber
of personsvotingfor the variousadjectives.
Forexample,in TableI there
were374 votesfor 'light'whena majorversionwasplayed,and 171 votes
whena minorversionwas heard,whichgives a 2.18 to 1 majorityfor
lightnessas a characteristic
of the majormode.Again,in TableII, there
was 1 vote for 'stormy'when a majorversionwas played,and 11 votes
whena minorversionwas heard,whichgives an 11 to 1 majority-an
overwhelmingly largemajorit;yfor stormyas a characteristic
of the minor
mode.Buttheseratios,2.18 to 1 and 11 to 1, do not takeinto accountthe
numberof peoplevoting,whichis an importantconsideration, for in the
caseof 'light'therewere545 votescast)andin thecaseof 'stormy' only 12.
Sincetherewere205 Os, and eachone listenedto 5 majorversions,the

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
 6

THE AFFECTIVE CHARACTEROF MUSIC lll

highestpossiblenumberof lrotesfor anyone adjectivewas 1025,of which

545 is 53% and 12 is only 1.29. Obviouslysomemethodfor ratingthe
importance of the adjectivesin termsof the percentageof peoplevoting
for themwas necessary,a schemewhichwouldreducethe high majority
ratiosof suchwordsas 'stormy'whenthe percentage of peoplevotingwas
at a minimum.Thiswasdonebycalculating ie differencebetweenthe per-
centagesof majorandminorvotesfor eachadjectiveandthe P.E.for each
difference.The ratioof the differenceto its P.E. (D/P.E.dtfr)is a direct
II
TABLE
SHOWING OFVOTES
NUMBER FORMINOR FOR
ADJECTIVES ANDMINOR
MAJOR VERSIONS
Votes D Votes D
Adjective Adjective
_.!
Major Minor P.t.E.diff Major Minor P.E.diff.
9z
.1 _

Pd
LttetlC IS I IS .96 Hushed 7 49 3 SS*
lelancholy
M 36 I38 I II . 49 Ponderous I6 33 3.o8*
.aintive
P1; 47 I4 I IO. 69 Sentimental 47 7I 3 °3-
sarning
Ye 39 II3 8.78 Fantastic 8 zo z. 86*
MIournful z8 9z 8.o8 Quiet 89 II8 lx 83*
,d
Sal S° IlI 7^93 Agitated l0 36 z. 8I*
Sober 49 I 10 7 4 Heavy I5 3° l. 69*

eading
P14 zS 77 7.o8 Tender 7 93 -3S*
MIysterious 6 47 7¢oz Stormy I II l. I9*
znging
Lc 68 I33 6.74 Majestic z1 3 z . o8*
Doleful I7 66 6.7I DigniEed 69 8l t- 53*
loomy
G1 3° 76 6-ls Grostesque I. S°*
Restless 8s I49 6.lo Emphatic 38 4S 1. lIF
Wteird S 38 S . 6I Passionate I; I8 I. 19*

ramatic 6z Io9 S . 4I Solemn 89 IOO I. I6*

MIystical 9 3S S. lI Lofty zS 31 I. IS*

rious 86 I 36 S. I6 Martial IO t4 I . o8*

epressing
Dx 36 76 S. I3 Yielding II I3 o. 8X*
Tragic 7 33 4 * 8I Tranquil 79 85 O. 7I*

reamy
D1 66 I°S 4*47 Obstinate 3 S o-s4*
Frrustrated 9 30 4* I7 spiring
A we win I 8 o.z7*
'ague 7 Sz 4-°3 Sensational S 6 o@z7*
tark 9 j9 3. 96*

* Ratiosnot statisticallyreliable.

measureof the importance of the adjectives,and in TablesI and II the

adjectiveshavebeenlistedin orderaccordingto the sizesof theseratios.8
In both of the tablesall the differences sig-
whichare not statistically
nificantaremarkedwith an asterisk(*), althoughin manyof thesecases
the ratioof D to its P.E.is so closeto 4 thatit maybe considereda satis-
factorydifference.
A gIancethroughthe list of adjectivesin these tables,with the sub-
stantialmajoritiesfor manyof them showsa moststrikingadherenceto
airmed characteristics.
thehistorically Fortheminormodethe melancholy-
8The P.E. of the proportionsand of the differenceshave been calculatedfrom
Yule's formulae:P.E.P- .6745oP = (pan)%; P.E.dflr- (P.E.2e + P.E.2p)%.

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
 112 HEVNER

mournful-gloomy-depressing qualityprobablystandsout as of greatestim-

portance,andthentheplaintive-yearning-longing quality.It has alsoquite
clearlya mysterious-mystical-weird-restless-dramatic effect,and a vague-
dark-dreamy-hushed-sentimental aspectThe moreextremeideas,majestic,
grotesque,sensational, etc. shownegligibledifferentiation. For the major
modethegreatestimportance is givento suchadjectives as happy,sprightly,
cheerful,joyous,gayandbright,andthenextgroupincludesplayful,grace-
ful, andevenquaintandfanciful.Eventhoughmostof the remainingad-
jectivesdo not discriminate reliably,the differenceswhichtheybringout
are alwaysin the expecteddirection.The one possibleexceptionis the
adjectivesoothing,whichgivesa negligibledifferentiation, althoughit has
beenconsidered by musiciansas characteristicof the minor.
(2) On the secondhearingof these 10 musicalcompositionsthe Os
weregivena schedulecontaininga list of 32 adjectivesarrangedin pairs
of opposites,suchas merry-melancholy, sad-happy, exciting-soothing,etc.,
andthe instructions wereto the effectthattheymustchoosethatadjective
from eachpair whichbest describedthe music.It was pointedout that
anyone composition mightnot seemto be eithertrulymerryor outstand-
ingly melancholy,but that the choicesthey madewould indicatemerely
that the musicseemedmoreinclinedto one than to the other.In most
casesthe studentsapparently foundlittle difficultyin followingthesein-
structions,for the omissions,on the average,amountedto less than517s,.
TableIII showsthe resultsof thesevotes.The percentage of votesfor the
majorand minorversionsare recordedin columnsone and two. Since
the studentswere obligedto choosebetweenthe adjectivesin eachpair,
60% in the majorfor the firstadjectiveof anypairwouldmean409 in
the minorfor the secondadjectiveof the pair.On firstthoughtit might
seemthat the simpleratiosbetweenthesepercentages wouldbe the best
methodfor ratingthe adjectivesin this table in the orderof their im-
portancefor characterizing the two modes.This wouldmeanthat 'bright'
wouldhavea ratioof 75.8 to 43.3 (or 1.75 to 1.00 majority),for char-
acterizing the majormode,and 'satisfying'a 63.4 to 30.5 (or 2.08 to 1
majority),etc. This method,however,is not advisablebecause,as we
pointedout in connectionwith TablesI and II, such ratiosdo not take
intoconsideration the numberof personsvoting,andtoo muchimportance
is given to adjectivesfor whichthe percentageof totalvotescast is at a
minimum. Forexample,fromthe figuresgiven above,the ratioof major
tominorfor 'bright'as a major characteristicgivesa 1.75 to 1.00 majority,
butthe corresponding ratioof minorto majorfor 'darkJ as a characteristic
of the mirorwouldbe 56.7 to 24.2 (a 2.30 to 1 00 majority),simplybe-

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
 THE AFFECTIVE CHARACTEROF MUSIC 113

causeof the fact thatthereare fewerpeoplevotingfor happyand there-

fore anyonevotecarriesmoreweight.Underthe conditionsof the experi-
ment (forcingthe studentsto chooseone or the otherof the adjectives)
thesetwo percentages maJtrepresentthe samefact,andanymethodwhich

TABLE III
SHOWINGPPRCPNTAGP OPVOTPSPORTHPMAJORADJPCTrVPS PORMAJOR
ANDMINORVPRSrONS
Adjectives E Low
Percentvotes D High Talent
Low High
Major Talent Sea Sea High Low
(Minor) }Major MinorP.E.diff. and and shore shore C.A.T. C.A.T
TrainingTraining
Satisfying 63.4 3°-S z3.oo II.08 7.90 9.40 7. 4I 6.z4 9-Sl
(Longing)
Bright 7S-8 43-3 zz.89 3.37 9-l° Io . 68 7.9o 7-70 8.38
(Dark)
Merry 67.6 3S. I z- 73 IS.4 9.oo I I . ll S-99 S-S4 8.40
(Melancholy)
Happy 7o.4 37-4 l- 4S I l . 98 IO . I 3 II . 40 7.80 S-60 g.40
(Sad)
Gay 7 I . 6 4o. I l I . 43 I3.04 8- 30 IO. o3 7.66 7-4I 8.oo
(Gloomy)
Exalting 7I. I 40.6 °-7S II. s8 7-9° 9.7o 7-l9 S-°9 9-°9
(Depressing)
Light 68. I 37-9 °-S4 Il.lI 8.69 9.zo 6.96 S-3S 9.60
(Serious)
Triumphant 6S-6 38.8 I 8. z3 8.60 6.30 6.68 6.87 6.76 s.98
(Frustrated)
Delicate 7S-9 s4.8 4. 93 4-7I 8. IO 4-39 7-SS 4-88 S-33
(Heavy)
Spirited 73-9 s4.8 I 3 . 4S IO . I 3 S*96 S-°9 4-3°
z.67 S-I4
(Dull)
Lyrical 66-S S°-7 IO. 7S S- 3I 3 -77 3. l8* 6.76 3-96 z.9I*
(Dramatic)
Vigorous S°-S
36.6 9. lI 8. zo I . 70* S-S9 I. Il
* I *So* t S*I 6
(Tranquil)
Exciting so.6 39. I 7.6l S*S9 3. 96* 4-34 I. 50 o.oo* 4.zo
(Soothing)
Secular 74-6 66.8 S-6S 3-7l* 4-49 4. SI 4-S4
o-gg*t 4-94
(Sacred)
Yielding s8 4 S3. I 3- SI* o . 38 tI *87 I- 30*' z.o6* S-3S I- 3S
(Obstinate)
Restless S9-9 SS. I 3. I8* 6. S7 - 4S* 3.9o* I. I6 z.67* I-8S
(Quiet)
Median 9-36+ 7.Io+ 6.I3 + 6.8I+: S ll+ S 6S+
.89 *S7 .S6 *S4 *39 *Sz
* Ratiosnot statisticallyreliable.
t Indicatesthatthe majority is in favorof theminoradjective.

introducesa discrepancy in the majoritiesfor brightand darkcannotbe

usedin treatingthesedata.In determining the relativeimportanceof these
adjectives,we havethereforerepeatedthe procedure usedin TablesI and
II, and have markedthe ratiosnot statisticallyreliablewith an asterisk.
In arrangingthis table,the percentagesrecordedare those in which the
largernumberof voteswerecastforthe majorversion;hencethis list of

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
114 HEVNER

of the majormode as they were

adjectivesrepresentsthe characteristics
perceivedby this groupof Os. The associated adjectives,characteristicof
the minor,alsoappearin the parentheses andthesetwo lists of adjectives,
withthe substantial shownin the ratiosin column3, showagain
majorities
a strikingadherenceto the historicallyaffirmedcharacteristicsof the two
moces.
SinceHeinlein,with his methodfor studyingthe characteristics of the
two modes,found that "anyfixityof feelingtone in relationto a given
modeis dependenton trainingto reactin a specificmannerto a purely
a furtheranalysisof the datawas madein
intellectualdiscrimination,"9
orderto studythe effectsof musicaltraining,of musicaltalent,andof in-
of the two modes.Since
telligenceon the perceptionof the characteristics
the 205 Os for the experimentwere all studentsin the Laboratory Psy-
chologycourse,theirratingson scalesfor musictalentandtraining,10 their
averagepercentilerankingson the six Seashoretestsfor musicaltalent,1l
and theirscoreson the MinnesotaCollegeAbilityTests,were available.
The 25 Os withthehighestscoreson the Seashoretestsandthe twenty-five
withthe lowestscores,eachgrouprepresenting approximately an eighthof
thetotalgroup,weresingledout; theirvotesweretabulated separately,and
theirpercentages, andthe ratiosof majorto minorvates,werecalculated
as above.Likewisethe highestand the lowesteighthof theseOs on the
musicaltalentandtrainingscales,andthehighestandlowesteighthon the
CollegeAbilityTest were separatedfromthe restof the groupand their
votestabulated.The effectof thesethreefactorson the perceptionof the
majorand minorcharacteristics may thus be studiedby comparingthe
ratiosfor thesehighestandlowestheights.In TableIII, columns4 and52
are the ratiosfor the lowestand highestgroup in the ratingscalesfor
musicaltalentandtraining.The actualpercentages fromwhichtheseratios
significant
were calculatedare not given, but all that are not statistically
aremarkedwith an asterisk.While it is truethatthe ratiosfor the group
with a largeamountof trainingare somewhatlargerthanthose for the
groupwithouttraining,it is alsoapparentthatthe studentswithouttrain-
andto a highlysatisfactory
ing do perceivethe differences, andthoroughly
reliabledegree.On the averagethe ratiosof the highlytrainedare only
slightlybetter.The medianof the ratiosfor the high groupis 9.36+0.89
and for the low group7.10+0.57. The ratioof this differenceto its P.E.
(2.26+1.10) does not allow us to predictthatthereis a realdifference

9 Op. rit.,140.
10W.S. Fosterand M. A. Tinker,Notes for In.rtructors,Experimentsin Psy-
rhology,rev.,ed., 1929,31.
1lC. E. Seashore,The PsZrhologyof MusicHTalent, 1919,1-288.

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
 THE AFFECTIVE CHARACTEROF MUSIC 115

betweenthese two groups.This lack of a substantialdifferenceis the

moreimportantin view of the greatdifferences in musicaltrainingwhich
thesegroupsrepresent.The medianscorein the highlytrainedgroupis
90 + 3.5, whilethe medianfor thegroupwithlittletrainingis 15 + 0.7.
An examinationof the actualpapersof the two Os who representthe
medianfor each group will give some indicationof the natureof the
differencein training.The low groupmedianstudenthas neverhad any
privatemusiclessonsof any sort, does not play any musicalinstrument,
has no musiciansin his family,and neverattendsoperasor concerts.He
can readmusica little for singing has sung two yearsin a High School
chorus,andhas a pianoanda radioin his home.The high groupmedian
subjecthashadprivatelessonson the pianofor two years,voicefouryears
andcellofouryears,hasplayedfouryearsin orchestras andsungsix years
in choirs,has threemusiciansin his family,can readmusicand provide
missingparts,attendsconcertsveryfrequently, andcansing or playbefore
small audiences.Yet even with such great differencesas these in back-
groundandtraining,the differences in the abilityof theseOs to perceive
the characteristicsof the majorandminormodesarenot large,and,what
is moreimportant, the studentswithouttrainingdo discriminate accurately
and reliablybetweenthe two modes.TherearelikewiselargedifFerences
in the averageSeashorepercentilescoresfor the two groupscomparedin
columns6 and7. The medianfor the highgroupon the six Seashoretests
is at the 88thpercentile(88.3+0.92) and for the low groupat the 40th
percentile(40.3+2.01). But the ef3ectof these dif3erences in musical
abilityon the sizeof the ratiosis verysmall.Therearethreenon-significant
differences for the highlytalentedgroupand four for the low group.The
medianratiofor the high groupis 6.13 + 0.56 and for the low group
6.8+0.54, and iis differenceof 0.68 in favorof the low groupis not
even as large as its deviation,0.78. The ratiois 68/78. This indicates
practically no differencein abilityto discriminate betweenthe two modes.
Finally,in comparing the lowestandhighesteighthof ie groupin Col-
lege Abilitypercentileratings,we findthereis a medianof 13+1.6 for the
low group,and95 +0.67 for the highgroup,butin spiteof thisverylarge
difference in intelligencetestscores,the diSerencebetweenthe averagesfor
the tatiosis againnegligible.The differenceis .43 in favorof the less in-
telligentgroup,andie deviationis 0.65. The intelligenceof the listeners
doesnot affecttheirabilityto discriminate betweenthe modesin any re-
liableor measurable way.
A comparison, then,of thesethreefactors, showsthattrainingis of more
importance thaneithermusicalabilit as measuredby the SeashoreTests,
or intelligencein discrimin,ating the characteristics
of the two modes,but

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
116 HEVNER

TABLEIV
0F THE
ANALYSES COMPOSITIONS

D % of
ComposiP.E.diff.strongModula Length
Structure, andHarmony
Tonality
tion beats tion
with
thirds
Martini I4.58 7S None Fmajor;allegromoderMostlyI to V. IV sug
Gavotte 4/4. 4 gestedoncebutonlyasa
ato;8 measures
measurephrasere passing toneandwiththe
peated. fifthomitted.
Bach IO.9; 7S None G major;
Lzistesso;I6 Writtenintwopartsand
Musette measures4/4. ABA organ point. Harmony
Bpreserves
form; same tendstowardvagueness
rhythm,similarhar because(I) manysecon
monyandmelody. darychordsareused;(z)
manychordsareambig
uousand;(3)manynon
harmonictonesfallonac
cents.
Rameau IO.8S 56 None I6 HarmonyismostlyI-I-IV
Emajor;allegretto;
Rigaudon measures4/4. ABA V-etc. In B sectionII7
form;A consistsof a andV7occuronce.
fourmeasure subject
Bhassimilar
repeated.
rhythmanda subject
infeeling.
similar
Rameau 9 s8 3; None Eminor;allegromolto Harmonyand tonality
Tambourine 44 measures4/4. verydefinite throughout.
ABACAform.Sub I and IV alternation
stac throughout
jectsverysimilar, forone
except
catoandgracenotes transitionalpassage.Lead
used in same way ing tone occurs fre
throughout. quently.
Schumann8.88 ;9 Momen D major; ;4 InAsection,I,IVandV
andante;
FolkSong tarilyto measures4/4. ABA prevailwithVIoccurr1ng
V at twiceandanunusual
form.BdifFersinrhyth ca
phrase mic figure,tempo, dence,VI, II7,III,I. Sec
endings. andtypeofmelody. tion B alternates I-V7,
thenmodulatestoV with
a V7-Iinthatkey.
BeethovenS.73 SS None G major;Tempodi A sectionalternates I to
Sonata minuetto;zomeasuresV. B sectionsomewhat
when
for3measures
3/4.A, a shorttransi vague
tion, B, and A re thereis analteredd fol
peated. lowedbyIV-III.

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
 THE AFFECTIVE CHARACTEROF MUSIC
117
TABLEIV (Continued)

D %of
Composi P.E.diff. strongModula Structure,
Length TonalityandHarmony
tion beats tion
with
thirds
Bach s.os ;5 Torela G major;animato;;4 In A, I-V prevailswith
Mfnuet tiveminormeasures3/4. ABA a few IV chords.In B,
form.A hastwo con modulationto relative
trastingphrases
repeat minor alternatingI-Vo.
ed. B maintainsthe Harmony .very clear
rhythmof A but the throughout;onlyonesecS
descendingfiguresin ondarychordoccurs.
the melody contrast
wih the ascending
figuresin a.
Durand 3 76 53 None Aminor;allegretto;I7 I-V successionswith IV
Chaconne measures ;/4. One occurring
at startof each
subjectwith twocon phraseand finalcadence
trasting phrases,re IV-V-I.
peated
Gluck 3. 65 36 To V F major;
andantino;
z8 I-IV-I-V. Modulation
Andante measures3/4. ABA to V. No secondary
(Ckpheo) form.8 measures
of a; chordsusedexceptin caS
B merelya develop denceII6rI4tVoI.
mentofA.
Arensky I.7z 33 To IV E minor; adagio; 8 Verycomplex,many7ths,
Chanson andthen measures 4/4. Onesub appogiaturosand two
Russe torelative jectwithtwophrases, modulations.I-VI-VIIr
minor repeatedandin repeti V-I-VI-VIIrI-VId
tionverydiffierent
har VoI-(a minor)VoIs (g
monywith two modu major)VoIV-VII13rVII7
lations. (e minor)I46-VI6-VoI.

thatneitherlackof trainingnor lackof musicalabilitynorlackof intelli-

gencecanobliteratethe abilityto discriminate the acceptedmeaningsfor
the two modes.
An attemptwas madeto discoverthe particular qualitiesin the music
whichaccountfor the diSerencesin the eSectsin the two modes.In order
to makethis study,the resultsweretabulatedseparatelyfor eachcompo-
sitionto determinewhichitemshowedlargedifferences andwhichshowed
smallor unreliabledifferences.
Thenan analysisof the musicwas madein
orderto accountif possiblefor thesedifferences. For the tabulation,six
pairsof adjectives
wereselected,satisfying,merry,happy,gay,delicate,and
exalting,andthe ratiosfor all six adjectiveswereaveraged.Thereforein
TableIV we have,notthediscriminative abilityof anyoneadjective, on the
basisof ten compositions,as in TablesI-III, but the discriminative im-
portanceof one compositionin termsof these six differentadjectives.

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions
118 HEVNER

TableIV lists the compositions in the orderof theirimportance in differ-

entiatingthe two modes,andgivesalsothe mostimportant factsaboutthe
length,tempo,complexity, andharmonization of themusic.It givesalsoan
indicationof the numberof timesthe thirdof the scaleoccurs.Sincethe
thirdis probablythemostimportant tonein establishingtheminorfeeling,
the numberof timesthatthe thirdappeared, andespeciallythe numberof
timesit appeared on a strongbeat,mightbe an importantfactorin estab-
lishing the mood effect.We have, therefore,tabulatedin Table IV the
proportionof strong-beats-with-thirds to the totalnumberof strongbeats
in thecomposition.
A studyof TableIV doesnotrevealanyclear-cutstructural differencesin
the compositions corresponding to theirimportance in differentiating the
efFectsof the twomodes.A comparison of the itemsin the firsthalf of the
tablewith thosein the secondhalf shows that thereis less modulation
amongthe first5 anda largerpercentage of strongbeatscarryingthe third
of the scale.Thebest 5 compositions alsoshowthatsimplicityof structure
and definitenessand uniformityin tonalityand harmonization emphasize
the differentmoodeffects.Thiswouldalsobe the logicalconclusion,since
varietyin melody,rhythm,or tempowouldincreasethe rangeof responses
andtheamountof overlapping for themajorandminor,whilevagueness or
lackof definitecharacter in the tonalitywouldincreasethe uncertainty on
the partof the Os andbringthe two modesclosertogetherin theireffects.

Su MMARY
fl) The previousmethodsof investigatingthe meaning,or moodef-
fect,of themajorandminormodeshavebeenunsatisfactory eitherbecause
(a) they left too manyfactors,suchas rhythmand tempo,uncontrolled,
or (b) theypresentedisolatedand disconnected elements(chords)which
werenot apprehended as "music",or (c) therewas no possibilityof a
cumulative effect.
f2) A new methodof investigatingthe effectsof the two modeshas
beendevisedin whichthematerialusedconsistsof twoversionsof thesame
composition,onewrittenin the majorkeyandonein theminorkey.
f3) With this method,all of the historically
airmed characteristicsof
thetwo modeshave beenconfirmed,in mostcaseswith very substantial
majorities,by 205 listeners,for 10 compositions
representinga varietyof
musicalforms.
f4) By the methodof groupdifferences, the factorsof musicalback-
groundand training,of musicalability (Seashoretests), and of intelli-
gence,havebeenfoundto be influential butnotin anycaseessentialin dis-
criminatingthe moodeffectsof the tvvomodes.

This content downloaded from 139.80.123.44 on Thu, 23 Apr 2015 18:49:39 UTC
All use subject to JSTOR Terms and Conditions