You are on page 1of 31

Inaudible Structures, Audible Music: Ligeti's Problem, and His Solution

Author(s): Jonathan W. Bernard


Source: Music Analysis, Vol. 6, No. 3 (Oct., 1987), pp. 207-236
Published by: Blackwell Publishing
Stable URL: http://www.jstor.org/stable/854203
Accessed: 07/02/2010 18:29

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=black.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

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.

Blackwell Publishing is collaborating with JSTOR to digitize, preserve and extend access to Music Analysis.

http://www.jstor.org
JONATHANW.BERNARD

INAUDIBLESTRUCTURES, AUDIBLEMUSIC:
LIGETI'SPROBLEM,AND HIS SOLUTION

Not long after his emigrationto the West in 1956, GyorgyLigeti decidedto
challengea well-establishedcompositionaltrend. Ligetiis knowntodayas one
of a relativelysmallnumberof composerswho in the late 1950ssoughtviable
alternativesto post-Webernianserialism,and it is no exaggerationto say that,
nowadays,whenasidefroma handfulof undisputedmasterpiecesmuchof what
waswrittenduringthe serialisteraseemshopelesslydated,Ligeti'smusicfrom
aboutthe sametime soundsas freshandoriginalas ever. His careeraffordsthe
music analyst the opportunityto study one composer'sturning away from
serialism:the nature of his objections, their implicationsfor the further
developmentof his technique,and the wayin whichthe methodshe arrivedat
achievea meaningfulorganizationof musicalmaterials.This articlediscusses
these issues and proposes analytic approachesdesigned to engage Ligeti's
solutionto problemsof musicalcompositionin whatturnedout to be the post-
serlalStera.
. .

Ligeti's emigrationbrought him into contact with a thriving European


(especiallyGerman)avant garde. Perhapsbecause he came to it later than
others, havinglived since the end of the war underconditionsof provincial,
state-imposedisolation, Ligeti responded to the stimulus of this activity
differentlyfromthewaymanyof his contemporaries haddone.Althoughhe had
arrivedwith scarcelyany knowledgeof twelve-notetechnique,let alone the
extensionof serialprinciplesto aspectsof musicalsoundotherthanpitch, not
even three yearshad passed before Ligeti was setting down his criticismsof
serial techniquesas they had come to be appliedto compositionduring the
1950s.' His previouslypublishedanalysisof Boulez'sStructures (PartIa), which
reflectedan extensivefamiliaritywith serial methods, demonstratedthat he
spokefroma well-informedposition.2
Ligeti'sdifficultieswith serialism,as expressedin the earlyDie Reikearticle
'Metamorphoses of MusicalForm'and in other, subsequentpublications,can
be succinctlysummarized.He found problematic'the organizationof all the
musicalelements'- thatis, pitch, duration,timbre,dynamics,modeof attack
- 'within a unified plan' because he 'detected within it a discrepancy:

MUSIC ANALYSIS 6:3, 1987 207


J()NA rHAN W. BERNARD

quantificationappliedequallywithinthe variousareasproduced,fromthe point


of view of our perceptionand understandingof musicalprocesses,radically
differentresults,so that therewas no guaranteethata singlebasicorderwould
produce analogous structures on the various levels of perception and
understanding'. Unity, therefore, existed only on the level of verbal
description, 'clapped on the musical events from the outside'.3 Even the
generalizationof serial procedures to engage more abstract and global
characteristics,such as types of motion, form, density and so forth, did not
tightenthe looseconnectionthathaddevelopedbetweencompositionalprocess
and the actual, resultantsound of the music. Pre-planninghad become so
importantthatit was the realcompositionalact.4
Evenin the treatmentof pitchalone,new musichadapparentlybegunto run
. . . *

up agalnst certaln .,lmlts:

The individualcharacterof the variousserialarrangements fadesas a result


of the superpositionof several horizontalseries, in which, wherever
possible, common notes occur at the same pitch. Such interweaving
obscuresthe singleserialthreads(especiallywhenallthe partsareplayedon
one instrument),andthe resultingintervalshavelittleor nothingto do with
the originalarrangement.Wheresucha procedureis coupledwith seriesof
durationsthe composer can hardly even retain an influence over the
intervals that are to result, let alone determine them. They follow
automaticallyfromthe type of procedure.In this way the pitchseriesloses
its lastremnantof function,paralysedby the emergingcomplex.5

The relegationof resultantsoundalmostto the statusof by-producthadled to


'decreasingsensitivityto intervals'and permeabilityof structures:'Structures
of differenttexturescanrunconcurrently,penetrateeachotherandevenmerge
into one anothercompletely. . . it is a matterof indifferencewhich intervals
coincidein the thickof the fray.' Smallwonder,then, thatcomposerswho had
adoptedserialmethodsdiscoveredthatit wasbecoming'increasinglydifficultto
achievecontrast'- theirmusicsufferingan inevitableflattening-out.6
Ligetidid givecertaincomposers,suchas BoulezandStockhausen,creditfor
havingtranscendedthe worstof thesedifficulties.Yet even they did not escape
completely unscathed: 'Although [their] works create the impression of
abundantcoherence,nonethelessthis coherence,arisingas it does apartfrom
the relationshipsestablishedduringthe compositionalprocess,is not freefrom
a certainqualityof "malgrelui".'7 WhatLigetiheardin muchof the musicbeing
writtenat that time was a preferencefor 'homogeneoussequencesof intervals,
particularlythe chromaticscale',with the resultthat 'the verticaldispositionof
this materialresults in a piling up of neighbouringtones. It is no longer
primarilythe intervalsthat constitutethe structurebut relationsof density,
distributionof registersand various displacementsin the building up and
breakingdownof the verticalcomplexes.'
These observationsseem to have broughtLigeti to a key realisation:if the

208 MUSIC ANALYSIS 6:3, 1987


I_IGETI'S PROBLEM, AND HIS SOLUTION

qualitieshe had notedwerein fact the truedeterminantsof auralshapein new


music, why not engage them directly, instead of through compositional
methods that could not control such qualities, except more or less
serendipitously?More than a decade after first expressinghis reservations,
Ligeti summed up his subsequentexperienceas a composerin a statement
which has almost the characterof a manifesto:'In workingout a notional
compositionalstructurethe decisivefactoris the extentto whichit canmakeits
effectdirectlyon the sensorylevelof musicalperception.'9
This statement,however,is less straightforward than it perhapsappearsto
be. Certainlyit should not be taken to mean that Ligeti felt he had learned
nothingfromhis exposureto serialismexceptwhatnotto do. He hasspoken,for
example,of havingabsorbedandappliedto his own worksuchaspectsof serial
thinking as 'the principle of selection and systemizationof elements and
procedures,as well as the principleof consistency:postulates,once decided
upon, should be carried through logically'.' Further, by making this
declarationLigetidoesnot categoricallyrejectmusicthatcannotbe heard,in all
its particulars,accordingto the way it is composed.One maynot actuallyhear
the row at everymoment(or even in somecases, as a literalseries,at all) in the
twelve-notecompositionsof Schoenbergand Webern,but this hardlynegates
the row'sstructuralimportanceor mitigatesthe worthof the music. 1Clearly
there are any numberof ways in which a 'notionalcompositionalstructure'
mightmakeitseffect,evenif restrictedto doingso directly.Knowingthatthis is
whatLigetiis afterdoesnot makemattersanyeasierfortheanalystof his music.
For considerthe following:
Technicallyspeaking,I have alwaysapproachedmusicaltexturethrough
part-writing.BothAtmospheresandLontanohavea densecanonicstructure.
But you cannotactuallyhearthe polyphony,the canon.You heara kindof
impenetrabletexture,somethinglike a very denselywovencobweb ....
The polyphonicstructuredoes not come through,you cannothear it, it
remainshiddenin a microscopic,underwaterworld,to us inaudible.12

How can we come to terms with this apparentdiscrepancybetweenwhat is


written and what is heard?What is the point of composingstrict canonic
structuresthatcannotbe perceivedas such?And if we do not hearthis 'micro-
polyphony',as Ligetitermsit, then whatdowe hear?Beforewe canattemptto
answerthesequestions,Ligeti'sideasabouthis musicmustbe exposedin some
detail.
Oneof the moststrikinggeneralfeaturesof Ligeti'sdescriptionsof his music,
both in his articlesand in his interviews,is his frequentrecourseto visual
analogies,especiallyones havingto do with space. Ligeti is underno illusions
abouttheultimatesignificanceof theseanalogies- he callsthe spaceof his pieces
'imaginary'and is carefulto distinguishthe sensein whichhis musicis spatial
from that in which Stockhausen'sGruppenis, for instance,or any otherwork
which involves literal dispersal of forces to different points within the

MUSIC ANALYSIS 6: 3, 1987 209


J()NATHAN W. BERNARD

performancespace- but still these expressionshave an inherentinterest.For


one thing, Ligeti seems susceptibleto visual and tactile parallelsto auditory
phenomena to a degree that approaches synaesthetic sensitivity: 'The
involuntarytranslationof optical and tactile impressionsinto acousticones
occursto me veryoften;I almostalwaysassociatesoundswithcolour,formand
consistency, and vice versa: form, colour and materialquality with every
acousticsensation."3HoweverLigeti's conditionmay have originated,it has
had an undeniableeffect on his compositionaldevelopment.There is, for
example, his famous childhood dream, the memory of which apparently
influencedthe compositionof Apparztions (1958-9) and probablysome later
worksas well. In this dream,Ligetifoundhimselfentangledin a giganticweb,
alongwithvariousinsectsandinanimateobjects,andbecamea captiveaudience
for the gradual transformation,through the insects' struggles to free
themselves,of this web 'universe'.14 It is temptingto speculatethatLigetimay
have been impelled by this intersensoryfacility to his particularchoice of
compositionalmethod, in which concentrationprimarilyupon 'conditionsof
the material' inevitably arouses 'associations with visual and tactile
sensations'.15
Ligeti'stendencyto 'spatialize',if thatwordcanbe used,alsoowessomething
to the state of new music at the time of his arrivalin Western Europe. In
'Metamorposesof MusicalForm'he notes 'the seemingconversionof temporal
relationsinto spatialones', as if a musicalcompositioncould in some waysbe
analogousto a painting.Underthese conditions,'the successionof eventsis a
mere expositionof somethingthat in its natureis simultaneous'.l6Actually,
whatseemsto happen,in Ligeti'sview, is thatthe time of a compositionevokes
space,andthatthe spatialanalogythus suggestedallowsthe composer(and,by
implication,the analyst)to traversethe structureof the music as if it were
presentall at once.l7 Furthermore,spatialmodelsof musicalstructureare of
particularinterestin the study of form, since the idea of 'form'in music is
essentiallyan abstractionfrom spatialconfigurations,from the proportionsof
objectsextendedin space. Musicalform, then, can be termedthe imaginary
spatializationof temporalprocesses.18
Not every spatialpossibilityis equally attractiveto Ligeti. His antipathy
toward'moment'form is revealing- for he feels that while mobilitymay be
inherentin musicalform, the form itselfcannotbe mobile. Momentform, he
says, is basedon a false analogyto visualart:for example,Calder'smobiles.l9
'Musicalmomentshavemeaningonly in that theypointto othermoments:not
the meaningsthemselves,but only the shifts and alterationsof meaning,are
comprehensible.'20 Instead of writing pieces the order of whose parts was
variable,it seemedto Ligeti to be 'muchmoreworthwhileto try andachievea
compositionaldesignof theprocessof change'.2lForthisreason,perhaps,Ligeti
has adopted,like Varese,the analogyto crystallisationin orderto describenot
only the finished productof his compositionalprocess22but also, somewhat
inconsistently,the process of eternal 'becoming'exhibited by a piece as it
progresses:

210 MUSIC ANALYSIS 6:3, 1987


I IC,ETI'S PROBI EM, AND HIS SOLUTION

The technicalprocess of compositionis like letting a crystalform in a


supersaturatedsolution.The crystalis potentiallytherein the solutionbut
becomesvisibleonlyat themomentof crystallisation.In muchthesameway
you could say that there is [in my music] a state of supersaturated
polyphony,withall the 'crystalculture'in it, but you cannotdiscernit. My
aim was to arrestthe process,to fix the supersaturated solutionjust at the
moment before crystallisation.23

The analogyto crystallisationexpressessomethingelse as well: namely,the


unidirectionalnature of the form-creatingprocess. For instance, Ligeti
characterizeshis electronicpieceArtikulation as 'a gradual,irreversibleprocess
from the heterogeneousdispositionat the beginningto the completemixture
and interpenetration of the contrastedcharactersat the end'.24This is entirely
consistentwith the atmosphereof his childhooddream,in whichthe changesin
the web 'seemedlike an irreversibleprocess, neverreturningto earlierstates
again.An indescribablesadnesshung over these shiftingformsand structure,
the hopelessness of passing time and the melancholyof unalterablepast
events.'25The allusionsto crystallinestructuresuggest also the presenceof
regular,even symmetricalpatternsin the music, aboutwhichI shallhavemore
to say shortly.
Finally,amonggeneralaspectsof Ligeti'smusicaloutlookhis firmgrounding
in the historyof Westerncompositionalpracticeshould not go unremarked
upon. This has obvious relevance to the elaborate canonic procedures
mentionedearlier;and in fact Ligeti has allowedthat he was 'very good at
counterpoint'as a student.26His educationalbackgroundplacedconsiderable
emphasis upon traditionalinstruction, and as a pedagoguehimself Ligeti
remainsconvincedthat even if the old techniquescannotbe used directlyby
contemporarycomposers,the studentcan neverthelesslearnthroughthem to
think logicallyin the musicalsense. Keeping moderncompositionup to the
standardsof thepastcannotbe a matterof indifferenceto anycomposertoday.27
The specificconsequencesof Ligeti'sattitudesand opinionsaboutmusical
structurearebestexaminedin the contextof analyticillustrations.We mayturn
firstto his FirstStringQuartet(1953-4),a workwhich, like mostof whatLigeti
wrote beforeleavingHungary,bearssignificantmarksof Bartok'sinfluence.
Among these is a penchantfor symmetricalconstruction- more precisely,
registrallyconsistentsymmetricalconstruction.Example1 is a reductionof bs
521-33,fromroughlythe midpointof the work.Excludingthe openfifthsin the
cello, the contents of each chord are mirror-symmetrical;the stacks of
bracketed numbers below the music show the adjacent intervallic
arrangements.Furthermore,the passageas a whole has a symmetricaldesign,
in two senses. First, temporallyspeaking,the seriesof fourchordsbeginsand
ends with the sameverticalarrangement(thoughnot the samepitches),andthe
two chordsin the middle sharea differentverticalarrangement.Second,the
verticalizedaggregateof all pitchesin the passage- includingthe cello- yields

MUSIC ANALYSIS 6:3, 1987 211


JONATHAN W. BERNARD

thearrangement shownto therightof Ex. 1. Thisis notmirror-symmetricalin


allits details,butit doeshavean overallsymmetry embeddedin it, as shown.
Thepitchesthatserveastheboundaries ofintervals
inthisembedded symmetry
areencircledforclarity.Theyincludethelowestfifthof thecelloandthelowest
pitchof thesymmetrical portionof eachof thefourindividual
chords.28
Ex. 1 FirstStringQuartet,bs 521-33

17

[3W
181

I)

03W [eY}

FF 141

I)
141 171 171 141 (
141

1-31 lll lll 1-31


11 171 [71 [1l
13S
1'31 lll lll 1-31 (i
141 171 171 141
1'81

71
3

Apparitions, Ligeti'snextmajorworkaftertheFirstStringQuartet,wasalso
the firstworkhe completedafterhis emigration,apartfromtwo electronic
pieces.Havingobservedthelossof sensitivity
tointervals
in serialmusic,Ligeti
decidedto seewhatcouldbe doneif thisnewlyevolvedconditionweretakenas
a givenand, in fact, exaggeratedby dispensingaltogetherwith intervalsas
structural components:
I composedsoundwebsof suchdensitythatthe individualintervalswithin
themlost theiridentityandfunctionedsimplyas collectiveintervalgroups
... this meantthatpitchfunctionhadalsobeeneliminated.... Pitchesand
intervalsnow had a purelyglobalfunctionas aspectsof compassand note
density.29

Thismaximized densitytooktheformof chromatically


filledspaces:'I inserted
so manyminorsecondsthat even the minorseconds,the chromaticism,
disappeared
in theharmonic sense.'30

212 MU S IC ANA LYS I S 6:3, 1987


LIGETI'S PROBLEM, AND HIS SOI UTION

Ligeti'sownremarksaboutApparitionsemphasizetheideaof transformation
fromonesoundgrouptothenextthroughwhathecalls'acontinuous reciprocal
betweenstatesandevents'.He continues:
relationship
The statesarebrokenup by suddenlyemergingeventsandaretransformed
undertheirinfluence,and vice versa:the alteredstatesalsohavea certain
effect upon the type of events, for these must be of ever new character,in
orderto be ablefurtherto transformthetransformedstate.In thiswayarises
an unceasingdevelopment:states and events, once they have occurred,
reciprocallyexcludetheirrepetition,thus areirretrievable.3l

Theseformalideasareevidentlycombinedwithfixedrepertories withinthe
variousdomainsof musicalsound; Ligeti has discussed the repertory of
durationsin somedetail,comparinghis employmentof it to a typesetter's
selectionof lettersfroma typebox,andhasmentioned others.32 In nocasehas
he enumerated the contents,but it is possibleto speculateanalytically about
whattheseareand,moreimportant,abouthowtheyareconnected- thatis,
how the transformations takeplace.The followinganalysisfocusesuponthe
interrelated domainsof pitch, compass(verticalspan)and note densityto
explorethefirstpartof thefirstmovement.33
In Ex.2 thescoreof bs 1-23is transcribed intogridnotation,whichprovides
a uniformsemitonalcalibration alongthe verticalaxis.(Numbersin the left
marginmarklocationsof C, with C4 corresponding to middleC. Numbers
alongthetopedgearebarnumbers.)Thegridmaybeanespecially appropriate
analytictoolin Ligeti'scase,sinceit is knownthatin theinitialcompositional
stageshe usesa kindof graphicnotation.34
The openingbarsshowa development of pitch/registral spacethrougha
strictlycontrolledgroupof intervals,severalof themdetermined by the total
numberof available partsin eachof thedivisionsof stringinstruments.35 The
initialminorsecond,(D#-E)1,is superseded locallyby a perfectfifth(spanof
[7])[13]above.Thenexttwoeventsincorporate allthreeof theseintervals - [1],
[7]and[13]- andbringin newonesasdirectresultants of spatialmanipulation
of theoriginalgroup.Theevent(C#= E)1 (b.8) doesnotatfirstappearto be so
related,butwiththelowerboundary of F#1=G2(bs9-11)it standsasonearm
of a symmetrical expansionfrom(D#-E)1.36 A by-product, as it were,of this
expansion is theinterval(C#-F#)1, or[5],whichnowappears asacomponent of
F#1=G2, combinedwith [1] and [7]. Further,the interval[5] describesthe
distancefromupperboundaryto upperboundary,C3 to G2. The cluster
F#1=G2is also[13]in total verticalextent,whichmeansthatthe intervalof
expansion[2]notedearlieris anintervalof projection aswell:thusF#1=G2is a
projection andfilling-inof E1-F2.Thesetwo [13]'stakentogetheryield[15],
readeitheras[13][2]or[2][13],whichis nextimmediately statedin twoforms:
oneas thetotalextentof theclusterin b.13, E2=G3;theotheras theinterval
frompreviousto newlow boundarypitch,C#1-E2.Note thattheuppermost
[15]is expressedas [7][1][7].

MUSIC ANALYSIS 6:3, 1987 213


--
*--
s-!+_ 1--| -t - -trit13
0 ------------
--vw - + t- - t--15 /t t r- tX4---
-- +
* --0-t t+t--- -- _t0 --+r-
--
St - i-t---
1L L lt-0 +00 I0
- It t|---- l-; _0q'
t| 1172' _- 0 -- ----t
t ; + ;_70
t151
--Z!-t----171--;
'' _ Tt--
|Ih
- + -
s - ----el
17
---,
;- 1--1I -'- | -*t119
;;0-0------
-L--;-t -: ---i I0-t-0-
-- > t -- t--+- T+t z! -:-"r_ftlLin
;--;
-L,
-+ | F; ,j-t'1-l
*b -
- t: 4-t
t - b t J ;--
-1\i i 't-
28 | __ -l I+
[N. - j;__
+ '--t';
t|15}
-- -+
-_ t -- --- "
__''-.t *;+
t_1-; +tA'._0
;-5,- ' W4- -_--5-St--- 1,1!
-t-t
}t----t
+;- i l !;
.4-t __ ' - T
_ --tt s -A:
----t---
t4 q-tF
-|)3;
._t t>-t
--+--
+ ___ - t.lt---
+bt i . t1,- _ t ,.!S r-
t + +

JONATHAN W. B ERNARD

Ex. 2 Appantions,I, bs 1-23

0_ t = . _ _ 4 _ _ _ , _ , t t { ; B i 4 _ -t- t

-r- t { l _ lC t1l 1 '+ t ! --


- + 1 1 t t, t ! j -"--t ff + +
_+,_e --t-- i i trl jlt < if |

0 } - - - -- , !/i- + -+- r - - - -8A 4 t t I < A

,, I i

l ^ + l i l
e -t0+ ----; l
B 1ll---t I i ! 5 t; I --, ; ;--
o-s L-- l 4 l l
t i i i
_ I'
i ' t; Z }_-t v 4 t
4i __,_i 1 - --* t t-- 1-- * ,6 I | I I I
--+r_-? t ^; (.e-t
tt t -
- t

S-+> ----

- ; -s-T - w - -- A - -

+_ -- t t t z1+ , tr-4 04 t | t i _s ; sE-r, 1t


- - -- @ |; )-tJ171 j

- J0 lt
--t - t I ' t -4 z---+--t-l-t-4--;
t m- - ^1
;
: (# j f --lIpi _ rw- t- t s ( t X t t
X + ; T -
t j t -i-4t --_' L-
+ +
... ,.
s _ + .1 s
t + ---t t t -,; - t t - t i t t ;-'

t t t ' t -- 1-l

- t - - - -t --
4 +

-. _ __ il 5 , AW 1 1# t i 1g; ._
+ + } X
-- - --+t- - S - -s-t- - t - 6 .
+ + XLo - - -tF r > L
t 8 Z I , , . . ,
I -1 t- -

0 t t ' + t

' 1)g
,\ 1 /+, . _ + z _ . t t w _

t
^' I tO
* I I T I 4 |

- 4

t--- i t + - + t t t 1, * + t; .

- TX __AX_-I i i : i :0 i ] 0 I . 1 1 1 1 f 1 . A ' ' '

Tl)tal span of sntervals (wherc relevant):

[1l 171 1Al 1131 llsl 151 11()1 151 [131 I 1 11 1'1 1 1r9' ' ''1

Intervalsthat are presentin essentiallyunrealizedpotentialin this complex


are [11]: FW1-F2 and [10]: FW1-E2. The formeris broughtinto play in b. 15
(bassoons), in the chord Ab1-G2-F#3. Generallyspeaking, this adjacent
doublingof intervalsize seemsto proceedfromthe ideaof doublingpreviously
expressedas [7][7] - F2-(C-G)3, then as [7][1][7] in (E-B)2-(C-G)3- and
[15][15],C#1-E2-G3.It is interesting,thoughnot necessarilysignif1cant,that
[11] is also the averageof [7] and [15]. The occurrenceof [5], as a filled-in
interval,in b. 16fallswithinthe spanof the lower[11]anddividesit [5][5][1];in
conjunctionwith the f1lled-in[10], D# 1=C#2, thatfollows(in the piano,b. 17)
it also expresses [11], which can now be regardedas a replica of Ab1-G2,
transposeddown [S] and now completely f1lledchromatically.The cluster
(D=G)2 in b.19 bears a relationto all three previousevents, both in that it
replicatesthe [S] A1=D2 and in that, restoringas it does G2, it brings[5] and

214 MUSIC ANALYSIS 6:3, 1987


of twopreviouschords:[5][1][7] and[7][1][7] [5][1][7][1][7]Just as [5][1][7]

I IGETI'S PROBLEM, AND HIS SOLUTION

[11]into the samespatialrelationshipthatthey occupiedin the juxtapositionof


eventsin bs 15-16.
The clusterin the stringsin b.19 is the firsteventin the piece to enterbefore
the previouseventhasended(discountingmomentaryoverlaps,as for example
in b.16). Here[13]is expressedas [5][1][7],as in bs 9-11;theoverlapof(D=G)2
with the uppermostcomponent,(D= A)2, expresses[7] as [5][2],asoccurredin
the relationshipof the clusterin bs 9-11 to eventsof earlierbars,but the means
in b.19 is different,with the spansof [5] and [7] being given simultaneously.
Furthermore, the [5] between successive lower boundaries, (D# -Ab) 1,
replicatesthe [5] (C#-F#)1 between successivelower boundariesin bs 8-9,
which means that the relationshipbetweenthe compositespans C#1-G2and
F# 1-A2is one of transposedreplication,by [2].
The entranceof the celestain b.21, on (C= B)4, bringsin the filled-ininterval
[11] at a distanceof [15] abovethe immediatelypreviousevent. The following
chord in the strings (bs 22-3) is the largest simultaneouslysounding
chromatically-filled spanheardso farin the piece, andcombinesthe structures
in bs 9-11wastransposedup [2] to its positionin b.19, so [7][1][7]in b.13 is now
transposedup [2]to its positionas a componentof thechordin bs 22-3. The total
spanof [21]duplicatesthatof the compositeof the firsttwo eventsin the piece:
D# 1-C3.
The last eventsin the grapharenot analysedin full, for they seem to belong
moreto the next partof the piece;but it is possibleto find, in theiroverlapwith
the preceding [21], the intervals [28] and [27] now emerging as filled-in
compositespansaftertheirmoreobliquestatementsas Ab1-C4(bs 19-21)and
F# 3-D# 1 (bs 15-17)respectively.
In retrospect,we canidentifya seriesof phasesthroughwhichalltheintervals
employedso farpass:
1) indirector obliquestatement;
2) successiveloweror upperboundariesof clusters,or spacebetweenclusters,
or compositespace;
3) compositespace,filledchromatically;
4) simultaneousstatement,filledchromatically;
5) incorporationas segmentinto largercluster.
Next, we can detailthe actualphasesthroughwhicheachintervalparticipates
ln t llS SerleS:
. . .

MUSIC ANALYSIS 6:3, 1987 215


J()NATHAN W. BERNARD

[1]: (4); (5)


[2]: (2); (5)
[5]: (2); (5)
[7]: (4)i (5)
[10]:(2); (4)
[11]:(1); (2); (3)i (4)
[13]:(2); (4)i (5)
[15]:(1); (2); (4); (5)
[21]:(2); (4); (5) (b.26)
It is evidentfromthe abovethatno two trajectoriesareidentical,andthatnone
of the intervalspassesthroughall five phases. With the exceptionof the last
phase, the series seems to representa process of gradualemphasis, as the
employmentof each intervalas a space-definingentity is made progressively
moreexplicit. The identificationof the fifth phase- incorporationinto larger
clusters- suggeststhat the developmentof spatialrelationsthroughvarious
intervalsizes in this piece does not proceedsimplyby meansof bringingeach
intervalinto prominenceand then maintainingit in that status.Once brought
to anexplicitlyspace-definingrole, an intervalcansubsequentlybe employedin
anyof its variousotherrolesand, as the piecegoes on, canbe absorbedinto the
textureto becomea componentin largersonorities,perhapseven sufferinga
temporaryor permanentcancellationas an explicitentity.
Conversely,as Ligeti'sdescriptionof the formsuggests,new intervalsmust
also arise. In bs 25-6, for example(not shown in the graph),the interval[8]
emergesexplicitlyfromthe lowertwo adjacentintervalsin the chord[7][1][7]:
F#2=A3. The boundarynotesin the cello/violachordin b.25 areF#2, C#3, D3
and A3; then, in b.26, F#2 and D3 becomethe boundariesof a separatechord
in the winds.A little furtheron (b.29), [8] becomesa segmentof a largerchord,
markingthe contrabasses'portion,(D# = B)1, of D# 1= Eb3. Phases(2), (4) and
(5) arethus represented.
In Atmospheres (1961) Ligeti is still working with chromaticallyfilled
complexesof sound, but the idea of a repertory - of durationsor intervals,for
instance- has been discarded.'Rhythm',as Ligeti has said of this work, 'is
completelyeliminated,[and] the absorptionof individualshapes into static
planesis accomplishedto the greatestpossibleextent.'37Atmospheres is widely
reputed, not without reason, as Ligeti's klangfarben piece, but Ligeti has
effectivelycautionedthe analystby sayingthat 'it is a rathersuperficialview to
lay too much emphasison timbre'in this workor otherworksof his, and that
in Atmospheres 'modificationsof timbre and dynamics are obviously very
significantbut the patternsemergingfromthemareeven moreimportant'.38 It
would seem, then, that the 'iridescence',as Ligeti calls it, causedby minute,
continualshiftsin doublingandin the locationof gapsin the chromaticfilling,
and by dynamicchanges,bowingchangesand so on servesto characterizeand
differentiatethe various'staticplanes'andimpelsmovementfromone planeto
the next. With this in mind, it is possibleto analyseAtmospheres throughthe

216 MUSIC ANALYSIS 6:3, 1987


LIGETI'S PROBLEM, AND HIS SOLUTION

samesortof gridnotationemployedforApparttions. A graphof bs 1-29appears


in Ex. 3.39
In b.1, the strings enter in three slightly separatedgroups:contrabasses,
(Eb=Bb)2; cellos/violas,C3=G4; violins, Bb4=C#7. Overlappingthe gap
betweenviolasandviolinsarethewinds:Ab3-CS,plusD2 in thecontrabassoon.
In the courseof bs 1-8 the winds'planeundergoeschanges,as severalpitches
dropout; the consecutivestagesaredisplayedin theexample.This reductionin
density,takingplaceas it does againsta backdropof chromaticallyfilled space
in the strings,causesthe windsalmostto disappear,as if theirplanehadmerged
with that of the strings. In particular,the departureof the flutes after b.4
changesthe winds' upperboundaryto B4 and preparesfor the assumptionof
this pitch as the cello/violaupperboundaryin b.9. At this point, all previous
structuresaresupersededand the cellosandviolassoundaloneuntil the restof
the ensemblerejoinsthemin b. 13. In the courseof bs 9-13, a rippleof dynamic
changepassesthroughthe clusterof cellos and violas- or rathertwo ripples,
movingin oppositedirections,as pairsof partscrescendo, oneby one, toforteand
then drop back to the ensemble pianississimo. The succession of dynamic
emergencescreatesthe symmetricaldesignshownin the graph.40
The stringsreach their new locationby transpositionof all three groups,
althoughnot in parallel:the contrabassesmove up [5], the cellos/violasup [4],
the violinsup [3]. A gap thus remainsbetweenthe uppertwo groups,but the
lower two have meshed completely,and a new regularityhas developed:th
segmentencompassinglowandmiddlestrings,Ab2= B4, correspondsprecisely
to the size of the violins' segment, DbS=E7. During bs 13-22a new sort of
dynamicfluctuationtakes place in the strings, dividing the contents of the
cluster into two interlaced collections on separate crescendo-diminuendo
schedules.Thesemomentarilyemergingverticalarrangements exhibitthesame
regularitiesand mutualcomplementarityas do the blackand white notes on a
pianokeyboard.
Meanwhile,somethingelse has happenedto the winds, whichnow occupya
space(notcompletelychromaticallyfilled)boundedby Gb3 andC#6, plus Eb1
in the tuba, joined at b.15 by E1 in the contrabassoon.From their original
locationat D2-CS,they haveundergonean expansionthataddstwo octavesto
their compass. This will be called an '[11]-[13] expansion', after the two
intervalsthat define the intervalsof change at lower and upper boundaries
respectively.The winds'fluctuationsat this pointaremorecomplexthanthose
of the stringsandfor the sakeof legibilityareshownin a separateexample(see
Ex. 4). As in the strings,crescendi
successivelyemphasizewhite-noteandblack-
note collections,but unlike the strings the winds projectstaggereddynamic
patterns; and besides dynamic change another form of 'disturbance'is
introducedinto the texture:rearticulationof individualnotes, shown by the
placementof crosses.As in bs 1-8, the windsin bs 18-22die awayby stagesand
soundas thoughthey havebeen absorbedinto the strings.
From b.22 to b.23 the stringscontracta totalof two octavesin range,from
Ab2-E7to G3-Eb6. This canbe calledan '[11]-[13]contraction',complementing

MUSIC ANAI YSIS 6:3, 1987 217


JONATHAN W. BERNARD

bs 1-29
Ex. 3 Atmospheres,

218 MUSIC ANALYSIS 6:3, 1987


s _ c x X - x cc E GS r rlr
D; ]

LIGETI'S PROBLEM, AND HIS SOLUTION

Ex. 4 Atmospheres,
bs 13-20,winds

13 15 16 IX 2()
_ . -- 1--
6- Cg _ _ __s__ X D1 DS DS
B B
- - x x BS BS BS

x
A >- r R
x AS AS AS
-- G _ 1-
- x X GS GS GS

x ES ES ES
- - x D
-
D
x
]
DS DS
r DS s

BS x BS BS BS
x A A F l l
-- x AS AS AS
x - G G |-/

x F F |
x E ]
x ES
x D D |
4 X C I
x B B t
x- x BS
x A |
AS x Aw
l l
GS x Gw

E E
I

ES
l

MUSIC ANALYSIS 6:3, 1987


219
JONATHAN W. BERNARV

the previousexpansionin the winds. The new spannow servesas the point of
departurefor a greatdownwardsweep, from D6, the upperboundaryof this
plane from b.24 on, to F2. The textural change as this happens is quite
dramatic:sulponticello, moltovibrato playingin everfasterfigurationalternating
in each partbetweentwo pitchestakesover from sultastosustainingof single
pitchesas the downwardsweepprogresses.4lJustbeforethe bottomis reached,
two moreplanesappear(b.25), flutesandclarinetsnestledsymmetrically within
the compassof the contrabassesplus threecellos. This smallerstringgroupis
left soundingalonein b.29 afterall otherinstrumentssuddenlyexit.
The graph displaysthe strikingsymmetriesoutlinedby the successionof
planarstatesfromb.13, wherethecompositewinds'/strings'rangeis Eb1-E7,to
the sweepof bs 23-9, whichrepresentsa symmetriccontractionfromthe outer
boundariesof that range, to the furthersymmetriccontractionto the small
stringgroupleft soundingin b.29.
The organizationof volumes of sound of varying density accordingto
schemesbaseduponverticalspanandsymmetricalconsiderationshasremained
a prominentfeatureof Ligeti's work since Atmospheres. Beginningwith the
secondmovementof Apparttions and continuing,with increasingexplicitness,
in subsequentworks, Ligeti integratedthis basic 'spatialconsciousness'with
other compositionalpreoccupations,notablythe high regardfor the riforous
contrapuntalproceduresof oldermusicthathe hadacquiredasa student. 2 Two
relativelybriefexamplesfromworksof the 1970swill serveto illustratetypesof
techniquesinspired by (if ratherremotelyrelated to) these procedures.In
'Bewegung',the third of the Three Pieces for Two Pianos(1976), the closing
section is basedon a rigorouspitch symmetry,canonicallyunfolded(see Ex.
5a). Each strandof the canonconsideredseparatelyis a pairof voices moving
note-against-notestrictly in mirror fashion, so that the intervals formed
verticallyexpand and contractsymmetrically.43 In bs 49-52 two such pairs
constitutethe entire texture, apartfrom rapid figuration (not shown in the
example)overlappingfrom the previous section and slowly fadingawayhere,
producing a double canon with the comesentering [9] belowthe dux.(This can,
of course, just as be
accurately described as a double canon in inversion.)In bs
52-8 the doublecanonis itself doubledto becomea quadruplecanon,with the
overallaxis of symmetry,A4, preservedfrom the precedingbars. Here the
originalpairof pairsbecomesthe centreof the entirestructureas two new pairs
areadded, [6] aboveand [6] belowrespectively.In bs 57-8 the canonbeginsto
dissolve,as the two upperpairsentertogether,followedone chordlaterby the
lower two pairs. During the final two chords of the piece (bs 59-63) the
symmetricallayoutis expressedentirelyin simultaneities.
In thereductionof Ex.5a, doublebarlinesmarkthepointsatwhichtheoneor
morecomes 'catchup' with the dux.This happensonceduringthe doublecanon
(b.50), once justbeforethe quadruplecanonbegins(b.52)andagainat bs 55,57
and 58. At these five points all voices resolve, as it were, into a mirror-
symmetricalchord,to whichthelasttwochords,alsomirror-symmetrical, serve
as a kind of culmination.The catch-upchordscan be heardas pointsof phrase

220 MUSIC ANALYSIS 6:3, 1987


S
" :r- " h['
Bl--b' - W

LIGETI'S PROBLEM, AND HIS SOLUTION

Ex. 5a 'Bewegung'(3 Piecesfor 2 Pianos,III), bs 49-63

v # -
- +
Su -
B
b 4 - 4 ^ j -

$ , b, .

4 (') b ^
< -s - bs
(We) _ +
b- s
i , . _

') b- - i s

articulation,with the shift to the thickertextureof the quadruplecanonandthe


finalabandonmentof canonicprojection(b.58) the principalpointsamongthe
five. In betweenit maywell be impossibleto hearthe canonicdesign, sincethe
ensembleof two pianosis exploited,hereas in muchof the restof the work, to
producea unified(as opposedto stratified)texture.(Whatis not apparentfrom

MUSIC ANALYSIS 6:3, 1987


221
J(:)NATHAN W. BERNARD

Ex. 5a is that none of the four canonicpairsis confinedto a single line in the
score;the two piano partscontinuallyexchangepitches at irregulartemporal
intervals.)It shouldbe possible,however,to hearsomethingelse: see Ex. 5b.
Here the contents of Ex. 5a have been transcribedinto grid notation. This
notation providesa more accurateidea of the spatial quality of the music,
enablingthe listener/analystto see it whole - as if, as Ligeti put it, the whole
passagewerepresentall at once. Fromthis graphit is evidentthat, despitethe
timelag betweenentrancesof partsenforcedby the canonicplan,one canreally
perceivethe generaloutlinethat emergesas a resultantof the canon, and that,
despite the fact that the internal details are slightly askew, the external
dimensionsstandout clearly.44
In the third movement of the Kammerkonzert (1969-70), we encounter
anotherapplicationof canonicprocedure,in a settingradicallydifferentfrom
that of the previous example. Bars 1-12 present one of Ligeti's meccanico
textures,a devicewhichis one of his principalagentsforiget[ting]ridof rhythm
as a conceptaltogether'.45In thispassage,aniniiialpitchis joined,successively,
by othersin closespatialproximity,forminga slowly,steadilyexpandingwebof
sound. The resultantof the canon can be interpretedin terms of gradually
shiftingsymmetricalrelationships.Example6 displaysthe overlappingstagesof
development.The first stage, at (a), surroundsinitialpitch E4 symmetrically,
then by addingC,"4 leavesE4 no longerquiteat the centre.In (b), at b.5 a new
groupof instrumentsbeginsthe canonoveragain;this time the eighthnote in
the canonicseries, C4, makesits appearance,pulling the centreeven further
downward.Withthe entranceof the f1rstsustainedpitch, in b.7, the thirdstage
begins:see (c). This eventuallytakesthe formof a 'whole-tone'divisionof the
spaceopenedup by the canon,as C#4, ES4 andF4 arecompletelysuppressed.
With respectto the originalaxis and initialpitch E4, spacehas been expanded
by [2] aboveand [4] below, a conditionwhichis neatlysummedup by the last
three sustainedpitches to enter - F#4, E4, and C4 respectively- and which
effectively'predicts'whathappensnext, shownin (d): a new meccanico section
beginsat b. 12in octaveASs. The twocentralASs, in octaves3 and4, arereached
by motion[4] belowC4 and [2] aboveF#4 respectively;the outerAS2 andAS5
simplyexpandoccupiedspaceoutwardfromthis centralarrangement.
Two worksof the 1960s,Lux aeterna(1966)for sixteen-parta cappellachoir
and Lontano(1967)for orchestrawithoutpercussion,are especiallyrewarding
subjects for study of Ligeti's canonic technique, principallybecause they
representa deliberateattempton the composer'spartto exertmorecontrolthan
previouslyoverthe spectrumof relativeclarity(transparency) to relaiiveopacity
of texture.LigetiregardsLuxaeterna,in retrospect,asa turningpointforhimof
comparableimportanceto Apparitions.46 In a briefessayentitled'Aufdem Weg
zu Lux aeterna'he notes that the work, which he composed directly after
f1nishingthe Requiembut whichwas conceivedas a separateworkeven though
its textis takenfromthe traditionalRequiemmass,wasdeliberatelydesignedto
havemorelimitedpossibilitiesfor opacity;its ensemblehasonly sixteenchoral
parts to the Requiem'stwenty and omits orchestralforces. Ligeti identif1es

222 MUSIC ANALYSIS 6:3, 1987


X .......
- .**I
......
.........
....... , 4
........... ,.I''I .......
.j.Iv ,,,j
,, , j.,,
. .,l, I *.... i +,,,,,,
. d..........................
.......................................... + v SI S ........
s+ ,s ., I''t'' ., !s If .. ..' ....
, '. L l ' ,.,,,,,,,,.,,,,,,,,;,,j,
............ ' .;., ' ' ,, '
.............................
f . .* In
-t , i t.....
\ s,
....... , .W. *...* -I c
.............................. I - ,' s I.I _-
!- _

JONATHAN W. BERNARD

Ex. 6 Kammerkonzert,
III, bs 1-12

4 _ __
[e 0 ---,- *-; --$[ -- -

(t.issossit %,,-.,,,.1,

(st) (d)

(h) b*- l

- - -
I
- " -

d -_ -

_19 -

t. .
141 -t

Ex. 7 Luxaeterna, bs 1-37

6
, . i . . l * ! ' I
l Xl l ! ' ' * * * . * ..................... ,

b l t . , . . . . . . . . . I . . . . . .

! '

8 ., . l + * w w w w w b b w w w w w w

I ,.,. l a b w w w w w + + g + w

Z . . . l
5
. ,.. l * fi + w w + w* ww6 * ww w k .... .................
.
., ...
t
I
! + ... ... j ' j' ' ' ' ' ' ;;
! 8 ' e ' *r ** g - , . . . _

*,,,,,1 :1T_ __=_


' ; t ; t' ' 1 1b' ' ' ' ' ' !' j _* t _

4 , . i * * ! z , . . . . . . . . , . . . . . . . . . s I . . , . + ; I;
. , . , . . . . . . , . . . , . . . . . . . . . . . . , I , . . . . i . . . . . . . . . . . i . . . . . . . . . . . . . . . . . . . .

t i!
..... ,. fI f j tbi} f ! . . .
t t * t * - t -
. . , . . . . . , . , , - . , . . . . . . . . . , . . . . . . fl . . . . . . . a . . . . .

:: ',;: . :, : I,;;::::::::: 0: ': :: .:: ':,:,:: !:: . . ':: !: .:::: 1;;:


t ;21 31 44 5 jh ja K i9 If 1l 112 1ll 1|4 1lS |Ih jl7 lIN 19 i20 ;

224 MUSIC ANALYSIS 6:3, 1987


+t ........
1+*;!.......
, ' '' - ' '.;'4
..... -4' '. '*' ,........
| .........
' ' '............
| * , ''' '.,r'' j' ''1*\
,*(|' t ..... '.' ......
,,,,,,' ' * '
t X. ....'\'..'- '.
'*.' ' ' ;&. '' ' *'
b . * j......
'' . ..
j......................
...... * j
' ' '+'! \ '' @j+' '* 'i 1*. t * '4' * 'i
* !**'.......
*' 4s' '.' * ,s *' +
.......... . ' *
4 +' .,, !*'
Isj . *. ' -,4 ........
.'' 'i t '. ,,T '.....
......................................................
. v
'.. 4 *' .... 4'
. *. t .... ,6,.,.,
...' ......
....' . I ..' . * ' .. . ...' .. .' .. . 4'_.. ......
6 ' - .. -' .. * 6 '.

LIGETI'S PROBLEM, AND HIS SOI_UTION

gradualtransformation as a principleof harmonicconstruction andstatesthat


thecounterpoint hasthefunctionof destroying oldstructures
andbuildingnew
ones.47To trackand interpretthis process,graphicnotationof a slightly
differentvarietyfromthatof previousexampleswill be useful.In Ex. 7 the
durationscale(horizontal axis)is consistent:one squareequalsone beatin
commontime,at a tempoof approximately crotchet= 56. Pitchnameshave
been dispensedwith;instead,the shadedareascorrespondto durationsof
individualpitches,projectedandprolongedby overlapped entriesin multiple
parts.In this,as in manysubsequent works,Ligetihasexplicitlydirectedthe
performers to maketheirentrancesas gently,evenimperceptibly, as possible
duringsustainedpassages.Thelisteneris evidentlynotintendedto focusupon
attackpoints;thusthegraphsimplyshowswhereeachpitchispresentinatleast
onepart.
Theorderof entrances of thefirstfournotes,(F-E-G-F,"
)4, canbedescribed
as two interlocking three-notegroupswith verticalintervallicarrangements
[1][2]and[2][1](a andb);the symmetrical arrangement hereensuresthatthe

Ex. 7 cont.

* * . . . . . K . . . . . . . . . . . . . . . . . . . . . .

... . .... I ...............

L . . . . . . . . . . . . .

_- _
! j:::;: [ fU! : ::::: ]
, . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 ' ' I i ' ! j ** * ** * . *

. . . . . . . . , , , , , j , j , j

, @b P 4 ! b ,; * | * - * b $ s b *- - ! ' ' ' l ' l ' F ' l '; l 6* +; t *t - - *- 4*+ - - - -


i # ' i j E ! ' * * ! } t I **i t * , j * 1: 1 j, ! . .. . ,
; 1e; ,;,; i; t i i' j . i + t ; t 1 + '; ' I ' t' + i t ' ' +; ' +; ' ' 4 ' [ j ',; ' I ' j t ;;;;

211
i| 221
i '231t [ |241t tt|2.sitt 1261llll2,'I i | 4l281 | !| | I f | | s 4t ttj] tt if ls ;| | I| | I| ii| i i;M
129 30 13 32 33 134 35 136 137 38 139

MUSIC ANALYSI S 6:3, 1987 225


J()NA I>HAN W. BF RNARD

availablesemitonalspaceis completelyfilled. As in the previousexample,from


the Kammerkonzert, the semitonalweb spreadingout from an initialpitch has
grownslightlyasymmetricallywith respectto that pitch. By and largethis is
typicalof Ligeti'spractice:here, as the dimensionsof occupiedspacecontinue
to spread outward, the general idea of a symmetricalcentre is effectively
conveyed,but the actual,specificlocationof thatcentreis constantlyin flux.
Pitch E4 dropsout in b.7, leavingthe cluster(F = G)4, [2] in extent(c). As
if in responseto this, Eb4 enters(b.7), then Db4 (b.9), markingoff striationsof
the semitonallycalibratedspace[2] apart(d). (Henceforth,in this analysisand
the one of the openingof Lontanothatfollows,such formationswill be referred
to as '[2]-striations'.)Meanwhile,in b.8, Ab4 enters (e), at first seeming to
functionas an extensionof the semitonalcluster(F =G)4. However,this effect
is only momentary,for F#4 soon exits, leaving(F-G-Ab)4, or [2][1], as the top
portionof the structure(p. PitchF#4 reentersin b. 11, againseemingto reassert
chromaticallyfilled space (g); but almostimmediatelyG4 drops out, leaving
[1][2]: (F-F#-Ab)4 at the top of occupied space (h). Ahe effect is of F#4
replacingG4, with a bit of overlap;the resultantis a duplicationof the [1][2]/
[2][1]relationshipthatopenedthe piece, one semitonehigherandwith greater
temporalseparationbetweenthe components.
The entranceof Bb4 in b.12 (i) comes in apparentreflectionof the [2]-
striationsbelow. In fact, this impressionis strengthenedby the subsequentC5
in b.13 (j); in between, however, the appearanceof A4 (bs 12-13) briefly
reassertschromaticallyfilled space as well as forming[2][1]:(F"-Ab-A)4and
[1][2]:(A-Bb)4-C5(showninterlockedat k). With the entranceof C5 and the
nearlysimultaneousexit of A4, [2]-striationsflanka [1][2]centre,(F-FS-Ab)4.
Furtherdevelopmentsemphasizesymmetry,but througha constantlychanging
web structure.First, just as Db4 disappears,a thickeningdevelopsat what
would be the centre if Db4 were still present(1). The formationis, literally
speaking,asymmetrical(m). The 'lump'is graduallydissipatedwith the exit of
G4 (b. 16) and of F#4 (b. 18);while this is happening,A4 reenters(b. 16), then
G4, just as F,"4 disappearsin b. 18, yielding another lump in reflective
symmetryto the firstone (n).
As A4 exits in b.20, the pace of events acceleratessomewhat.The graph
showsclearlyhow a 'ripple'of semitonalfilling(o) movesup throughthe space
in use fromthe bottom, startingwith E4 abovethe alreadypresentEb4. For a
split second, in b.22, the interval(Eb-Ab)4 is completelyfilled. Behindthis
ripplecomesanother,less well defined(p). The graphshowsmomentarygaps
in the continuityof individualpitches, as well as the fact that the entirespace
traversedis only (Eb-G)4this timeandis neverfilledallat the sameinstant.The
rippleanalogyis particularlyapposite,since the changein the textureat this
momentsoundsas thougha disturbanceof some sorthas been introduced- or
perhapsthe rippleis simplya premonitionof somelargereffect aboutto break
out upon the surfaceof the music. In any case, just as the second ripple is
makingits way up the semitonalspectrum,A5 entersdramatically,alongwith
its octaveA4, whichfunctionsas a continuationof the f1rstripple'sascent(q).

226 MUSIC ANALYSIS 6:3, 1987


LIGETI'S PROBLEM, AND HIS SOLUTION

ThepitchA5is quitea bithigherthananyotheryetheardin thepiece.The


locationof this seeminglydeliberatediscontinuityis, however,actually
determinedby the dimensionsof the spacepreviouslyoccupied,or, more
precisely,whatthesedimensions havecometo be. Thegraphshowsthatafter
theexitof DS4 in b.1S, theupperandlowerboundaries remainfixedatC5and
ES4 respectively, untiltheentranceof A5. Thisspace,[9]in extent,is doubled
bytheentranceof A5,forA5is [9]aboveC5.Atthatmoment,in b.24,wehave
two spatialregions:one ratherdenselypacked,the othercompletelyempty.
Immediately thereafterthe lower,denseregionbeginsto thin out: first,C5
exits;thena kindof descendingrippleoccursacrossbs 26-31,leavinglarger
intervallicintersticesin its wake(r). The effectherebringsto mindLigeti's
description ofaprocessthattakesplaceinSan FranciscoPolyphony,inwhichthe
texture'getslessdense,asif someonewentthroughit witha comb,thinningit
out'.48Alreadyin bs 25-6the [2]-striationshavereappeared: (ES-F-G-A)4(s).
These are momentarily obscuredby the descendingripple,emergeagain
partiallyin b.29 (t) as (F-G-A)4,thenfinallyarepresentedin unmistakable
fashionby the successiveexitsthatattenuate occupiedspacefrombelow:ES4
(b.32),F4 (b.34)andG4(b.35),leavingA4 (u). Meanwhile, fromabove,the
successiveexitsof C5in b.24andBS4 in b.34, togetherwiththeremaining A4,
presentthe by now familiar[1][2] pattern(v). Other [1][2] and [2][1]
configurations arealsopresentastheprocessof attenuation continues(w).
At b.35 the onlypitchesleft soundingcomprisethe emptyoctave,A4-A5.
Thiseventmarksa returnto themaximal clarityof thesingle-pitchopeningbut
cannotbe consideredexactlyequivalentto it, sincethe octaveintervaldoes
definea regionof space,howeverequivocally. Octave-bounded spacesplaya
considerable rolelaterinLuxaeterna;here,theoctave,besidesservingasapoint
of arrivalandasanobvioussectionaldivision,hasother,long-range functions.
The chordthatentersin b.37(bassesfalsetto,F,"-A-B) incorporatesA4. This
intervallicstructure,03][2],is identifiedby Ligeti as the principalstable
harmonyof thepiece.4Hereit servesasa spatialpivot(seeEx. 8):thevertical
distance[10]fromA5downto B4(successive upperboundaries) in
is duplicated
thedistancefromlowerboundary F,"4 (b.37)downto AS3, thelowerboundary
of thenextstableconfiguration in thepiece,firstenteringin b.51.
Ex. 8

MUSIC ANALYSIS 6:3, 1987 227


-* ' " ' ha b. b^-- ; ;_ ;, b. ---

JONATHAN W. BERNARD

AsforA4'srole,it is ofevenlonger-range import:inonesenseacomponent of


thefalsettochordatb.37,butin another,in thebarsimmediately preceding the
entranceof thatchord,thetrue'replacement' of theinitialF4 of thepiece.In
otherwords,A4-A5canbeheardasintervallic space,butit canalsobeheardas
a singlepitchwitha prominently audibleovertone.Theintervaldefinedby F4
andA4is duplicated, [9]lower,intheboundaries ofthestableharmony heardin
isolationin bs 58-61,AS3-C4.Theinterval[9]is thusconnectedto bothA4and
A5:thelatteron thelocallevel,theformerovera longerspanof time.50
LigeticallsLontanoa 'sisterwork'5lto Lux aeterna,and indeeda close
relationship betweenthetwopiecesis evidentfromtheoutset- notsimplyinthe
generalsimilarity producedbytheuseof canonicprocedures in both,butmore
specificallyin thatthe pitchserieswhich,usedin canon,'generates' the first
sectionof Lontano(bs 1-41)is nearlyidenticalto thatof thefirstsectionof Lux
aeterna(bs 1-37),[3]higher.52Example9 presentsacomparison. Apartfromthe
obviousdifferenceof chosenmedium,however,the treatment of thecanonic
strand in Lontano producesconsiderabletexturaldifferencesfrom its
predecessor. Foronething,thestrandinLontanobecomesdoubledattheoctave
abovebeginningin b.14,withthetwelfthnoteof theseries;foranother,non-
canonicpedalpitchesextendcertaincanonicpitchesbeyondthedurationthat
they wouldhaveif they wereprojectedsolelyby the canon;for a third,in
Lontanocontinualshiftsoccurin theorchestration of thecanon,allowingone
choirof instruments to begina segmentof thecanoniclineas anotherchoiris
finishingandproducing therebya greaterdegreeof overlapthanis possiblein
thefirstsectionof Lux aeterna,withits consistentsettingof foursopranoand
fouraltoparts.53
Theopeningtenbars(seeEx. 10)havethesamebasicshapeasdothefirstsix
of Lux aeterna(a), andafterthatthe [2]-striations openup belowin similar
fashion:F,"4in b.ll, E4 in b.l3 (b). Already,however,the courseof
development hasbegunto divergefromthatof theearlierwork.Notethatthe
pitchA4,onceintroduced inb.9(c),continuesto sounduntiltowards theendof
thesection,whereasF,"4, its counterpart in Lux aeterna,exitsfrom thetexture
forseveralbarsas the [2]-striationsareformed.In Lux, F,"4's absencehasthe
Ex.9

Lux aeterna
1 2 .3 4 5 6 7 X 9 1() 11 12 1.R 14 15 16 17
L D
Lontano +8X-------------- __
@1_ 1Z__ , 6_ b. 0, W. 1,, + , S _
1 2 3 4 5 6 7 X 9 1() 11 12 13 14 15 16 17

MUSIC ANALYSIS 6:3, 1987


228
-1$^ " #^ t #^
-----bz
5 # " , , t s b- *; # b , b s b-

LIGETI'S PROBLEM, AND HIS SOLUTION

effect of extendingthe striationthroughmuch of the rest of the textureat that


point;in Lontanothe retentionof A4 contributesto a greateropacityoverall,by
comparisonwith the otherwork.
The imitationof [2]-striationin the subsequententriesof higherpitchesin
Lux also occursin Lontano,but by differentmeans:the introductionof octave
doubling,as E, F,"andG,"5 enterin bs 14-15(d). The explicitoctaverelationto
the alreadypresentE, F,"and G,"4 beginsto establisha two-tieredstructure,
with eventsin two separateregistersmovingin parallel.At this point, though,
separateentrancesin the tworegistersserveto suggestthatthisparallelismis not
yet quiteestablished:A5 entersin b. 15, to complementthe alreadypresentA4
(e); C,"5 and B,"4, also in b. 15, arefollowedcloselyby C,"6 and B#5 (bs 15-16,
f and g); then the apex of the canonicstrand,D,"5, precedingD,"6 (bs 15-16,
h and i).
The suddenceasingof (E-F,"-G,"-A)5 in b. 19 leavesa hole in the texture(j).
The break is short-lived,but the momentaryseparationactually serves to
emphasizea paralleldevelopmentin the descentof the canonicstrandfromD,X,
throughC,"to B (k and1). The [2]-striationnow becomesa featureof the upper
regionof the texturein the two registers,as it did in the one registerof Lux
aeterna'sopeningpassage.Alreadyin b.20 we hearGX5 reentering,followedby
A,"5 and F,"5 in bs 21 and22 respectively(m).The hole is thusfilledagain,but
not completely,andnot quitein the sameway. In fact, thesethreepitches- G,X,
A,"andF,"- arethe nextsegmentof thecanonicstrand,afterD,X,C,"andB, and
they projectanother[2]-striation.
The graphshows, from aboutb.22 onwards,a distinctlymorepronounced
spatialseparationof the tworegisters,astheycontinueto moveevermoreclosely
in parallel.In fact, all pitchesin one octavenow haveoctave-correspondents in
the other, except for E4, the lowest of all. The graphsuggestsone possible
reasonfor the omissionof E5: preciselyto providesomemeasureof separation
between the two registers.The [2]-striationD,"-C,"-Bis maintainedon both
tiers,butotherwise,andespeciallyfromb.24, the textureentersits mostopaque
phase,with the spaceF,"-B in bothregisterscompletelyfilledchromatically(n).
The orchestrationhere,exceptforone sustainedclarinetpart,is allstrings,thus
Ex. 9 cont.

18 1') 't) 'I " '.R '4 'S rh '7 'S '9 A(} 31
_= DD __ _

- - - - - -- - - - - - - - - - - +8 Is

IS I'} 't) 'I " '3 4 S 36 37 'X '9 A() 31 Ar .AR

MUSIC ANALYSIS 6:3, 1987 229


stl.,.t?
., ,+
...
+P , . -t.T-.8tT-+.
i , | l.
.,, tu......T
.-;. ' ,*--,*--t*------------,
;t. .I T_, . *+* , *- . t ;,'---tt.'-
... ff ....... l....! t *.;* ** i*. i* , *.------- .t _ t ..*
+ . _ *!.- .. ; .t.....................................................................................
__' 1 ; - . tt *-_._. -t--* ,.. *. ,. . *+l..+ T
. '-.+-t.l.X.l
... ...
! ..!. * t* .+ *T
tjt .........
___. *t -,'_+
#__; *11
t - _-* *I*.......
v !- T- | I
|,,f......................
.. , +i fT,
+,* ....................
.t
.____
. . ....
-___
......
.. +.....
.......... _._......
...............
................
........
...............
.........
_ .,,
.j.........
.. t_....L._......
. _. *;.....,t,.t.+*,;{_........
.+
.........................
...............
... .. & . ,.*- j*,,
.t
. . _t-, 1
.t........
............
t .......4''.
*....
. . t.t ...........
+F
. .j ::'1
e*g ...... *!............
e$.. +, +* -,e*:'i
............ _ . i_,
!t,.,*ti*_
;,+....
_ . .. ;--
t __ t
. _.i . . _ . . . e .+ . .....+ .....t t
t .* . e. . - 4 .. - t 1
. ++ . - - - + -. +. 4...............
*. +F. + ls++. . ,
.......................
_ we e+* * ..... ,.. *4 . * _ I ....
_ -...............
. * +* *, _.....
* . * * . * _ ,. ;
. * fv- .Z. ,,.,
* t. S. *. ....., +.,-,*--
.P . . _* + .t... * . + * _. .....
*'- .. * . 4

PROBLEM, AND HIS SOLUTION


LIGETI'S

cont.
10
Ex.

il, lill 14
---4IlLt
t '1 1lF
_tlItutt--0-,-++t
!; 1t i-X+X , ,
4-tl
j . t ,
ts
, .
++l!0-
j . ,
. > , z , . . . , , t ,

'
. ,

* *
, t .

!
.

'
.

* '
. .

'
.

'
.

'
.

'
. .

t
. .

' , + , _ . * * i t
* i t t * * + 4 4
+ ; t t;
1 * w I
, . ffi I . i
t + i * * * + +

I * ' * * . s

X ,,; . .;
., , .
, . + ,,
. . |

., .
...., ..
* * ..,
! * * *
t ' t * * f t * * $

--- *t- e,,--,.-.-z-..


*-j *-t
- -- - - ---
- -t --
T ---- . F
* jS t * t t , f + . t

't |
t^t- t' '

__ _ ;__________-S
_t

*i

_-- ' ' . .


' . .
I ;-
, . _

! I _1
.

-; * -
, .
. .
.

I 1;11 ' . .
. .
, .
. .
.

5 . .
. .

........
! ' -!t t :, .. ;, .
. . . . . . . . . . . .

* * * *
*

.
.
. .
. .
. .

= a
, . .
.
. . , . . . .

,.+_
l
-_ ......
t
t ' * ^ - - * * * t .
^, , . p .........
-! ............................
_._

i
..........
. .
. .
+ .
- .
. .
. .
. .
. . .
. . . .
. . . .
. . .
. .
. .
. .
. .
. .

* +
f f
. .

* .
+
* +
.

. * - . . . *
- - + . . . t i
+ -
+ * *

++ *+ + +
* .
*

*
t f * .
, f
*
P P *
t *

-* - t- -........
-. . . - t
. - . .- - + -
.

*- -! . . .
..
t
. .
. .
.
. .
.

!
, .
. .
. .

fl
. .
.

' ' ' '

+- +
' ! '

t * - - - - -
......... . * * *
, t
*- j- *- -
. .
. .
, ...

.. ....
. . .
. . . .
. . .
. . +
. . . .
. + . . .
. . . . . ,
* * . . . . . . _
. ,
! ' * * . . .
. . . . . .

. , . * * . . . . .
* * . . . . .
. . . . . .
. . . , .
. . . . .
. . . . .
. ,
.; . . . . .
. . .
. . .
* t --
. . .
.
. .
. , . ,
t
* t F
t * * *
. .
.
.
, t
. . .
*
* *
* * * . . t
. . .
. . . .
. *
f . .
. .
. . . .
* . . .
. . . . .
.. . . . .
. t . , .
P I .
, .
,
. * . ,
. .
........ t
. .
,
.
.
.
.
.
. . .

* X,, * ..., '


+

. * t -t i+ l i
! t-**s--+ *..... . t
+
. . .

.. ' I ..............
I
+! ..I, t + t
..
. .
,

.. . I f ..................
I ! *-...... ......... + t --
t
. .

, t
..
. t
. .
. . . . . .
* +! *
t- i | -i I -
.
. .
. .
. .
.
..
! t i

t F i 4 t it- -
++ t--t
.
. .
. . .
..
. . . . . . .

t t ! t

--
_ . _.+
t
. .
. .

*, t * , t t
; + . . .

'. !
. .
*

i t
. . .

* ' * t - * s

_ X _ _
. . .

'i' ! j * .
. .
. .

t .
. f
. . ;

,. ....... iF ' * ' * * F tt t n - - -


.
. .

4 . s . ' * - - - - t t
t - t-1-
+- -
. -

. j. . . i F___ t
* * l
! ' F;
;
! *
t !
- t * t * . . r- t f
- - - ^ - + __
-4
, j .... T + ' ' * t ' ' *' * t t F I _ !
-t . -- * t t t ** !F tt *;
.
. ..
I
- --
v
-
t+;t t
.
.
. .
. .
..
.
t t ! ' -- ** t
f; +
, ' 4
I i t
--4--t
L
+
'
.

*; * 4 s
F + *l * tt
* Ib t
1*;-- 1 l F '
- ____
+

t-* + t
t....
- .

. .
, ;
1,
,
| .

t i
t tt*
*
t
{ .
'. . -- t - t
t
t
itt i
+

' t
. t ------_
---t------
J
. . i, !
i . ; _
,
. .
., . .,

*[,* -
t0 t - - - -

l
.
...

t! ! !
- ---r -
- - -

t
.

*I
. . .

L 'I
. .
_ _ _
_ __.
_ _ _ _ _ _ _

!
_ _
t _ _
. _
* . _ .
_
_ _ _ _ _ .
_
_ _
. . _
_ _ l
I
2
. _

t ;T . _
_ .

41
.

+ _ _ .
_ . .

,
_ . _ _ _ _

37 38 Bg Z

. _
_
_ __ _ _ _ _ .
_ _ . _ . .
._

34 3S 36
_

. . _
.

31 32 33
- -
I -I l I . 261
'ZSI
Xe,
1tjCI
2X t1?1,
281 291 30
I
2J1 j 241
1
221

231
MUSIC ANALYSIS
6:3, 1987
IONATHAN W. BERNARD

emphasizing homogeneity of texturebycomparison withwhathasgonebefore.


This passagehas the effect of cancellingmost previouslyestablished
relationships andpavingthewayfortheshapeofthesection'sconclusion. Asfor
the [2]-striationpreservedabove,it is not disrupted,appropriately enough,
untiltheentranceof thepitchC (b.31),whichin severaloctavesservesthesame
functionhereasdo theoctaveAsin Lux.
Aftertheentranceof C(S,6, 7), thetexturebeginsto thinnoticeably.Pitch
D# dropsoutin b.32,thenE4andB (4,5)inb.33.Thegraph(ato andp)reveals
thatthethinningprocesssystematically exposes[2]-striation in bothregisters,
and that the pitchesformingthese striationsthen proceedto drop out,
beginningwiththe lowestandendingwiththe highest,thusattenuating the
textureuntilonlyCis left.Combined withthisreemergence of the[2]-striation
is the persistenceof C# (5, 6) untilb.38. As in Lux, the descentof theupper
boundary(q andr) producesthe intervallicarrangement [1][2];alsoas in the
earlierpiece,the soundingof the lastthreenotesof the canonicstrandalone
(Db-Bb-C)fromb.37on constitutea formof [2][1].
Lookingback,we findthatthefluctuations in thedensityof thetexture,as
revealedbythegraph,areto a certainextentreflectedin theseriesof pitchesin
thecanonicstrand.Alreadynotedarethe[2]-striations attheouteredgesof the
rangeof thestrand:E-F#-G#andD#-C#-B,whichoccurasdistinctsegmentsof
thecanonicseries(twelfthto fourteenth andnineteenth to twenty-thirdpitches
respectively). Alsoworthmentioning are,forone,theveryopening,whichasin
Lux producesa narrow,compactbandof pitchesfromtwo symmetrically
arranged formsof [1][2],andfromnumerous repetitionsof Abamongthefirst
nine pitchesin the series;andfor another,the segmentextendingfromthe
twenty-fourth pitch(G#) to the thirtieth(Ab), in whichtheconstricted range
andconstantchangesin directionproducetheclimaxof opacitynotedearlierin
bs 25-32.
Thesecanonic'resultants' arealsonoticeable asfeaturesof theLuxgraph.At
this point, a side-by-sidecomparisonof Exs 7 and 10 is instructive.The
similaritiesbetweenthetwopiecesshowupmorein generalshapeandcontour
thanin specificdetails.Internalrelationships, in particular,
arequitedifferent.
It is wellto remember thatowingto theoctavedoublingin Lontanothevisual
comparison withLux is moreaccurately made(especiallyfromaboutb.21 in
Lontano)betweenLux andhalfthe textureof Lontano.Evenso, however,the
realisationof notes22-8of t}lecanonin Lux(o andp, Ex.7)differsconsiderably
fromthatof notes24-30in Lontano(n, Ex. 10)becauseof thelargernumberof
partsin thelatter.Furthermore, themotiontowardthebareoctavesattheends
of the sectionsunderanalysisis managed,in Lux, partlythroughan internal
'emptying'of occupiedspacewhichdoesnot takeplacein Lontano.(Seethe
fissurethatappearsin the wakeof the descendingrippler in Ex. 7.) This,as
notedin theanalysisabove,comesaboutin apparent responseto theregionof
emptyspace,equalin sizeto thespacealreadydenselyfilledbelow,openedup
byA5.Butin Lontanotheentranceof C7is a lessdramatic eventthanthatofA5
in Lux simplybecauseoctavedoublingis alreadya factof the texturein the

232 MUSIC ANALYSIS 6:3, 1987


LI{GETI'S PROBLEM, AND HIS SOLUTION

orchestralpiece.Theprogress fromthispointto theconcluding octavespacings


in Lontanosoundslesslikeanemptying ofspacethanafocusinguponoctavesthat
arealreadypresent.
Ligeti'sproblemcouldbe summarized in thefollowingquestion:Howdoes
onearriveatacompositional practicethatsatisfiestheintellectaswellastheear?
Theworstexcessesof theserialisteraindulgedtheformerattheexpenseof the
latter,butparticularly forthosewhoshunnedtheserialistpath,afterlearning
whatit ledto, theexperience ofserialismmadeit effectivelyimpossible todevise
alternativesinwhichconsciousmethoddidnotplaya prominent role.Ligetihas
saidthatwhenhe composeshe alwaysbeginsby imagining thewayhe wantsa
piecetosound,ingreatdetail,frombeginning toend;thenhefiguresouthowto
producethatsound.Thismightseemtobeexcessattheoppositeextreme,were
it notthattheactofwritingthepiece,forLigeti,invariably changestheoriginal,
imaginedplan.54This mustmeanthatthe imaginedsoundcanonlybecome
audibleif it is basedon consistentprinciples.Theinaudiblestructure doesnot
justifytheaudiblemusic,butwithouta structure it willnotbe possibletoknow
preciselywhat the musicshouldsoundlike. Hereis wherethe purposeof
contrapuntal structurein Ligeti'smusicbecomesevident:it is a rule,hidden
thoughnot secret,that can be appliedflexiblyand with greatcontrolover
temporaland spatialdimensions.Becauseit is noticeablynot a mechanical
procedure,it does not automatically dictatein all respectsthe soundof the
emergentmusic,but neitheris it simplya convenientexcuseon whichthe
composermayhanghis preconceived notions.Therecanbe littledoubtthat
Ligetiderivesa deepsatisfaction, bothaestheticandtechnical,fromemploying
canonicprocedures as behind-the-scenes devicesthatenterinto an intricate
reciprocal relationshipwiththesoundingsurface,shapingthefinalresulteven
as theyareshapedto meettheexigenciesof thework.It is a satisfaction thatis
readilycommunicable to thelistenerandanalyst.

NOTES

1. GyorgyLigeti, 'Metamorphosesof MusicalForm'(1958), in Die Reihe, Vol. 7


(Fonn-Space), Englishedn (BrynMawr:Presser,1965),pp.5-19.
2. Ligeti, 'DecisionandAutomatismin StructureIa' (probably1957),in Die Reihe,
Vol. 4 (YoungComposers), Englishedn (BrynMawr:Presser,1960),pp.36-62.
3. Ligeti, 'Fragenund Antwortenmit mir selbst'(1971),trans.GeoffreySkelton,in
Ligetiin Conversation(London:Eulenberg,1983),pp.124-37.
4. Ligeti, 'Uber Form in der neuen Musik,' Darmstadter BeitragezurneuenMusik,
Vol. 10 (1966),pp.23-35.
5. 'Metamorphoses,' pp.5-6.
6. Ibid.,pp.8,10.
7. 'UberForm',p.31: my translation.
8. 'Metamorphoses', p.6. Specifically,in Structures
Ia he notedthat 'Whenwe hear
this compositiona complex networkunfolds - of coarseror finer weave . . .

MUSIC ANALYSIS 6:3, 1987 233


IONATHAN W. BERNARD

consistingof a significantlyorderedflockof sounding"points":theseareorganised


to formthreadsof variedthickness,whichnow standout plastically,now become
less distinct.The threads,for theirpart,arewoventogetherwith greateror lesser
density'('DecisionandAutomatism',p.61).
9. 'Fragenund Antworten',p.l31.
10. 'Fragenund Antworten',p. l 31.
11. Ligeti, 'On MusicandPolitics',trans.Wes Bloomster,Perspectives of New Music,
Vol. 16, No. 2 (1978),pp.19-24
12. PeterVarnai,'BeszelgetesekLigetiGyorggyel'(1978),trans.GaborJ. Schabert,in
Ligetiin Conversation, pp. l 3-82.
13. Ligeti, 'Zustande,Ereignisse,Wandlungen',Melos,Vol. 34 (5 May 1967),pp.l65-
9: my translation.
14. Ibid. An Englishtranslationof Ligeti'saccountof this dreamappearsin Ligetiin
Conversation, p.25n.
15. 'Metamorphoses', p. l 5.
16. Ibid. The latterphraseis Ligeti'sown quotationfrom Adorno.See TheodorW.
Adorno, Philosophyof ModernMusic, trans. Anne G. Mitchell and Wesley V.
Bloomster(New York:Seabury,1980),p.188.
17. 'UberForm',p.24.
18. Ibid., pp.23-4, 34.
19. Ibid.,p.34.
20. Ibid., p.26: my translation.
21. 'Metamorphoses', p. l9; emphasisin original.
22. 'Fragenund Antworten',p.125.
23. Varnai,p.l5.
24. 'Metamorphoses', p.l5.
25. Varnai,p.25n.
26. PaulGriffiths,GyorgyLigeti(London:Robson, 1983),p.20.
. .

27. Ligeti,'UberneueWegeim Kompositionsunterricht: EinBericht',in ThreeAspects


of New Music(Stockholm:Nordiska,1968),pp.9-44.
28. It is surelysignificant,in thisconnection,thatin his ownanalyticwritingLigetihas
notedinstancesof pitchsymmetryin Bartok;see 'Uberdie Harmonikin Weberns
ersterKantate',Darmstadter Beitragezur neuenMusik,Vol. 3 (1960), pp.49-64.
Registrallybasedsymmetry,in particular,playsmoreof a role in Bartok'swork
thanhasgenerallybeenrecognised;seemy article'SpaceandSymmetryin Bartok',
3rournal of MusicTheory,Vol. 30 (1986),pp.185-201.
29. 'Fragenund Antworten',p.128.
30. JosefHausler,'ZweiInterviewsmit GyorgyLigeti' (1968, 1969),trans. SarahE.
Soulsby,in Ligetiin Conversation, pp.83-110.It shouldbe notedthat,eventhough
in the resultanttexture of clusters the traces of Bartokianstyle have all but
vanished, still the originalidea of clusters predates Ligeti's emigrationfrom
Hungaryandis attributedby him to his familiaritywithBartok'smusic(Griffiths,
p.26). See, for instance,the third movementof Musicfor Strings,Percussionand
Celesta,bs 20ff. The firstversionof the firstmovementof Apparitions wasa work
entitledVtziok,completedin Budapestin 1956.

234 MUSIC ANALYSIS 6:3, 1987


LIGETI'S PROBLEM, AND HIS SOLUTION

31. 'Zustande,Ereignisse,Wandlungen',p.169: my translation.


32. 'Metamorphoses', p.14; 'Fragenund Antworten',pp.131-3.
33. This planof attackleavesotheraspectsof Ligeti'ssoundstructureunengaged.The
readershouldbearin mind,however,thattheseotheraspects,mostlyof secondary
importanceaccordingto the composer,areorganisedalongindependentschemes
(againaccordingto Ligeti)andthus arepresumablyseparablefor the purposesof
analysis.
34. UrsulaSturzbecher,'GyorgyLigeti' [interview,c.1970], in Werkstattgesprache mit
Komponisten (Cologne:Gerig, 1971), pp.32-45. Obviously,however,I make no
claim to having recapitulated,in any substantivesense, Ligeti's compositional
process.As he himselfsaidof his analysisof Webern'sOp.29, 'Analysiscando no
morethanmerelyconjureup the shadowof the actualcreativeprocess'('Uberdie
Harmonikin WebernsersterKantate,'p.53n: my translation).
35. Throughoutthis articlethe word 'interval'is used exclusivelyto mean vertical
distance,whetherchromatically filledor not. Thus 'interval'reallymeans'vertical
compass'in Ligeti'sterminology.Despitethe potentialconfusion,I havedecided
to use 'interval'in a sense not meantby Ligeti's use of the word, since 'vertical
compass'or even 'compass'is syntacticallyawkward.
The double verticalbars in the graph indicateclusters:completechromatic
filling. The fact that highestand lowest points, and boundariesof instrumental
groupswithinclusters,are markedwith pitch namesis not meantto suggestthat
these boundariesare more'pitch-like'thanthe notes withinthe boundaries;they
areshownin this fashiononly to makeit easierfor the readerto graspvisuallythe
infa mationconcerningintervalsizescontainedin the graph.
36. The shortdashor hyphenis used hereandsubsequentlyto separatesinglepitches
or to referonlyto the boundariesof filledspaces;the longdoubledash(= ) refersto
thecontentsof clusters:chromatically filledspacesboundedby thepitchesnamed.
37. Sturzbecher,p.39: my translation.
38. Varnai,p.39.
39. The horizontalbracketin bs 1, 13 and 25 signifiesthatthe verticalsencompassed
areattackedsimultaneously.
40. The pitch namesin the graph,here and in bs 18-20(andalso in Ex. 4), markthe
individualpointsof maximumloudness.
41. Pitchnamesin bs 23-5 showeachseparateset of entriesverticalized.
42. The quasi-canonicpassageof the secondmovementof Apparitions (bs 25-37), in
which46 stringparts(24 + 8 + 8 + 6) playingfourdifferentlines begintogether
but then immediatelyget out of phase, seems to be the first exampleof this
integration.For extensivecriticalcommentaryof widely varyingusefulnesson
Ligeti'swork up to 1965, see ErkkiSalmenhaara,Das musikalische Mater2alund
seine Behandlungin den Werken'Apparitions,''Atmospheres,' 'Aventures,'und
'Requiem'von Gyorgy Ligeti, trans. from the Finnish by Helke Sander,
Forschungsbeitrage zurMusikwissenschaft, Vol. 19 (Regensburg:Bosse, 1969);Ove
Nordwall, GyorgyLigeti:Eine Monographie,trans. from the Swedish by Hans
Eppstein(Mainz:Schott, 1971),pp.9-77.
43. The canonicstructureof this passageis alsodiscussed,incompletely,hy Reinhard

MUSIC ANALYSIS 6:3, 1987 235


JONATHAN W. B ERNARD

Febel in 'GyorgyLigeti: Monument- Selbstportrait - Bewegung(3 Stuckefur 2


Klaviere)',ZeitschriftfurMusittheorie,Vol. 9, No. 1 (1978),pp.35-51;Vol. 9, No.
2 (1978),pp.4-13.
44. Foranotheranalysisthatexploreslarge-scalesymmetryin Ligeti'smusic,see Pozzi
Escot, "'Charm'dMagicCasements":GyorgyLigeti'sHarmonies',in Contiguous
Lines:Issuesand Ideas in the Music of the '60's and '70's, ed. Thomas DeLio
(Lanham:UniversityPressof America,1985),pp.31-56.
45. Hausler,p.108.
46. Sturzbecher,p.44.
47. Ligeti, 'Auf dem Weg zu Lux aeterna',Osterreichische Musitzeitschrift,
Vol. 24
(1969),pp.80-8.
48. Varnai,p.44.
49. Ibid., p.29. Ligeti callsthis harmonya 'typicalLigeti signal'anddescribesit as 'a
fourthmadeup of a minorthirdanda majorsecondor the otherwayaround'- that
is, [3][2]or [2][3],whichformulationseemsto confirmthatthe ideaof inversional
equivalenceis validin generalfor analysisof his music.
50. The reader'sattentionis directedto two other analysesof Lux aeterna:Clytus
Gottwald,'Luxaeterna:Zur Kompositionstechnik GyorgyLigetis',Musica,Vol.
25 (1971), pp.12-17; RobertCogan,New Imagesof MusicalSound(Cambridge,
Mass.:Harvard,1984),pp.39-43.Cogan'sspectrumphotosseemto someextentto
coincidewith my graph,althoughdetailin the photos is difficultto makeout at
certainpoints.
51. 'Aufdem Weg zu Luxaeterna',p.86.
52. Therearecloserelationshipsbetweenthe canonicstrandsof othersectionsas well,
althoughtheyvaryas to orderwithinthe piecesandas to levelof transposition,and
somestrandsand/orpartsof strandsin eachareunique.
53. BruceReiprich,in 'Transformation of Colorationand DensityinGyorgyLigeti's
Lontano',Perspectivesof New Music, Vol. 16, No. 2 (1978), pp.l67-80, has
discussed this orchestrationalsegmentation. However, Reiprich complicates
mattersunnecessarilyby referringto each of the segments- eight in all - as a
separatecanon.
54. 'LigetiTalksto AdrianJack',MusicandMusicians,Vol. 22, No. 11 (1974),pp.24-
30.

236 MUSIC ANA LYSI S 6:3, 1987

You might also like