2010-Meaningful Scribbles jmm9 - Libre

Peer-Reviewed Paper JMM. The Journal of Music and Meaning, vol.
9, Winter 2010
Meaningful Scribbles:
An Approach to Textual Analysis of Unconventional Musical Notations
Douglas C. Wadle, Adjunct ProIessor,
CaliIornia Institute oI the Arts School oI Music
Abstract
This paper draws upon recent studies oI writing systems, buttressed with
material drawn Irom general linguistics and semiotics, to develop an
approach to the analysis oI musical notation as a system oI signiIication in its
own right (as opposed to a mere representation oI musical sound). Scores are
herein understood, Iollowing Nicholas WolterstorII`s Iormulation in his paper
'Towards an Ontology oI Artworks (1975), as 'a record oI the artist`s
determination oI correctness-conditions. Our analysis must, then, provide a
clear explication oI the means by which these conditions are communicated
even where these means are not readily understood as drawing on any known
signiIying convention. With this in mind, the method oI analysis is designed
to be Ilexible enough to accommodate any notated document purporting to be
a musical score regardless oI the particular notational conventions used. A
description oI the relevant linguistic and semiotic terminology and its uses is
Iollowed by a discussion oI their application to the study oI elements oI
standard notational practice. A sequence oI steps, through which the analysis
oI an unconventional 'graphic score is to proceed and by which musical
meanings are to be assigned to the score`s markings, is presented. This
sequence is illustrated through a progressive sample case that oIIers a range
oI possible musical interpretations Ior variations on a simple notation oI the
author`s devising. Finally, a discussion oI the possibilities Ior evaluating
unconventional notations, once musical meanings have been ascribed to all
markings within the score, will be undertaken with reIerence to John Cage`s
'Variations I.
38
Peer-Reviewed Paper JMM. The Journal of Music and Meaning, vol.9, Winter 2010
1. Introduction
While scholarly attention has been paid to the notational component oI music, these studies
1
do
not explicitly outline a method whereby a musical notation can be treated as a distinct system oI
signiIication (as opposed to a mere representation oI musical sound). The extant study that has
come closest to this aim is Leo Treitler`s 'The Early History oI Music Writing in the West
(1982), which applies Peircean semiotic terminology to the classiIication oI regional musical
scripts oI the late medieval period. Recent studies
2
in linguistics reveal some Iundamental
properties oI writing systems Ior which Treitler`s study does not Iully account. It is my intention
to build upon these studies, augmented by relevant ideas Irom semiotics and general linguistics,
in proposing an approach to the analysis oI any notated document purporting to be a musical
score regardless oI the particular notational conventions (or lack thereoI) used. The desire to
undertake such an analysis may strike some as odd. AIter all, the purpose oI a musical notation,
iI it is to be a musical notation, is to describe or guide the perIormance oI the musical work
notated. How, exactly, a notation does this is still a matter oI some philosophical debate, but Ior
present purposes, Nicholas WolterstorII`s Iormulation will suIIice: 'Most scores Iunction and are
meant to Iunction to guide perIormances. But what is true oI every score is that it is a record oI
the artist`s determination oI correctness-conditions (1975: 136). Given this Iunction, the
importance oI a clear understanding oI the medium by which these 'correctness-conditions are
communicated should be evident. II we are to identiIy proper instances oI a work (oI notated
music), then we must know what the notation signiIies.
This paper was written with the radical notational experiments so-called 'graphic
scores such as those created by the New York School
3
in the 1950s and 60s and their Iollowers
in more recent decades in mind. In these works, the notation is not construed simply as a
means oI describing a sound structure or delimiting perIormance techniques. Rather, a level oI
strategic ambiguity is introduced by the use oI unconventional and oIten newly-invented
symbols. These symbols, when not tied by tradition to a pre-existing musical meaning, interIere
1
For a recent sampling, see Leo Treitler (1982, 1992), Kenneth Levy (1987), Anna Maria Busse Berger (1993), and
Alma Colk Santosuosso (1989) regarding the historical development oI musical notation; see Gardiner Reed Ior a
practical guide to modern notational practice and (1987) Ior a survey oI proposed notational innovations; see Jane
Alden (2007) Ior a recent study oI the historical precedents Ior Earle Brown`s notational innovations and Mieko
Kanno (2007) on modern innovations in prescriptive notation.
2
See Coulmas (2003), Harris (1995), as well as Gerhardt Augst, ed. (1986), New Trends in Graphemics and
Orthographv (Walter de Gruyter).
3
'The New York School is a term that has gained some currency in reIerring to the collected works oI composers
John Cage, Morton Feldman, Earle Brown, and Christian WolII.
39
with our ability to identiIy the 'correctness-conditions indicated by the score. In such cases a
new reading strategy is required (oIten aided, but not Iully accomplished, by the inclusion oI
perIormance instructions, though in some cases no directions are provided at all
4
). The piece oI
music notated in such a way results Irom a dynamic interaction between the ambiguities oI the
notation and the Iixing oI those ambiguities into particulars to guide a perIormance. The
approach presented herein should serve as a basis Ior such readings. This approach is presented
as a sequence oI steps through which the analysis oI an unconventional 'graphic score is to
proceed and by which musical meanings are to be assigned to the score`s markings (section 4).
A progressive sample case (section 4.1) walks us through this sequence, oIIering a range oI
possible musical interpretations Ior variations on a simple notation oI the author`s devising. In
preparation Ior this, necessary concepts Irom linguistics and semiotics are introduced (section 2)
and applied to conventional musical notation, which serves as a background Ior our
consideration oI unconventional notations (section 3).
While my primary concern is with radical notational experiments, the present techniques
can also be applied to other notational approaches, including wholly traditional ones. For
instance, a study oI the prescriptive notational innovations oI Helmut Lachenmann, would reveal
much about the composer`s conceptualization oI the perIormance techniques he calls Ior. A
study oI the relative success oI several oI these innovations might, in turn, reveal something oI
the way musicians experience and interpret musical notation. An analysis oI the use oI (usually)
standard notation to represent the spatial conceptions oI musical sound in the work oI composers
such as Varese, Xenakis, and Tenney might prove similarly illuminating oI their work. It should
also be stated that the discussion will raise philosophical questions regarding reIerence and
meaning as well as the ontology oI the musical work. These will have to await Iuture studies Ior
their Iull treatment. The question oI the evaluation oI musical works notated by unconventional
means will be addressed in reIerence to John Cage`s 'Variations I (section 4.2) with the
Iollowing caveat: I will conIine myselI to the evaluation oI the appropriateness and inventiveness
oI the notational means employed weighed against the apparent intent oI the composer. I am not,
at present, concerned with establishing criteria Ior the aesthetic assessment as works of music oI
works oI music notated by unconventional means. That issue can only be properly addressed
once the Iull analysis oI such works merging a textual analysis oI the sort I propose with
phenomenological analyses oI a sampling oI perIormances oI the work in question and
ethnographic study oI the decision-making processes required Ior each realization/perIormance
4
See, Ior instance, Cornelius Cardew`s Treatise (1967) and The Great Learning (1971).
40
oI the work has been undertaken. Obviously, to present such an analysis, in addition to
outlining the proposed method Ior undertaking the Iirst part oI that analysis, would require more
space than I could reasonably devote in a single article.
2. The Study of Writing Systems
This section will introduce the requisite terminology Ior the analysis, deIining terms Irom
linguistics and semiotics and explaining their interrelated Iunctions. Most studies oI writing
systems draw upon the tools developed by linguists to make possible the analysis oI spoken
language. Such analysis is predicated on the assumption that:
. no matter how we answer the question oI whether writing Iollowed, preceded or accompanied
the recognition oI diIIerent levels oI linguistic structure, it is clear that, by virtue oI the Iact that
every writing system maps onto a linguistic system, it embodies and visibly exhibits the dissection
oI units oI language and thus linguistic analysis. (Coulmas 2003: 151)
A writing system marks distinctions appearing, at some level, in the sonic Iorm oI a language. In
so doing, writing can Iunction as a medium Ior the analysis oI spoken language (at the level oI
those distinctions). In broader terms (not privileging spoken language), two distinct expressions,
the written and the spoken Iorms oI an utterance, are correlated by a structural analogy. For
instance, iI we have symbols representing complete spoken words, then we have a logographic
writing system. II we have symbols representing each possible syllable oI the language, then the
writing system is syllabographic. Where letters correspond to phonemes, the system is
alphabetic. These smallest units are those by which the structural analogy is articulated.
However, the units oI these expression-planes (written or spoken) can be Iurther analyzed
though not always (or oIten, even) into meaningIul parts Ior the conventions governing their
Iormation.
Like linguistic sound systems, writing systems are structured and can, accordingly, be analysed in
terms oI Iunctional units and relationships. The distribution oI these units is governed by
restrictions limiting their linear arrangement in Iorming larger expressions. These restrictions can
be understood as operating on the graphic level alone. In this sense every writing system is to be
analysed as a system in its own right irrespective oI other kinds oI linguistic structure to which its
units and compound expressions may reIer. (Coulmas 2003: 34)
Our purpose then includes, not only, an explication oI the nature oI the analogy between
expression-planes, but also a study oI the sub-analogical structures oI one or the other
expression-plane (i.e. the structural aspects oI an expression-plane that exist below the level oI
41
the analogy coordinating that expression-plane with another).
According to linguist Roy Harris, the most Iundamental characteristic oI writing systems,
at least in reIerence to their analysis, is the presentation oI the writing within a graphic space.
The graphic space makes available two- and three-dimensional spatial structures Ior the
representation oI one-dimensional, time-bound activities to which writing is taken to correspond,
such as spoken language or musical perIormance (Harris 1995: 46). Harris distinguishes two
Iorms oI writing systems, scripts and charts. Charts may be understood as graphic
communications that assign meaning to locations within the graphic space, scripts those that
assign meanings to arrangements oI marks in relation to one another (Harris 1995: 93). OI course
there is no absolute instance oI either case a chart with no markings conveys no meaning,
neither does a set oI symbols with no arrangement in graphic space. A chart operates by
assigning a gradation oI values in some parameter to (at least some oI) the graphic space`s
dimensions, as in a map, a seismograph, etc. Scripts still depend on temporal axes determining
the order in which the symbols are to be read. In written English, a primary temporal axis is
assigned along the horizontal, a secondary temporal axis along the vertical dimensions oI the
graphic space. Given the syntagmatics oI the graphic space in charts, one need only assign the
correct parameter and demarcation oI unit values to each axis in order to read the chart.
Any script is comprised oI marks made within a graphic space. For these marks to be
meaningIul, they must create a distinction within the expression-plane that is understood to
correspond to a distinction in the content-plane oI an intended message
5
. Nina Catach, Iollowing
the inIluential linguist Louis Hjelmslev, has reIerred to such marks as 'pleremes. In certain
cases, a plereme may be divisible into smaller markings that do not carry meaning in themselves.
Such is the case with the morpheme/grapheme (word/letter) distinction in alphabetic writing
6
.
Marks that do not carry meaning within a writing system are called 'cenemes (Catach 1986: 3-
4). The plereme/ceneme distinction is similar to Andre Martinet`s notions oI first and second
5
In some instances this content-plane will be construed as the analogous structure in the expression-plane oI spoken
language, which, in turn, reIers to another content-plane with which it shares some structural analogy (and so by
extension with which the written expression shares an analogy) though the analogy to the content-plane is not
necessarily oI the same structural Iorm as that between the spoken and written expressions that reIer to it. For
example, the word /book/ as spoken and 'book as written share a structure based upon phoneme/grapheme
correspondences, but either /book/ or 'book, as singular units distinct Irom other nouns in the English language,
reIer to the object //book// as a singular entity among other physical objects a grosser analogy than existed between
the two expression-planes.
6
Actually the graphemes do have a reIerence to some phoneme but it is not a reIerence that is operative in the
consideration oI the reIerence oI the word. This phonemic reIerence operates at a diIIerent level oI description, the
determination oI which will be described later in this section. But even this phoneme correspondence runs into
trouble, as some phonemes require two letters (digraphs) while single letters can reIer to more than one phoneme.
42
articulation the smallest units oI meaning and the non-meaningIul units oI which those
meaningIul units are comprised, respectively
7
. As we shall see below, the concept is not Ilexible
enough to deal with all oI our concerns with regard to writing systems; however, it will be seen
to be adaptable into a useIul tool Ior speciIying the level oI description at which (a given portion
oI) an analysis operates. The identiIication oI marks as either cenemes or pleremes will be
contingent upon a tradition, the writing system, which is comprised oI an inventory oI symbols
(themselves, pleremes or combinations oI pleremes), a code governing the correlation between a
symbol and its meaning, and a convention governing the structural arrangements oI subordinate
markings into meaningIul units. In the case oI cenemes, we will have either marks that are
combined to Iorm a plereme or that are supererogatory visual inIormation added to the plereme,
serving to make clariIying distinctions between similar (but distinct) pleremes or Iunctioning
merely as ornamentation. An example oI such a clariIying mark would be the horizontal line
through the middle oI the Arabic numeral, 7, which some people use to help distinguish 7 Irom
other marks to which it might bear a resemblance T or Z, Ior instance. The seriIs that adorn the
letters oI various Ionts are an example oI ornamental cenemes.
The set oI rules by which written symbols are constructed Irom visual marks (cenemes
and/or pleremes) is one instance oI what Umberto Eco calls an 's-code (meaning structural
code): 'S-codes are systems or structures` that can also subsist independently oI any sort oI
signiIicant or communicative purpose. They are made up oI Iinite sets oI elements
oppositionally structured and governed by combinational rules that can generate both Iinite and
inIinite strings or chains oI these elements, (Eco 1976: 38). While a code governs the
Iunctional correlation between the symbol and reIerent, creating a sign, an s-code governs the
structuring oI larger units Irom smaller units irrespective oI the semantic import (or lack thereoI)
oI either level
8
. It is this irrelevance oI semantic import that allows us to describe an s-code Ior
the combination oI cenemes into pleremes within a script-based writing system. Eco tends to
highlight binary oppositions oI strings oI elements in discussing s-codes; however, he explicitly
states that this binary opposition is not a necessary Ieature oI s-codes. An s-code may even
operate as positioning elements within a continuum (Eco 1976: 178). This observation is crucial
when regarding writing systems the arrangements oI cenemical markings is not a simple binary
7
The idea oI articulation is based upon a mono-linear conception oI spoken language, understood by linguistics as a
concatenation oI phonemes partitioned into morphemes (Martinet 1964: 25).
8
This immateriality oI meaning is more evident in the Iollowing description oI a sample s-code: '. we should think
oI magnetized marbles which establish a svstem of attraction and repulsion, so that some are drawn to one another
and others are not, (Eco 1976: 124).
43
Iunction but occurs at some number oI positions within the two-dimensional graphic space
think oI the possible positions oI vertical and oI the horizontal lines allowed in the Latin
alphabet`s capital letters as exhibited in L, T, and H. This structuring oI non-signiIicant elements
(cenemes) rests upon the Iact that the symbols oI writing can be divided beyond the level oI
second articulation, in a seeming contradiction to Andre Martinet`s original use oI the term
9
.
The theory oI articulation, when wed to the designations oI Iirst and second articulations
at the level oI morpheme/phoneme only, presents limitations that are particularly acute when
applied to the study oI writing systems. It does not account Ior the ability to combine
meaningIul units oI diIIerent hierarchical levels and to allow them to Iunction, within a particular
syntax, as equal to other elements that include a greater or lesser number oI hierarchical levels.
For instance, a noun phrase containing several words, each oI which is independently
meaningIul, still operates as a single part oI speech and can, as such, be paired with a single-
word verb to make a complete sentence as in the case oI 'The jogging man Iell. This Iact is
represented in the tree diagrams oI linguists. Hjelmslev reIers to elements that operate at the
same hierarchical level as being oI the same degree and points out that identical elements may
diIIer in degree depending on context: '.an entity can sometimes be oI the same extension as
an entity oI another degree. (1963: 43). Eco has noted that phonemes, themselves, may be
analyzed into component parts, their combination into the recognizable phonemes oI a language
being governed by an s-code (1984: 169-179). This has particular relevance to the study oI
writing systems. Consider the Iollowing example Irom the historical development oI the Latin
alphabet as related by Stephen Roger Fischer in his Historv of Writing:
In the third century BC, Rome`s Iirst headmaster oI a private school, Spurious Carvilius Ruga,
observed that the Roman alphabet needed a /g/, so he took the Etruscan C and gave it a hook
G to supplement their alphabet with this sound. In this way, Ruga voiced` the Roman C with
a single stroke, displaying his recognition that the only diIIerence between the two sounds was a
voiceless (C)/voiced (G) contrast. (2001: 142)
Now, suppose the innovation Ruga introduced was adopted as the convention by which all
voiced consonants were graphically distinguished Irom their otherwise identical unvoiced
counterparts. The addition oI the 'hook would take on the meaning 'voiced in reIerence to the
9
It is important to remember that Martinet explicitly stated that the study oI writing was outside the realm oI
linguistics, though having obvious relations to it (Martinet 1964: 17). He clearly understood that writing operates
according to diIIerent rules than does spoken language. This is due to the one-dimensional nature oI spoken
language, on the one hand, and the (generally) two-dimensionality oI writing (as per Harris and in contradiction to
the characterization by Coulmas, above). This two-dimensionality can be Iurther expanded by the possibility oI
mixed systems, like musical notation, where a single mark may simultaneously signiIy as a position in a chart and as
a symbol in a script (more on this in section 3).
44
other phonetic attributes indicated by a basic letter-Iorm. Note, also, that the mark would be
irrelevant in respect to vowels, as they are all voiced and have no unvoiced counterparts. A
repeatable marking is here used to carry a particular meaning in reIerence to the larger symbol to
which it is applied. It is not merely a distinguishing characteristic, it is a distinguishing
characteristic that takes on a deIinite meaning at each instance oI its appearance and, thereIore, is
entitled to be called a plereme.
In accommodating these observations, we can adapt Martinet`s terminology to indicate a
moveable distinction that deIines a level oI description pertinent to a given discussion. By way
oI example consider the sense oI the statements 'Please sit down and 'Have a seat. We
identiIy the Iirst articulation (in the sense I am proposing) at the level oI the proposition and the
second articulation at the level oI the words. At this level of description, the individual words are
not meaningIul. It is the meaning oI the proposition that is relevant. This is why the two
propositions can have the same meaning without sharing any words. We can say that an s-code,
Ior a given level oI description, will govern the Iormation oI elements oI Iirst articulation Irom
those oI second articulation. Furthermore, these elements oI second articulation will, by
deIinition, be oI the same degree and will comprise an inventory. We can now state that
'degree is the status oI an element as being oI Iirst or second articulation at a given level oI
description. The inventory will be the collection oI elements oI second articulation at a given
level oI description. Members oI this inventory may be compound, Iormed in whole or in part
by units that are meaningIul (pleremes) at another level oI description, and they or their
constituent parts may Iunction as members oI other inventories (oI elements at diIIerent degrees).
Similarly, an element within one oI these compounds may also be a member oI the same
inventory as that compound. This is a technical explication oI what we have already seen in the
example oI the noun clause. Each word within the noun clause can serve as some part oI speech,
just as the clause itselI does. So, the inventory oI elements that Iunction as some part oI speech
will include 'the running man, 'the, 'man, and 'running. The subclass oI that inventory,
nouns, will include 'the running man and 'man.
The last piece oI this puzzle, prior to turning our attention to musical notation, is a
consideration oI the relation between semantics, syntax, and s-codes. For a syntactical structure
to be present it is necessary Ior the elements oI the inventory to be deIined according to their
Iunction within that syntax. This is not tantamount to assigning a semantic reIerence. The
semantic reIerence is essential Ior meaning but not Ior the proper Iunctioning oI syntax (hence
the possibility oI grammatically correct nonsense). These Iunctions are assigned to subclasses
45
within the inventory oI parts oI speech, as seen above. The s-code is broader still, as it
determines the disposition oI potentially (but not necessarily) meaningIul units within a
particular system. We might say, then, that linguistic syntax is the ordering oI the subclasses oI
Iunctional uses oI morphemes and certain combinations oI morphemes operating as single parts
oI speech, so syntax is a particular sort oI s-code. In the case oI alphabetic writing, an s-code
governs the concatenations oI alphabetic characters (graphemes) in the Iormation oI words
(morphemes). In relation to musical notation, an s-code provides the convention Ior the proper
Iormation oI symbols in the script. For instance, it identiIies the locations, in reIerence to a
notehead, at which we may place a stem, but it does so without any concern as to whether that
stem indicates anything. At another level oI description, a diIIerent s-code governs the proper
Iormation oI a sequence oI notes. At yet another level oI description, another s-code rules the
proper Iormation oI a sequence oI groups oI notes chords within a harmonic system. As
chords can be said to IulIill Iunctional roles in respect to a system oI tonality, we can state that
this last s-code is syntactic.
Where a symbol violates the s-codes normally in operation in the sort oI text in which it
appears, the correlation between expression and content is hindered or broken. Such hindrances
can be introduced intentionally, so as to require an active engagement with the text on the part oI
the reader. It is precisely this type oI symbolic production
10
that we consider when we turn to
'graphic scores. In deciphering such works, the interpreter will need to reIer back to the
physical Iorms oI the symbols, themselves, as well as their arrangement in graphic space, in
order to Iind a basis Ior the interpretation. BeIore presenting a method Ior that undertaking, we
should examine the workings oI standard notation, employing the terminology developed in this
section, as 'graphic notations oIten deliberately play upon aspects oI standard notation.
3. Conventional Musical Notation as a Writing System
As with linguistics, it is almost always the sounded aspect oI music that is considered in analysis.
When applying the methods oI linguistic and semiotic analysis to music, this presents the
problem oI reIerence
11
. This problem does not arise, however, when we discuss musical notation
10
This sort oI symbolic production is reIerred to as 'invention by Eco (1976: 245).
11
Nicolas Meeus, Ior instance, has circumvented the problem oI reIerence by introducing the notion oI 'analytical
pertinence (Meeus, 2002: 166) to replace 'meaning. An intuitive move oI the same sort seems to be made by
many music theorists, linguists, and cognitive scientists (see Patel (2008)) who have compared the operations oI
music and language, arriving at the Iormulation: music has syntax but no semantics. II we understand 'analytical
pertinence to mean, 'having a Iunction in respect to the other elements oI the same degree, then we see that it is
precisely analytical pertinence that makes musical syntax a possibility.
46
because the notated symbol does, in Iact, have a reIerent, in the traditional sense
12
. This will
allow us to circumvent certain observations oI musical semioticians: 'Notes cannot be Iurther
segmented, but they can be Iurther analysed into what linguists would call their distinctive
characteristics` their parameters. Neither pitch nor duration can be realised in themselves
outside the note to which they belong, oI which they are qualities... (Meeus 2002: 164). Such
statements reIlect the same Iocus on the sequential processing oI sound-as-heard evident in the
work oI Martinet and other linguists. We have already seen, with respect to language, that a
diIIerent principle is at play within a writing system one that unIolds in a multi-dimensional
expression-plane. And so, we will Iind that those very properties, which Meeus calls
unrealizable in themselves when considering musical sound, can be indicated by distinct musical
notations. For example:
A duration has clearly been indicated, though we have no designation as to what sort oI sound
should instantiate that duration. But, oI course, Meeus is right to claim that duration cannot be
reali:ed in itselI because there is no physical representation oI these properties oI sound except
as a constituent element oI a sound. In philosopher`s terms, this is the distinction between the
intension oI the note and its extension between the properties constituting the notes and the
class oI realizations oI the note as notated. However, in musical notation, we have markings that
signifv these intensional properties. It is these marks that we study when we study music as a
writing system. The physical maniIestations oI these markings constitute the extension oI the
pleremic units comprising the writing system (musical notation). The identity oI a marking as a
member oI some pleremic content-class connects that marking to its signiIication some
intensional property oI musical sound within the notational system.
The means oI indicating these properties within the graphic space are multiIorm, as
Harris points out:
. notes are diIIerent Irom one another in shape. and their sequence relative to one another is
signiIicant, but the location oI each in the graphic space provided by the Iive-line stave,
Iunctioning as a chart, is also essential to the musical message. In order to read` music, we
have to deploy two diIIerent processing techniques simultaneously. (1995: 94)
12
This reIerent seems to be some combination oI a description oI a sound to be realized and an indication oI a
perIormance technique to be employed. These descriptive and prescriptive aspects oI notation have led to
interesting arguments regarding the Iunction oI the score in respect to the ontology oI the musical work. See in
particular Goodman (1968), WolterstorII (1975), and Levinson (1980).
47
Indeed, I am unaware oI any system oI writing that is as thoroughly mixed as musical notation.
This suggests that we should identiIy what is signiIied by the placement oI various notational
markings, and what is signiIied by the script elements (the shape oI these notational markings).
Immediately, we note that pitch is designated by the chart-aspect oI musical notation and
duration and sequence are determined via the musical script. As Gardiner Read states in his
manual oI notational practice:
Properly written, a musical note. indicates without question two aspects oI musical sound. It
is Iirst a symbol indication by its position on the staII and by the cleI used a deIinite pitch to
be played or sung. And second, it establishes by the exact appearance oI its three integral
parts the relative time duration oI this musical sound. (1979: 63)
This basic division oI the chart and script aspects oI musical notation has deep historical
roots going back perhaps as Iar as the ninth century (Levy 1987: 61). The earliest notation
indicated pitch contours, correlated to script Iorms, to be applied to syllables in the recitation or
chanting oI liturgical texts. The leIt-to-right sequencing oI events, then, is directly derived Irom
Latin writing. These early neumes served as mnemonics to aid perIormance, only. There was no
indication oI the pitch relationship between successive syllables, and the contours did not
represent exact intervals. It was not until about the tenth century when neumes started to be
placed in vertical relation to one another according to pitch-height, thus introducing the Iirst
chart-based notation (Treitler 1982: 244). As the chart-based representation oI pitch became
increasingly reIined with the introduction oI the staII (Iirst a single line, then having lines added
one by one) and the cleI, the original neume Iorms were simpliIied. The speciIicity oI the chart-
based pitch notation had made the shapes oI their script Iorms redundant, allowing that script
Iorm to accrue other meanings distinct Irom the designation oI pitch. The adoption oI script
Iorms to convey duration was Iirst Iully codiIied in the Franconian notation oI the thirteenth
century. Now speciIic note-Iorms signiIied speciIic durational values (Parish 1957: 110). It was
in the Baroque era that our modern notation was codiIied (Parish 1957: 195). As musical
practice demanded new indications, as with the shiIt Irom the modal to the tonal system, script
elements were added. Accidentals, Ior instance, were introduced into notational practice to
indicate harmonic relationships oI the pure IiIth, an innovation derived Irom the perIormance
practice oI early polyphonic music in which (unnotated) musica ficta where used to avoid
tritones notated in the score. In other instances already existing Ieatures oI notation were used to
make new distinctions. This led to a situation in which many oI the markings contributing to the
48
Iorm oI the musical note have reIerence within one or both realms oI notation (chart and script).
Both oI these realms have their own s-codes or set oI s-codes. In the case oI the chart-aspect, the
matter is relatively simple: the staII and cleI create localized gradations oI pitch height, positions
upon which are marked by noteheads. Within a multiple-staved system, there are as many
localized gradations as there are staves. However, the staves oI a system share a localized axis
determining the sequence oI events, which runs leIt to right Irom the beginning oI one system to
the next, the order oI systems being sequential Irom the top oI the page to the bottom. The
various s-codes operating at diIIerent levels oI the script is a more complicated matter.
We can identiIy distinguishing marks constituting a given note-Iorm (notehead, stem,
Ilags or beams) and analyze the Iunctions they serve in the diIIerent planes oI notation, the chart
and the script. Let us take, Iirst, the notehead. Whether or not the notehead is Iilled-in or hollow
is immaterial to its ability to signiIy the property oI musical pitch as represented graphically on
the staII. Furthermore, nothing in the shape oI the note, itselI, has any reIerence in respect to
pitch. As script, however, it is the placement on the staII that is immaterial. The exact Iorm oI
the notehead is one oI the identiIying characteristics oI a given durational value indicated by the
script. Considering the Iilling or emptiness oI noteheads within the script-based aspects oI
notation, does not, in itselI, mark any signiIicant diIIerence. In Iact, we see Iour diIIerent
notehead Iorms: hollow with double lines on either side, hollow, hollow and slanted, Iilled and
slanted. The slanted distinction, however, is too easily lost particularly in handwritten scores
to be trusted as a clear distinguishing characteristic, so we Iind the stem attached to instances oI
the slanted notehead (either Iilled or hollow as in the quarter note or the halI note). The stem,
itselI, does not represent any alteration in the durational property oI the note indicated by the
note-Iorm. Indeed, the stem is redundant as a distinguishing mark, serving simply to aid in the
distinction oI slanted note-Iorms Irom their non-slanted counterparts: as there are two attributes
oI the note Iorm that contribute to its identity as halI note or whole in respect to one another (the
slanting or not oI the notehead and the presence or absence oI the stem), neither attribute alone
can be said to carry durational meaning (hence making them pleremes); they can only be said to
be distinguishing marks (cenemes). The status oI the stem as ceneme is Iurther reinIorced by the
Iact that its presence carries no inIormation in making the distinction between a halI note and a
quarter note, Ior instance. The Iilling-in oI the notehead serves as the distinguishing
characteristic between our halI and quarter notes. Why, then, does the note require a stem, iI the
Iilled-in notehead is a clear enough distinction on its own? Durations have always been
conceived as proportionally related and have proceeded Irom a smaller number oI distinctions to
49
a greater number oI distinctions over time. New note-Iorms, marking smaller temporal divisions,
were oIten created by adding additional distinguishing marks to the next larger (or when creating
larger units, the next smaller) temporal unit. The intricacies oI the Renaissance mensuration
system make the identiIication oI singular mathematical relationships between these successive
temporal divisions impossible each level oI the hierarchical system could be divided into two
or three units in a three-tiered system
13
. Also, the distinguishing marks were not repeated. Each
new note-Iorm carried a unique distinguishing mark.
This changed with the adaptation oI the Ilag Iirst introduced as a distinguishing mark
added to the minim to Iorm the semiminim (Parish 1957: 144) as a means oI indicating a
halving oI the value indicated by the rest oI the note-Iorm. This is only applied to a certain
subset oI note-Iorms in musical script, namely those with Iilled-in noteheads (and stem). The
resulting note-Iorms indicate a relation to some duration, d, as Iollows:
double whole note: 2d
whole note: 1d
halI note:
1
/
2
d
quarter note:
1
/
4
d
eighth note:
1
/
2
(
1
/
4
d)
sixteenth note
: 1
/
2
|
1
/
2
(
1
/
4
d)|
thirty-second note:
1
/
2
(
1
/
2
|
1
/
2
(
1
/
4
d)|)
sixty-Iourth note:
1
/
2
|
1
/
2
(
1
/
2
|
1
/
2
(
1
/
4
d)|)|
one hundred twenty-eighth note:
1
/
2
(
1
/
2
|
1
/
2
(
1
/
2
|
1
/
2
(
1
/
4
d)|)|) .
To the objection that the marking oI the Ilag is merely a distinguishing element (a ceneme) in
musical script, as with the various Iorms oI the notehead, I answer that the repeatability oI the
Ilag always carrying the same meaning, which is easily expressed as a mathematical Iunction,
makes this mark a plereme. The Ilag Iunctions much like the dot, which is also additive, and can
also be expressed as a mathematical Iunction: d
1
/
2
d Ior one dot, d
1
/
2
d
1
/
2
(
1
/
2
d) Ior two, etc.
The situation becomes very similar to the hypothetical on Ruga`s introduction oI the hook to
Iorm the grapheme, G, discussed in section 2. It will be remembered that the hook expresses the
phonemic distinction 'voiced when applied to consonant graphemes (letter-Iorms). This creates
a subclass oI graphemes that are compounds oI pleremes. Our musical analogy to the
hypothetical on Ruga`s innovation would make the notes with open noteheads analogous to the
13
See Anna Maria Busse Berger (1993) Ior a detailed study oI the origins and development oI the symbols used to
indicate the mensuration oI early music.
50
vowels, the closed noteheads analogous (in reIerence to their compound nature) to consonants,
those with Ilags being analogous (to a point) with the voiced consonants. The compound script-
Iorm oI the eighth note Iunctions as a unit oI the same degree as a whole note, though the whole
note is not a compound symbol. They are both meaningIul at a shared level oI description.
We have reviewed, up to this point, the Iunction oI the notehead in both the graph-based
and script-based dimensions oI musical notation, and the Iunction oI the stem and Ilag in the
script-based dimension. The distinct characters oI the two dimensions oI the mixed writing
system oI musical notation lead to the immediate intelligibility oI (by now conventional)
unconventional notations in which, Ior instance, the Iilled notehead without stem or Ilag is
placed on a staII to indicate pitch without any speciIied rhythmic content. Or, on the other hand,
when we see stems and Ilags absent noteheads as in some notations Ior spoken word rhythms
or lute tablature, or where we have x-shaped noteheads attached to stems (with and without
Ilags) to help visually diIIerentiate metal instruments Irom others in compound percussion parts
we have no problem assigning durational values to the second tier oI script-based note-Iorms
(quarter notes and smaller values). This is reinIorced by the Iact that multiple noteheads may be
attached to a single stem and that two stems may be attached to the same notehead the graph-
based designation oI the stem as identiIying which voice a pitch belongs to in instances where
more than one voice is notated on a single staII. While the stem serves only as a ceneme,
marking a distinction between certain note-Iorms, in the script-based dimension, it serves this
wholly diIIerent (pleremic) purpose (at times) in the graphic dimension.
I alluded, above, to the introduction oI script elements to accommodate changes in
musical practice. The use oI accidentals, Ior instance, leads to two diIIerent indications
regarding diIIerent aspects oI pitch perception: pitch-height, designated (roughly) by the chart-
based aspects oI musical notation, and harmonic relationships, designated by the script-based
aspects oI musical notation (in combination with the chart aspects). Articulation markings are
attached to particular note-Iorms as additional descriptive inIormation about the intensional
properties indicated by the composite note. Accidentals and articulation markings both tend to
be simple, visually, and cannot be Iurther analyzed, as script, into lower level pleremes. They
can, however, be compounded. I won`t concern myselI with their structure here, but rather with
their combination with note-Iorms to create a composite whole. There is a rule Ior the proper
combination oI these markings with a note-Iorm they are members oI an inventory that are
combined according to an s-code. We may have a single note-Iorm, one appended with a
staccato marking, and another with an accent and a sharp sign, all on a single staII. We reIer to
51
all oI these as a singular note, and the note is the unit oI Iirst articulation at this level oI
description. Each oI the units oI second articulation the note-Iorm, the accidentals, the
articulation markings, and the staII position oI each serve one or more oI three Iunctions: a
descriptive indication oI some property oI sound, a prescriptive indication to undertake some
action resulting in a sound, or the modiIication oI some indication oI another marking within the
notation. The note-Iorm and the notehead, Ior example, are members oI the class oI descriptive
Iunctions. Tablature notation and Iingering indications are members oI the class oI prescriptive
notation. The Ilag, dot, and accidentals (when considered in respect to pitch-height)
14
are
members oI the class oI modiIicatory Iunctions. These Iunctional classes allow us to describe a
syntax (as a special kind oI s-code) Ior the Iormation oI notes at the level oI description at which
we are operating. We see that the main descriptive elements are indicated by the primary note-
Iorm and its placement on the staII, with secondary descriptions being located immediately
above or below the notehead (articulation markings). This is also the location Ior certain
prescriptive indications, such as Iingering markings, and the status oI articulation markings as
unequivocally descriptive (rather than prescriptive) is in no way secure. ModiIiers occur to the
leIt oI the notehead, in the case oI those applied to pitch (accidentals), and to the right oI the
notehead or the stem iI applied to rhythm (the dot and the Ilag, respectively). We are able to give
the description above because each element oI the inventory can be assigned a Iunctional class
membership. We cannot speak oI a syntax Ior the Iormation oI note-Iorms Irom the inventory oI
notehead, stem, and Ilag, because these are not all Iunctional units as a ceneme, the stem does
not serve a Iunction in respect to the other parts oI the script. We can, however, speak oI the s-
code Ior the Iormation oI note-Iorms, as the s-code has no requirements regarding reIerence or
Iunction. Enough has been said in relation to standard notation Ior present purposes. Let us now
14
A detailed consideration oI the prescriptive and descriptive aspects oI notation, and certain conditions Ior the
acceptance oI innovations in these areas has been oIIered recently by Mieko Kanno (2007). UnIortunately, the
discussion is marred by the absence oI any criteria by which a notation can be assigned an unequivocal identity as
either descriptive or prescriptive and by a too ready acceptance oI the idea that any given notation is wholly one or
the other. Leo Treitler was closer to the mark when he stated that '. in varying degrees, people read even staII
notation |as prescriptive|, translating the signs directly into Iinger movements, (Treitler 1982: 241). It is likely that
diIIerent musicians Iall in diIIerent locations upon a spectrum in which notations are considered, at one end oI the
continuum, wholly descriptive and at the other, wholly prescriptive. Among the Iactors that would seem to play into
an individual`s position on this spectrum Ior any given notation are the quality oI one`s ear, the instrument one plays,
and the use to which one puts the notation. A pianist is probably more likely to conceive staII notation as
prescriptive as there is only one action that instantiates a given pitch within the instrument`s range. A trombonist, by
contrast, can play certain notes in the middle to upper register in a wide variety oI slide positions, suggesting a more
descriptive interpretation oI staII notation. Certainly, though, some notations, such as tablature, are unequivocally
prescriptive, some notations are unequivocally modiIicatory, but descriptive notations can always be construed, by
the perIorming musician, as prescriptive. The implications Ior this Iact bears Iurther examination.
52
turn our attention to the synthesis oI the inIormation presented in this and the previous section
into an approach to the analysis oI any even the most radically idiosyncratic musical
notation.
4. Outlining an Analytical Approach to Alternative Notations
The dominant conventions treating the dimensions oI the graphic space, in a particular class oI
texts, may be expected to be operative where one expects the chart aspect oI a text to be
meaningIul and where no marking indicates otherwise. In the case oI musical texts oI this type,
we can expect the deIault treatment oI graphic space in two dimensions to be a temporal
progression Irom leIt to right along the horizontal axis and a pitch height continuum along the
vertical axis. In certain instances, such treatment may be insuIIicient and recourse to other
methods oI ordering the elements will be preIerable, such as those used in the analysis oI visual
art (in reIerence to composition, color, etc.) and the psychology oI visual perception
15
. In other
instances, the composer will designate unconventional parameters attached to each dimension oI
the graphic space.
The case is much diIIerent with the markings contained within that graphic space (the
script-based elements oI a writing system), which may be independently identiIied as pleremes
or cenemes. These might be Iurther arranged, at various levels oI description, into compound
symbols. The analysis should begin by 1) determining what parameters are to be assigned to the
two-dimensional axes oI the graphic space construed as a chart where there is no internal
indication suggesting otherwise, it would be wise to Iollow the standard temporal and pitch
continuum interpretations oI the graphic space. Then, beginning with the lowest perceived level
oI description and repeating these steps Ior each subsequent level oI description, 2) identiIy the
s-code Ior the script-based aspects oI the notation (i.e. the apparent logical structure governing
the Iormation oI the symbols at the given hierarchical level); 3) determine which marks are to be
treated as pleremes and which are to remain cenemical at the present level oI description. Once
all levels oI description to be addressed are deIined, proceed as Iollows Ior each, revising the
analysis as necessary: 4) assign Iunctional class designations to pleremes (descriptive,
prescriptive, modiIicatory) in the script- and/or the chart-based aspect(s) oI the notation; 5)
propose a syntax existing between the Iunctional classes, once assigned (this will require that all
members oI the inventory under consideration be pleremes, though their exact reIerence, beyond
15
See, in particular, Rudolph Arnheim (1974), Art and Jisual Perception. A Psvchologv of the Creative Eve
(University oI CaliIornia Press).
53
membership in a Iunctional class, need not yet be assigned); 6) propose semantic reIerences (in
the script- and/or the chart-based aspect(s) oI the notation) that will not lead to nonsense when
building outward Irom the lowest hierarchical level containing pleremes (i.e. the lowest level oI
description). In practice, this step may precede or coincide with step 4. When attempting to
assign semantic meaning to pleremes, Iollow the course: (i) identiIy all markings that have
equivalences in conventional musical notation correlated to some musical meaning
16
; (ii) identiIy
marks, designated as pleremes, that do not have equivalents in conventional musical notation but
which may be identiIied as meaningIul within some other conventional system oI writing; (iii)
identiIy the remaining marks that are to be treated as pleremes; (iv) build a catalog oI candidate
meanings Ior each mark, drawing on the results Irom steps (i) and (ii), where possible, and
checking the candidate designations against one another within the perceived syntax Ior
intelligibility; (v) once having eliminated nonsensical candidate designations, determine some
preIerence criteria by which one set oI coherent designations is selected; (vi) assign semantic
reIerences to pleremes according to the preIerence criteria oI step (v). OI course, in the majority
oI cases, the composer has already provided much oI this inIormation in the perIormance
instructions, substantially lightening the work load oI the analysis.
The progressive sample case, to which we now turn, is not intended as an exhaustive
instance oI the approach detailed above. Rather, it serves as an illustration oI certain aspects oI
this approach as they arise when attempting to assign musical meanings to unconventional
notations when they are not accompanied by clariIying perIormance instructions. As such,
reIerence will be made to the speciIic steps, introduced above, as they appear.
4.1. A Progressive Sample Case
Imagine a graphic space oI undiIIerentiated background on which outlines in black oI equal-
sized diamond-shapes are arranged (ex. 1). There is nothing to distinguish one Irom the next
except Ior its placement within the graphic space. II we are to interpret this as a text, we will
need to look to the spatial distribution oI the diamond-shapes as the potential locus oI meaning.
There is no diIIerentiation amongst the Iorms oI the constituent symbols, so any potential
meaning is conveyed through the coordinates oI the mark`s location in the graphic space a
16
For instance, in an eighth note in which the Ilag is placed adjacent to the un-stemmed, Iilled notehead, we can still
identiIy the usual meanings oI each component present: notehead marking a vertical position within a gradated pitch
height spectrum, the Ilag halving the durational value oI the rest oI the notation. This example possesses a deviant
morphology that violates the s-code Ior the Iormulation oI a proper note (in musical script), yet we are still able to
incorporate the musical correlations oI the pleremes present in whatever interpretation we ultimately give to this
symbol.
54
prime example oI a chart. It is as though the text indicates 'x here at every instance oI a
diamond-shape and 'x not here at every other location. All that would be required to give the
text a concrete meaning would be the identiIication oI the parameters to be correlated to each oI
the two dimensions oI the graphic space (step 1 oI the preceding section).
Example 1
Imagine now that some oI those diamond-shapes are Iilled in. Having introduced a
distinction in the script Iorms, we can address step 3 (the designation oI which marks oI the
script are to be construed as pleremes and which are to be construed as cenemes), in which
cenemes/pleremes will be isolated through the perception oI diIIerences that are taken to be
substantial enough to constitute some new class or subclass oI marks in relation to the others.
The text now indicates 'x here Ior all unIilled diamond-shapes, 'v here Ior all Iilled diamond-
shapes, and 'neither x nor v here Ior all other points within the graphic space. These marks take
on the qualities oI a plereme and so may be assigned a Iunctional and a semantic reIerence in the
later steps oI our analysis. A vertical line descending Irom the lowest point oI the diamond-
shape, iI occurring on each diamond-shape and iI equal in length, color, width, etc., makes no
new distinction between the (potential meanings oI) symbols within the graphic space adding
these always-equivalent marks does not change the Iormulation 'x here, 'v here, 'neither x nor
55
v here (ex. 2). Indeed, it is the composite oI the diamond-shape and the vertical line that now
Iorms the plereme. This plereme is constructed oI two marks, which will be considered cenemes
at this level oI description. However, where some diamond-shapes lack this added mark while
others possess it or where these vertical lines are not uniIorm throughout, the vertical line
becomes at least a distinguishing mark and at most a plereme that, combined with the diamond-
shape, creates a compound symbol.

Example 2
The reader may have noticed that we have proceeded Irom step 1 to step 3, the reason Ior
which is that the marks discussed at the outset oI the example are not (perceived as) reducible to
subordinate parts. Now that we have introduced a second mark, we can approach step 2, on the
s-code governing the combination oI the discrete marks oI the script into compound symbols.
On this Iront, though this example is still quite simple, we can note that diamond-shapes are
permissible as Iree-standing entities while vertical lines only appear as components oI a
compound symbol. Furthermore, according to the description above, vertical lines only appear
as descending Irom the lowest point oI a diamond-shape. So, iI we accept descending vertical
lines oI uniIorm length Irom either Iorm oI diamond-shape and those two diamond-shapes,
themselves, we have Iour distinct classes oI symbols deIined by this s-code. While it is possible
(in a return to step 3 with this revised example) to treat these Iour classes as indicating x, v, p,
and q, the repetition oI the pleremes indicating 'x and v within the markers designating 'p and
q seems to suggest that the more appropriate way oI understanding the Iour classes oI symbols
would be 'x (without p), 'v (without p), 'x with p, and 'v with p. My preIerence Ior this
treatment derives Irom the identiIication oI the line and the diamond-shapes as discrete elements
oI the same hierarchical level, which appear as such by their combination according to an s-code
56
at the level oI gestalt perception
17
. II we consider an alteration to our example in which the line
exists as a Iree Iloating mark, not always attached to a diamond-shape, then the diamond plus
line suggests a compound proposition ('x and p). Where the line only appears appended to a
diamond, it may still be meaningIul as a compound proposition, or it may be construed as a
modiIication oI the proposition expressed by the diamond-shape, resulting in a new, singular
proposition.
The role these pleremes take in such a proposition will comprise their Iunctional
designation, the subject oI step 4. II, as Iirst proposed, we consider each plereme to be an
independent proposition, we can ascribe a Iunction to each, as seen in section 3, whereby each
represents some intensional property oI a sound, whether that be an intensional property
regarding the qualities oI that sound (descriptive) or its manner oI production (prescriptive).
When we accept the vertical line as a modiIier, it modiIies whatever intensional property is
assigned to the diamond-shape (Iilled or un-Iilled). For example, we might construe the line (or
its absence) as indicating a duration, the diamond-shapes as indicating two diIIerent timbral
qualities. In this case, the line and the diamond-shapes perIorm descriptive Iunctions.
Alternatively, we might decide that the presence or absence oI a vertical line designates which oI
two instruments or instrumental groups should perIorm the sound, while the diIIerent diamond-
shapes indicate that the sound should be either loud or soIt. In this case, the vertical line is
prescriptive, the diamond-shapes descriptive. II the line Iorms a modiIicatory role let us say
that it indicates that the designation oI the diamond-shape is to be exaggerated and the
diamond-shapes perIorm the descriptive Iunction oI designating a loud or soIt dynamic, we will
get a result, not oI compound propositions, but oI Iour distinct singular propositions indicating
one oI Iour dynamic levels. Once such Iunctions have been deIined, a syntax regarding the
combination oI these Iunctional symbols into a composite text can be proposed (step 5, which
will be addressed once the example becomes complex enough to sustain this aspect oI the
17
In other words, prior to considering any ascription oI meaning, the diamond and the vertical line seem to be,
visually, oI the same hierarchical level we see them each as selI-contained (irreducible) and complete in
themselves. As such they comprise an implied inventory that can be combined by an s-code into higher level
structures (diamond-shape with vertical line). Where one element oI some inventory is taken as meaningIul and
where we do not possess a code Ior the ascription oI meaning to each mark, I preIer to take a liberal stance, allowing
the possibility oI meaning to elements oI the same level. This is not always the way things work in notational
systems it certainly is not in the case oI musical notation but it is a possibility in a case where we are attempting
to derive the code Irom a document without any known convention behind it. The liberal stance is taken out oI a
desire to present the richest possible interpretation oI the text. Should a substantial statistical sampling oI the
markings in the text strongly indicate that an element is a ceneme, by its relations to other markings, even though it
combines with pleremes at some level oI description, then this liberal stance should, oI course, be abandoned.
57
discussion).
Now, iI we take this grouping oI diamond-shapes and vertical lines as a member oI the
class oI texts 'musical scores, we have recourse to the code oI musical notation in correlating
the pleremes and symbols contained therein to musical meanings (step 6.i.) and notational
Iunctions (step 4). For instance, we may note that the current usage oI diamond-shapes in
musical notation is (primarily) to designate a nodal point along a string at which a Iinger is
lightly placed so as to produce a harmonic. A graphic space oI undiIIerentiated background with
some distribution oI unIilled diamond-shaped Iigures (ex. 1) understood as a musical score may
then be interpreted as demarcating nodes oI varying pitch heights along the length oI a string.
Notice that the addition oI the designation 'musical score suggests two things: Iirst, the marks,
even though undiIIerentiated, can be ascribed a meaning derived Irom the traditional musical
script; second, the chart`s dimensions can now be interpreted according to the s-code oI the
chart-based aspect oI standard notation
18
. The addition oI Iurther pleremic content, such as the
staII (ex. 3), also taken Irom the inventory oI symbols used in musical notation, will alter the
interpretation by introducing new elements that must either be treated traditionally or not,
possibly to the exclusion oI treating some other element in the traditional manner. Whatever
semantic designations are ultimately derived Irom standard practice will depend on which
aspects oI that practice are perceived as the least violable.
18
These dimensional designations are hardly unambiguous, particularly in reIerence to the vertical axis, which is
conventionally correlated with pitch height in a musical score but may also, in this example reIer to some placement
along a string. Nodes do not produce pitches the pitch height oI which corresponds to a shortening length oI string
as the Iinger 'ascends along the string. They are related to proportions oI strings: the 5
th
partial will sound at any
node dividing the string into some number oI 5
th
parts. So placing one`s Iinger according to the height indicated by
the vertical axis oI the text will not (necessarily) produce pitches that express the same contour, resulting in possible
cognitive dissonance between sound and symbol one reason that sounding pitch is oIten notated in parenthesis
above the nodal designation. An alternative interpretation would be to produce harmonics that do represent this
contour.
58
Example 3
Accepting the lines, added to example 3, as a musical staII, a systematic demarcation oI
pitch height or diatonic pitch identity can be understood to have been introduced, though there is
no deIinite reIerence (a cleI) anchoring this demarcation. An ambiguity as to whether it is pitch-
height or node that the notation indicates remains, though it might be clariIied (Ior instance)
through a designation oI a speciIic string, on which each node is placed. This designation may
be made (according to convention) by connecting the diamond-shape to a standard notehead
below it by a vertical line or by placing a Roman numeral immediately above or below the
diamond-shape, indicating a natural harmonic on one oI the numbered strings oI some instrument
such as the violin, viola, cello, bass or guitar. This last additional mark (the Roman numeral),
thereIore, implies an instrumental designation (and serves a prescriptive Iunction). Where beIore
we were thinking in terms oI 'some string we are now thinking in terms oI 'some instrument
containing some number oI strings, at the very least, and are implying, by the numerals
indicated, the minimum number oI strings the instrument must have. Should a particular
instrument be decided upon, a host oI considerations leading to possible revisions oI the
interpretation, based upon the perIormance practices and notational conventions associated with
that instrument, are introduced. These would include the inIerence oI some tuning relationship
between strings, which will drastically aIIect the resulting sound oI our interpretation.
II we proceed, not by designating particular strings but by drawing vertical lines
descending Irom the lowest point oI each diamond-shape oI example 3, we obscure the musical
signiIicance oI the marking '^ and introduce a symbol with no conventional musical meaning,
initiating steps 6.ii. and 6.iii. Should only some diamond-shapes contain this additional mark,
we might interpret it as some alteration oI the nodal interpretation a plereme serving an as yet
undeIined modiIicatory Iunction in relation to the prescriptive Iunction oI the diamond.
59
Assuming that some have Ilags, as in a conventional eighth-note, at the end oI their stems (ex. 4),
the diamond reasserts itselI as a plereme, as interpreted beIore, while the vertical line becomes
the musical notation, 'stem (here we have completed step 6.vi., circumventing 6.iv. and 6.v. Ior
the sake oI brevity). The stem, no-stem, stem-and-Ilag conIigurations take on conventional
rhythmic meaning, granting them the status oI plereme constructed oI these elements according
to an s-code resembling (though not identical to) that which governs the construction oI note-
Iorms Irom stem, Ilag, and notehead in traditional notation. We can read the rhythmic meaning
oI these new symbols because the stem and Ilag distinction can convey rhythmic meaning, as
pleremes in musical script, absent any notehead or reIerence to the chart-aspects oI musical
notation, as seen in our discussion oI lute tablature, etc., in section 3. As the stem, itselI, is not,
generally, a plereme but only becomes one in such headless notations, the presence oI the Ilag,
which is a plereme in standard notation, upon some stem is necessary Ior the establishment oI the
rhythmic meaning oI the isolated stem within the notation.
Example 4
II we Iill some oI these diamond-shapes seen in example 4, we upset our conventional
interpretation, as the nodal pleremes depend on an open diamond-shape. Yet, this Iilling oI some
oI the diamond-shapes may bring to mind the white notation oI the 15
th
and 16
th
centuries,
particularly iI applied only to stemmed note-Iorms and always occurring in the case oI stemmed
and Ilagged note-Iorms. II, as a Iurther variation on the example, we introduce the option oI a
halI-Iilled diamond, some oI our symbols will resemble certain note-Iorms employed in Italian
Quattrocento notation, an idiom that Ialls outside the standard knowledge base oI present-day
perIorming musicians. For those that see these aIIinities with historical notation schemes , the
60
note-Iorms oI our example might be interpreted as conveying the rhythmic inIormation ascribed
to their historical counterparts
19
. II, instead, we introduce certain logical operators such as '&
(and), '! (iI. then.), and '" (or) between the various note-Iorms oI example 4, these will
upset the leIt-to-right temporal sequencing oI events Ior those interpreters that know symbolic
logic notation. For instance, two note-Iorms appearing on either side oI '& might be taken to
coincide with one another, whereas two note-Iorms appearing on either side oI '" might be
interpreted as indicating the perIormance oI one, the other or both note-Iorms at a given
moment
20
. In both the case oI symbolic logic and Quattrocento notation, this sort oI designation
oI semantic content derived Irom a writing system other than that oI standard (modern)
notational practice (even when that other writing system is a musical one), is an instance oI step
6.ii.
The example has now taken on suIIicient complexity Ior the consideration oI possible
syntactic relations between symbols (step 5). II the pleremes (diamond, diamond-plus-stem,
diamond-plus-stem-and-Ilag) are identiIied as comprising an inventory oI script elements, and
each is assigned a Iunctional meaning, we can attempt to derive a syntax Ior their combination
into larger meaningIul structures. The same can be done regarding the chart-based indications oI
the markings. II their positions are assigned a Iunction in respect to other allowable positions
within the chart, then a syntax ruling the combination oI larger meaningIul units Irom the
designation made by chart placements can be described. The simplicity oI example 4 still does
not inspire much in the way oI a syntactic interpretation, but on the script level, iI we assign
rhythmic values oI one unit to the diamond, halI a unit to the diamond-plus stem, and a quarter
unit to the diamond-plus-stem-and-Ilag, we could extrapolate the rule: sequential groups oI two
symbols will always have the internal relation d .5d, where d is the durational value oI the Iirst
symbol oI the grouping. Without assigning speciIic durational reIerences, we can state this as:
sequential groups oI two symbols will always have their second element be the next smaller
durational unit Irom their Iirst element.
This example has walked us through many oI the steps oI the analysis, hopeIully
19
See Phillip Schreur (1989), Ior a critical translation oI the Tractatus Figurarum, which details the intricacies oI
ars subtilior notation in Italy in the Quattrocento.
20
The use oI '" is potentially ambiguous as it is also a bowing indication within standard musical notation;
however, the placement suggested (between, rather than over, note-Iorms) and the presence oI '&, which has no
musical counterpart, make the ascription oI meaning based upon the symbolic logic denotations oI these symbols the
more deIensible conclusion. Similarly, once meanings drawn Irom symbolic logic notation are allowed, the
diamond-shape might be interpreted as it is in modal logic, indicating 'it is possibly the case that.; however, the
use oI these diamond shapes within more complex note-Iorms and the absence oI logical quantiIiers within the
notation support a non-logical interpretation.
61
demonstrating the exactitude with which they should be completed. It must be pointed out that
the discussion did not move through various levels oI description. This is a Iunction oI the
simplicity oI the sample case. However, a situation in which the necessity to treat other levels oI
description arises can easily be imagined. Picture a larger graphic space comprised oI an
arrangement oI squares containing the same elements as example 4 but in varying conIigurations.
Now the analysis at the level oI description already undertaken will have to account Ior the
contents oI these additional squares, and a higher level oI description will have to be explicated
in the analysis oI a) those squares` relationships to one another in the graphic space as, possibly,
chart indications and b) whether the individual arrangements oI their contents are to be taken as
compound script conIigurations, which might take on Iunctional and/or semantic meanings.
Other levels oI description can be added, ad inIinitum, by introducing ever larger graphic spaces
in which squares oI the next lower level are arranged in reIerence to one another and/or serve as
compound script elements. And so the discussion oI levels oI description has not been Ior
naught.
OI course, it is usually the case that graphic scores include detailed perIormance
instructions that answer the various steps oI the analysis, assigning musical meanings to the
symbols Iound within the score. OIten some ambiguities remain ambiguities that can be Iixed
with recourse to the method introduced here. The possibility oI evaluating radical notational
experiments, once the notational ambiguities have been resolved (by the application oI our
method, by the perIormance instructions, or by a combination thereoI), will be demonstrated
with reIerence to Cage`s 'Variations I, below.
4.2. The Evaluation of ~Graphic Scores After the Analysis: 1ohn Cage, ~Variations I
The Iorgoing presentation indicates that musical notations (conventional and otherwise) have the
character oI compound propositions designating intensional characteristics oI a sound, whether
descriptive or prescriptive, modiIied or not; and that, owing to the mixed nature oI musical
notation, the propositions will generally be compounds oI indications made by the script- and
chart-based aspects oI the notation. Returning to WolterstorII`s deIinition oI the Iunction oI the
score in reIerence to the musical work, the speciIicity oI the intensional properties indicated in
the notation determines the conditions by which a proper perIormance oI the work (i.e. a proper
member oI its extension) can be identiIied. Where the composer has speciIied instrumentation,
articulation, pitch, rhythm, and dynamic, these are all selected as relevant to evaluations oI the
extension oI the sonic structure notated in the score. Where some oI these attributes are not
62
notated or are leIt ambiguous, we say that these intensional characteristics oI the sound are leIt
indeterminate, leading to a wider latitude Ior a perIormance`s acceptance as a member oI the
class oI a score`s extension. It is through an involvement with such works, particularly those oI
the New York School, that the present study Iound its impetus.
'Variations I is one oI several pieces by John Cage, the scores oI which consist oI a
number oI transparencies that are to be overlaid upon one another in some manner to make
determinations, according to the (rather cryptic) perIormance instructions, regarding the sonic
elements oI the particular perIormance (ex. 5). It is striking that, in using transparencies, Cage
has made, not only the markings contained within it, but also the score, itselI, indeterminate. In
this case, Cage has created six transparencies. One oI these contains an assortment oI dots oI
Iour diIIerent sizes, which are to indicate (based upon their size) the number oI individual events
contained within the larger identity oI that dot. The thirteen smallest are to be individual sounds,
the next tier oI seven dots indicate two sounds, the three oI the second largest size represent three
sounds, and the Iour largest represent Iour sounds. These sounds may be played simultaneously
or as what Cage terms 'constellations, which I interpret as meaning temporally distinct yet in
close proximity to one another. The remaining transparencies each contain Iive lines at varying
angles and points oI intersection. Beginning with the dot-bedecked transparency, one goes about
making determinations Ior the perIormance oI each plurality (the three larger sizes oI dots), each
individual sound oI which has Iive intensional properties decided by a diIIerent one oI the Iive
lined transparencies, or by a diIIerent orientation (always aligned with the edges oI the dot
transparency) oI the same lined transparency
21
. Each line is to be assigned a diIIerent intensional
property, which it deIines by the relation oI that line to the position oI one oI the dots indicating a
plurality oI sounds. The Iive properties are: 'lowest Irequency, simplest overtone structure,
greatest amplitude, least duration, and earliest occurrence within a decided upon time (Cage
1960b: i). The shortest distance Irom the dot to that line determines its relation to the correlated
characteristic. The distances are 'to be measured or simply observed (Cage 1960b: i). Note
that these properties are all relational designations meaning that, Ior instance, the dot that is
closest to the line selected to represent the earliest occurrence will be the Iirst sound heard. It is
leIt to the perIormer to assign particular properties to particular lines.
21
Note a perplexing ambiguity in the instructions to the score: pluralities are to be played simultaneously or as
'constellations, and yet each individual constituent oI a plurality is subject to determinations made by diIIerent
lined transparencies or a diIIerent arrangement oI the same lined transparency, one oI the lines oI which is used to
determine the temporal placement oI that individual sound in reIerence to the total duration oI the piece.
63
Example 5 (John Cage, 'Variations I)
In spite oI their cryptic nature, these instructions have answered the various points oI the
analysis laid out in section 4. The designation oI meanings ascribed to the chart dimensions
(step 1) are addressed through the rules Ior use oI the lined transparencies. The simplicity oI all
the symbols employed by Cage, as single marks with no subordinate parts, obviates the need to
identiIy the s-code governing the Iormation oI the symbols (step 2). These symbols, straight
lines and dots, all take on some meaning designated by the perIormance instructions, making
them pleremic (and so step 3 is completed). The identiIication oI Iunctional class designations
Ior script elements (step 4) is also made (implicitly) within the perIormance instructions. The
dots are grouped into Iour categories each indicating a diIIerent number oI events. The actual
determination oI the Iive intensional properties Ior each sound is made by bringing the lined and
dotted transparencies together according to the syntax outlined in the instructions (answering
step 5). As Ior semantic reIerence (step 6), this too is provided when Cage deIines the number oI
events per dot and the properties to be assigned. This leaves us with little work to do in
reIerence to the analytical method, but it allows us to proceed to the assessment oI the score as a
piece oI (musical) writing.
64
Given the exceedingly large number oI possible sonic realizations oI this score
22
, we are
in the situation alluded to at the beginning oI this section regarding the evaluation oI extreme
notational experiments. II the score is the record oI the conditions deIining a proper instance oI
the associated musical perIormance and iI that record is conveyed by intentionally ambiguous
means, then we must conclude that the composer intends any justiIiable interpretation oI this
record as a proper interpretation, upon which a proper perIormance oI the work may be built.
This allows us to rescue a criterion Ior the judgment oI extreme cases oI graphic notation, as
works oI music; namely, these perIormances must be based upon a justiIiable interpretation. By
moving through the steps oI section 4, as we have, we can arrive at a basis Ior these justiIiable
interpretations. Though it would be impossible, in almost all cases, to reconstruct the
arrangements oI transparencies used in a given perIormance Irom that perIormance, this does not
preclude the possibility oI the perIormers making a determination as to whether or not the
perIormance was a proper instance oI the particular arrangements oI the transparencies used in
its construction
23
. This sort oI assessment, however, lies outside oI our present purpose and
requires a conIluence oI various methodological approaches (noted in the introduction). We do
have another approach to the evaluation oI the work available to us (one that does conIorm to the
interests oI the present inquiry); namely, the assessment oI the ingeniousness with which Cage
connects sound to notation Ior the reader/perIormer oI the score.
The economy oI means by which Cage creates this immense web oI possibility is quite
remarkable: dots oI Iour sizes and Iour sets oI Iive lines. He has managed to transcend the two-
dimensionality (and its traditional musical interpretation) oI the graphic space his notation
occupies: The system oI measuring a dot`s distance Irom the lines oI an overlaid transparency
22
For each oI the seven points that represent two sounds, there will be 20 x 19 possible arrangements oI the
transparencies to determine the intensional properties oI the elements oI the plurality in question. Similarly Ior each
oI the three dots representing pluralities oI three sounds there will be 20 x 19 x 18 possible arrangements oI the
transparencies and Ior each oI the Iour dots representing pluralities oI Iour sounds, there will be 20 x 19 x 18 x 17
possible arrangements oI the transparencies. For each transparency conIiguration, there will be 5! Possible
assignments oI intensional properties to the lines contained within the transparency. This totals a possible
3,046,528,558,080,000 variations oI assigned properties to be applied to the pluralities. A close reading oI the
perIormance instructions reveals that there is no direction Ior making determinations Ior the thirteen individual
sounds: 'In using pluralities, an equal number oI the 5 other squares (having 5 lines each) are to be used Ior
determinations, or equal number oI positions each square having 4 (Cage 1960b: i). It would appear that Cage is
leaving the determination oI all perIormance aspects oI the smallest dots to the player`s discretion, adding one more
layer oI indeterminacy to the work.
23
A key ontological point is thereby raised; namely, that the interaction with the score is a necessary condition Ior
any instance oI this work. This is in contrast to more traditionally notated works where a perIormer might learn his
or her part by ear without any impact on the identity oI the perIormed work. This also suggests that an audience is
not in a position to pass judgment on perIormances oI the work (a suggestion which is, incidentally, perIectly in line
with Cage`s aesthetics).
65
completely avoids any leIt to right (or right to leIt) or top to bottom (or bottom to top) readings.
Each line may be thought oI as a zero point elongated in a single dimension and Irom which all
the various dots oI the ground score may be conceived to radiate. And so any given arrangement
oI a transparency creates a two-dimensional text, the graphic space oI which is to be
reinterpreted as a network oI Iive independent one-dimensional charts. The intensional
properties oI a sound all clearly descriptive, in this case are expressed through chart
relationships, each line taking a meaning, as in a plereme, but also Iunctioning as the
instantiation oI one oI these one-dimensional charts. Dots indicate some number oI events or,
equivalently, some number oI graphic spaces, in which the dot is to mark a single position. That
number is indicated by distinctions in the script (the size oI the dots). Cage`s use oI the dot bears
a resemblance to the chart-based Iunction oI the notehead in conventional notation. Both the
traditional notehead and Cage`s dots are used to designate locations in a chart that determine
intensional characteristics oI the event(s) the dot marks the traditional notehead designating the
approximate pitch height oI a given event as well as its numeric position in a sequence oI events,
Cage`s dots determining timbre, temporal position, duration, volume, and pitch height in
reIerence to one or more sets oI Iive axes. The choice oI Iive lines Ior the remaining
transparencies suggests an analogy with the traditional musical staII. Having already detailed
how these lines redeIine the graphic space by operating as simultaneous radial axes, we need
only note that the traditional staII is the symbolic marker oI both the localized vertical/pitch
height axis and the general horizontal/temporal axis to see how the idea oI Iive lines, correlating
the chart-based aspects oI the notation to musical properties (two oI which are pitch height and
temporal placement), have been so inventively reconceived.
'Variations I requires a radically diIIerent approach to its interpretation than does a
work in traditional musical notation. It is no longer possible to read the instructions to the score,
open to page one and proceed. The act oI reading becomes the primary Iunction oI the
perIormer. It requires a substantial amount oI preparatory work (including the resolution oI
ambiguities in the instructions) beIore the generation oI any sound. Indeed, it requires a direct
physical manipulation oI the materials oI the notation. Notation ceases to be a means oI arriving
at a particular sound construction or even a constrained range oI possible sound constructions. It
becomes a task oI creative engagement with the score that, in the end, will sound. That
sounding, however, is not conditioned by the aesthetic preIerences oI the composer or (to a lesser
extent) the perIormer. It is, rather, the complicated outcome oI a reading strategy that must, in
this case, be non-linear, keeping track oI multiple axes oI inIormation at conIlicting angles. It
66
will most likely require the creation oI a secondary document synthesizing the determinations
made by the particular arrangement oI the indeterminate score.
5. Conclusion
It is my hope that the Iramework presented in this paper a set oI relevant terms, drawn Irom
recent studies oI writing systems and buttressed with material Irom semiotics and general
linguistics (sections 2 and 3), along with a systematic procedure Ior their use (section 4,
illustrated in section 4.1) will Iacilitate IruitIul discussion regarding the very real (and very
particular) complexities contained within 'graphic scores. Moving beyond the physical
contents oI the score and their interpretation, the discussion enlarges to embrace strategies Ior the
evaluation oI the work (section 4.2). Even where the sonic extension oI the score is so variable
as to admit almost any collection oI sounds, as we saw with 'Variations I, we can still have
grounds Ior judgments oI the score as a text setting Iorth intensional properties oI sounds to be
satisIied in a perIormance oI the work. This, in turn, Iorms the Ioundation oI judgments
regarding particular perIormances (even iI the judgments can only be made by the perIormers,
themselves). What constitutes the 'work as a whole, and whether this work is a work oI music
or something else, remains an open question on both ontological and deIinitional Ironts. A
complete analysis oI the work will depend on the answer to this question. Once answered, the
complete analysis will need to employ a combination oI ethnographic, phenomenological, and
textual approaches applied to those perIormances, interpretations, etc., construed as Iorming a
part oI the work.
References
Alden, Jane (2007). 'From Neume to Folio: Medieval InIluences on Earle Brown`s Graphic Notation.
Contemporarv Music Review 26(3/4), 315-332.
Busse Berger, Anna Maria (1993). Mensuration and Proportion Signs. Origins and Evolution. OxIord: Clarendon
Press.
Cage, John (1960b |1958|). 'Variations I. For Any Number oI Players. New York: Henmar Press.
Catach, Nina (1986). 'The Grapheme: Its Position and Its Degree oI Autonomy with Respect to the System oI
Language. In: Augst, Gerhard (ed.), New Trends in Graphemics and Orthographv. Berlin: Walter de
Gruyter. (1-10).
Coulmas, Florian (2003). Writing Svstems. an Introduction to Their Analvsis. Cambridge: Cambridge UP.
Eco, Umberto (1979). The Role of the Reader. Explorations in the Semiotics of Texts. Bloomington: Indiana UP.
67
Eco, Umberto (1984). Semiotics and the Philosophv of Language. Bloomington: Indiana UP.
Eco, Umberto (1976). A Theorv of Semiotics. Bloomington: Indiana UP.
Fischer, Steven Roger (20101). A Historv of Writing. London: Reaktion Books.
Goodman, Nelson (1968). Languages of Art. An Approach to a Theorv of Svmbols. New York: The Bobbs-Merrill
Co.
Harris, Roy (1995). Signs of Writing. New York: Routledge.
Hjelmslev, Louis (1963 |1943|). Prolegomena to a Theorv of Language. WhitIield, Francis J. (trans.). Madison:
University oI Wisconsin Press.
Kanno, Mieko (2007). 'Prescriptive Notation: Limits and Challenges. Contemporarv Music Review 26(2), 231-
254.
Levinson, Jerrold (1980). 'What a Musical Work Is. Journal of Philosophv 77(1), 5-28.
Levy, Kenneth (1987). 'On the Origin oI Neumes. Earlv Music Historv 7, 59-90.
Martinet, Andre (1964 |1960|). Elements of General Linguistics. Palmer, Elisabeth (trans.) London: Faber and
Faber.
Meeus, Nicolas (2002). 'Musical Articulation. Musical Analvsis 21(2), 161-174.
Parish, Carl (1957). The Notation of Medieval Music. New York: W. W. Norton and Co.
Patel, Aniruddh C. (2008). Music, Language, and the Brain. OxIord: OxIord UP.
Read, Gardiner (1979). Music Notation. A Manual of Modern Practice, 2
nd
ed. New York: Taplinger Publishing Co.
Read, Gardiner (1987). Source Book of Proposed Music Notation Reforms. New York: Greenwood Press.
Santosuosso, Alma Colk (1989). Letter Notation in the Middle Ages. Ottowa: The Institute oI Medieval Music.
Schreur, Phillip E (1989). Tractatus Figurarum (Treatise on Noteshapes). A New Critical Text and Translation.
Lincoln: University oI Nebraska Press.
Treitler, Leo (1982). 'The Early History oI Music Writing in the West. Journal of the American Musicological
Societv 35(2), 237-279.
Treitler, Leo (1992). 'The Unwritten` and 'Written Transmission` oI Medieval Chant and the Start-up oI Musical
Notation. Journal of Musicologv 10(2), 131-191.
WolterstorII, Nicholas (1975). 'Towards an Ontology oI Art Works. Nous 9(2), 115-142.
68

2010-Meaningful Scribbles jmm9 - Libre

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

2010-Meaningful Scribbles jmm9 - Libre

Transféré par

Droits d'auteur :

Formats disponibles

Peer-Reviewed Paper JMM. The Journal of Music and Meaning, vol.

Vous aimerez peut-être aussi