From Allen Forte: Schenker’s Conception of Musical Structure, in Journal of Music Theory, III/1

(April, 1959), 7 – 14, 23 – 24. Reprinted in Norton Critical Scores: Dichterliebe, edited by
Arthur Komar.1
Note: I have modified Forte’s footnotes in two ways: first, I’ve added a number of footnotes of my own,
pointing out where some of the unexamined assumptions of Schenkerian analysis might be fueling some of
Forte’s or Schenker’s statements, and second, where the reference is sufficiently obscure as to be best
omitted. Those footnotes which present my own commentary are always prefaced with my initials SLF.
(Sometimes I’ve appended my commentary directly to Forte’s footnotes.)
From time to time I may insert a larger-scale note into the text proper; such notes are formatted as you see
here, in smaller italic text and enclosed in indented paragraphs.

I can think of no more satisfactory way to introduce Schenker’s ideas, along with the terminology
and visual means which express them, than to comment at some length upon one of his analytic
sketches. For this purpose I have selected from Der freie Satz a sketch of a complete short work,
the second song from Schumann’s Dichterliebe. I shall undertake to read and interpret this
sketch, using, of course, English equivalents for Schenker’s terms.2
Here in visual form is Schenker’s conception of musical structure: the total work is regarded as
an interacting composite of three main levels. Each of these structural levels is represented on a
separate staff in order that its unique content may be clearly shown. And to show how the three
levels interact, Schenker has aligned corresponding elements vertically. I shall make a quick
survey of this analytic sketch and then give a more detailed explanation.
The lowest staff contains the major surface elements, those elements which are usually most
immediately perceptible. Accordingly, Schenker has designated this level as the foreground. In
deriving his foreground sketch from the fully notated song, Schenker has not included all its
actual note values. Those which he does include represent in some cases the actual durational
values of the work; but more often they represent the relative structural weight which he has
assigned to the particular tone or configuration. The sketch omits repeated tones, and shows
inner voices in mm. 8 – 12 only, indicating that there they have a greater influence on the voice-
leading.
On the middle staff Schenker has represented the structural events which lie immediately beyond
the foreground level. These events, which do not necessarily occur in immediate succession in
relation to the foreground, comprise the middleground. It should be evident now that the
analytic procedure is one of reduction; details which are subordinate to larger patterns are
gradually eliminated – in accordance with criteria which I will explain further on.
Finally, on the upper staff, he has represented the fundamental structure level, or background,
which controls the entire work.
Now let us consider the content of each level in some deatil. This will provide an opportunity to
examine other important aspects of Schenker’s thought, all derived from his central concept.

1 SLF: I’ve added a copy of the score to the attached analysis, for convenience.

2 The rendering of Schenker’s technical expressions into English presents a number of problems, not the least of

which is the fact that there are already, in some cases, two or more published versions of the same term. It is to be
hoped that with the publication of Der freie Satz (now being translated [SLF: it is nowadays easily available as Free
Composition]) a standard nomenclature will be established.
A series of sketches such as this can be read in several directions. For the purpose of the present
introductory explanation it would seem advantageous to begin with the level which contains the
fewest elements and proceed from there to the level which contains the most – thus, reading from
top to bottom or from background to foreground. By reading the sketches in this order we also
gain a clear idea of Schenker’s concept of prolongation: each subsequent level expands, or
prolongs, the content of the previous level.
The background of this short song, and of all tonal works, whatever their length, is regarded as a
temporal projection of the tonic triad3. The upper voice projects the triad in the form of a
descending linear succession which, in the present case, spans the lower triadic third. Schenker
marks this succession, which he called the Urlinie, or fundamental line, in two ways: (1) with
numerals (and carets) which designate the corresponding diatonic scale degrees, and (2) with the
balken [i.e., beam] which connects the stemmed open notes (I shall explain the black noteheads
shortly). The triad is also projected by the bass, which here outlines the triadic fifth, the tonality-
defining interval. Schenker calls this fundamental bass motion Bassbrechung, or bass
arpeggiation. Like the fundamental line, it is represented in open note-heads. The fundamental
line and the bass arpeggiation coordinate, forming a contrapuntal structure, the Ursatz, or
fundamental structure which constitutes a complete projection of the tonic triad.4 Thus, to
Schenker, motion within tonal space is measured by the triad, not by the diatonic scale.5
Observe that in this case the most direct form of the fundamental structure would be the three-
interval succession in the outer voices:
fundamental line: 3 2 1
bass arpeggiation: I V I

3 SLF: This is a clear statement of the single most significant shortcoming of Schenkerian analysis: note the phrase

of all tonal works. By stating that this applies to all tonal music, Schenker is in fact proposing what in the sciences
would be called a ‘theory’ that aims to account for observed phenomena in a comprehensive manner. But the tonic-
triad notion is scientifically irresponsible as a theory: Schenker offers no mechanism by which his theory can be
disproven, which is the goal of all scientific inquiry. He only offers a set of works which either demonstrate his
assumptions, or which can at least be massaged into supporting his thesis. It is presented as a fait accompli and we
move on from there. As such, it really belongs more in the realm of theology or philosophy – systems of humanistic
thought and not repositories of verifiable truth. As is well known by scientists obliged to debunk religious
fundamentalists with dreary regularity, if you begin an inquiry looking for something specific, you will be certain to
find it, no matter how arcane or ridiculous it might be. Some of the most amusing examples of such misplaced
‘analysis’ can be found in the voluminous commentary of those folks who are determined to prove that the
authorship of William Shakespeare’s plays is Francis Bacon or the Earl of Oxford of Queen Elizabeth or John Florio.
Some of the ‘proofs’ are startling in their elaboration and complexity, but none of them are remotely valid in terms of
logical integrity.
4 Each tonal work manifests one of three possible forms of the fundamental line, always a descending diatonic

progression: 3 – 1 (as in the present case), 5 – 1 and 8 – 1. Variants upon these forms arise when the bass arpeggiation
disposes the fundamental line components in different ways.
5 SLF: This represents yet another oddly askew aspect of Schenker’s ideas – they are based predominantly on the

harmonic practices of Western Europe from about 1700 through 1900. Diatonic scales and triads are far from
being the only materials available which measure tonality; one can think of whole-tone scales, octatonic scales,
pentatonic scales, or the ragas of the music of India. Schenker was drawing ‘universal’ principles from his survey of
an extremely limited set of samples.
 Note: from this point onwards, the union of fundamental line and bass arpeggiation is written with a slash
between the two components, therefore 2/V means a second scale degree over V in the bass, or 3-2/I-V
would refer to third scale degree over I in the bass followed by second scale degree over V in the bass, and
so forth.

The background sketch shows that this succession occurs consecutively only in the last part of
the song. The song begins unambiguously with 3/I;6 however, it does not progress immediately
to 2/V and from there on to 1/I; instead, the first interval is prolonged as shown in the sketch: the
upper voice C# first receives an embellishment, or diminution, in the form of the third-spanning
motion, C#-B-A (represented in black noteheads), and then moves over a larger span (shown by
the beam) to B on the last eighth-note of m. 8, where it is supported by the bass V. (This V is not
to be equated with the final V [m. 15], which effects a closure of the fundamental line.) Schenker
then shows how this initial prolongation is followed by a restatement of 3/I and the completion of
the succession 3/I - 2/V - 1/I.
To recapitulate, there are two prolongational classes shown in this background sketch. The first
includes diminutions, or prolongational tones of shorter span (represented by black noteheads);
the second includes the larger prolongational motion from 3 to 2 (connected by the beam) which
comprises the controlling melodic pattern of the first phrase. Schenker regards this larger
prolongation motion as an interruption of the direct succession 3-2-1/I-V-I, and represents it by
placing parallel vertical lines above the staff following 3-2/I-V. The fundamental structure, which
is in this case the uninterrupted succession 3-2-1/I-V-I therefore may be considered as the
essential content of the background.7 In reading Schenker’s analytic sketches a distinction must
often be drawn between the background level in toto, which sometimes includes prolongations of
primary order as in the present case, and the essential content of that level, the fundamental
structure. Thus, “fundamental structure” designates a specific contrapuntal organization which
assumes several possible forms, whereas “background” is a term which may include other events
in addition to the fundamental structure, as in the present instance, where it includes two
prolongations, each belonging to a different structural order. This distinction, not always clearly
drawn by Schenker, is indispensable to the full understanding of his sketches and commentaries.
In this connection I point out that within each of the three main structural levels several sub-
levels are possible, depending upon the unique characteristics of the particular composition.8
The idea of the interrupted fundamental line provides the basis for Schenker’s concept of form.
For example, in the typical sonata-allegro form in the major mode, interruption of the

6 This is rather funny, considering that there’s absolutely nothing ‘unambiguous’ about the opening. I make further

comments along the way about this opening. Also see my essay Problems With Schenkerian Analysis.
7 It should be apparent that Schenker’s major concept is not that of the Ursatz, as is sometimes maintained, but that

of the structural levels, a far more inclusive idea. SLF: objections such as mine regarding the logical flimsiness of the
Ursatz notion were already being heard loudly when this article was originally published in 1959. Furthermore,
analysts may claim that the Ursatz is of relatively lesser importance, but certain of their analyses tend to prove
otherwise as they go to distinctly extraordinary lengths in order to preserve that Ursatz idea.
8 Undoubtedly Schenker compressed many of his sketches in consideration of the practical requirements of

publication. Mr. Ernst Oster, who has in his possession a large number of Schenker’s unpublished materials – which
he plans to present along with commentaries at a future date – has brought this to my attention. Schenker’s
unpublished sketches of Brahms’ Waltzes, Opus 38, for example, are executed on several superimposed staves, so
that each structural level is shown distinctly and in detail.
fundamental linear progression at the close of the exposition normally gives rise in the
development section to a prolongation which centers on V. Of course, the prolonged
fundamental line component varies, depending upon which form of the fundamental structure is
in operation and upon which specific prolongation motions occur at the background level.
Before explaining the middleground, I should like to direct attention again to the diminution
which spans the third below C# (black noteheads). By means of the numerals 3, 2, 1, enclosed in
parentheses, Schenker indicates that the motion duplicates the large descending third of the
fundamental line. This is an instance of a special kind of repetition which Schenker called
Übertragung der Ursatzformen (transference of the forms of the fundamental structure).
Throughout his writings he demonstrates again and again that tonal compositions abound in
hidden repetitions of this kind, which he distinguishes from more obvious motivic repetitions at
the foreground level.9
We can interpret the content of the middleground most efficiently by relating it to the
background just examined. The first new structural event shown at the middleground level is the
expansion of the smaller prolongational third (black noteheads) by means of the upper adjacent
tone10, D, which serves as a prefix. The sketch shows how this prolongational element is
counterpointed by the bass in such a way as to modify the original (i.e., background) third. That
is, the figured-bass numerals in parentheses indicate that the second C# (black notehead) is a
dissonant passing-tone, and therefore is not to be equated with the initial C#, which serves as the
point of departure for the fundamental tone. The adjacent tone D recurs in m. 14, where
Schenker assigns more structural weight to it, as indicated by the stem.11 I reiterate that
conventional durational values are used in the analytic sketches to indicate the relative position
of a given component or configuration in the tonal hierarchy – the greater the durational value,
the closer the element to the background.
In addition to the prolongation described in the preceding paragraph, the middleground
contains the essentials of the prolongational middle section (mm. 9 – 12) which appears in more
detail in the foreground sketch. Schenker regards this entire middle section as a prolongation of
the background fifth formed by 2/V. Its main feature is the inner voice which descends from G#
to E, a middleground duplication of the fundamental line’s third. The bass which counterpoints
this inner voice arpeggiates the tonic triad, E-C#-A. Schenker shows how the arpeggiation is
partially filled in by the passing note, D, and by slurring E to A he indicates that he considers that
motion to be the controlling bass motion, within which the C# functions as a connective of
primarily melodic significance.12 Here we have an example of the careful distinction which

9 SLF: motives may sometimes reduplicate the fundamental structure, as well.

10 Schenker’s abbreviation, “Nbn.,” stands for Nebennote, or in English, adjacent tone (not “neighbor tone”).

11 SLF: I don’t understand Forte’s remark here; in both previous instances of this same D (measure 2 and its ‘clone’

measure 6) the note is also stemmed, as it is in measure 14, so as far as I can tell there’s no difference. In measure 14
Schenker does not add the notation (Nbn) which stands for ‘adjacent tone’. Perhaps the F-natural in the piano part
has something to do with that, but since the F-natural was reduced out in the foreground, there’s no way of following
his reasoning.
12 The author adds here a footnote calling attention to Schenker’s remarks: “The bass carries an arpeggiation of the

fifth down through the third without, however, invalidating the interruption.”
Schenker always draws between major bass components or Stufen, which belong to the
background level, and more transient, contrapuntal-melodic events at the foreground and
middleground levels.
A brief consideration of three additional events will complete our examination of the
middleground level. First, observe that the diatonic inner-voice descent in the middle section,
G#-E, is filled in by a chromatic passing-tone, G. Schenker has enclosed this in parentheses to
indicate that it belongs to a subsidiary level within the middleground. Second, observe that just
before the inner-voice motion is completed on the downbeat of m. 12, the G#, its point of
departure, is restated by an additional voice which is introduced above it. Schenker has pointed
out that in “free” compositions, particularly instrumental works, the possibility of more
elaborate prolongation is greatly increased by introducing additional voices, as well as by
abandoning voices already stated. The final event to observe here occurs in the middle section:
the motion from B, the retained upper voice, to C# on the downbeat of m. 12. This direct
connection does not actually occur at the foreground level, but Schenker, feeling that it is
strongly implied by the voice-leading context, encloses the implied C# in parentheses and ties it
to the actual C#, thereby indicating that it is an anticipation.13
In the foreground sketch Schenker represents for the first time the metrical organization of the
song. As I have already mentioned, he shows here some of the actual durational values, in
addition to using these as sketch symbols14. This reveals the position assigned to meter and
rhythm in his system: he considered them to be important structural determinants at the
middleground and foreground levels15 but subsidiary to the fundamental tonal organization,
which, he maintained, was arhythmic.
Let us now examine some of the relationships which Schenker has shown in his sketch of the
foreground, this time beginning with the bass. In m. 2 he encloses the bass-note A in
parentheses16 and marks it with the abbreviation, Kons. Dg. (Konsonanter Durchgang or
“consonant passing-tone”). By this he indicates that the tenth which the bass A forms with the
upper-voice C# transforms the latter, a dissonant passing-tone at the middleground level, into a
consonance at the foreground level. In this way he also intends to indicate the function of the
chord at that point. Since it supports a passing-tone in the upper voice it is a passing chord. In
addition, it belongs only to the foreground and therefore is to be distinguished from the initial

13 SLF: This practice of adding ‘implied’ tones is one of the most controversial of Schenker’s practices and definitely

gives abundant fuel to arguments that Schenkerian analysis is far too easily ‘massaged’ to fit the underlying tonal
theory.
14 SLF: this practice tends to confuse the reader, so it shouldn’t be surprising to learn that most later Schenkerian

analysts don’t do it. In Schenker’s foreground analysis the flags in measures 3 and 4 are probably rhythmic in nature,
but they also seem to imply that the C# and A relate to a more significant level, which is also true. Most readers are
uncomfortable with this lack of clarity.
15 SLF: although he said very little about “how” they were important and, in fact, tends to be rather arbitrary about

the entire issue.

16 SLF: Forte identifies this as measure 3, but measures are counted starting with the first complete measure, and not

with upbeat measures, so I’ve corrected him.

tonic chord, a background element.17 Two of Schenker’s most important convictions underlie
this treatment of detail: (1) that the study of strict counterpoint provides the indispensable basis
for a thorough understanding of the details, as well as the larger patterns of a composed work,
and (2) that the function of a chord depends upon its context, not upon its label. This can be seen
in his notation of the chords in this sketch. Although he uses the conventional Roman numerals
he provides them with slurs, dashes and parentheses to show their relative values in the tonal
hierarchy. Thus, the long slur from I to I indicates that the IV and V chords lie within the control
of that chord, while the abbreviation Vdg. (Vordergrund) shows that the succession belongs to
the foreground. And in the middle section, mm. 8 – 12, the parentheses show that the chords
between V and I are subsidiary chords. These arise as part of the prolongational complex at that
point and stand in contrast to the stable background chords I and V. These arise as part of the
prolongational complex at that point and stand in contrast to the stable background chords I and
V.
Now let us turn to the melody. We can most efficiently examine its structure by first comparing
each foreground prolongation (slurred) with the larger middleground prolongation immediately
above it, and then by relating both the foreground and middleground to the background. In this
way we see that the foreground prolongation of the first section spans a descending third twice,
thus duplicating the successively larger thirds at the middleground and background levels. In the
middle section the melody undergoes more elaborate development. There, by means of
connecting beams, Schenker shows how the upper voice skips down to the inner voice and back
again. The ascending skips comprise a sequence of two fourths, which are marked by brackets
and emphasized by a typically Schenkerian exclamation point. This sequence lends support to his
reading of the implied anticipation of C# in the upper voice of m. 12, mentioned earlier.18
The foreground of the middle section provides a good example of Schenker’s concept of
“melody” (he avoided the term in his writings) as a self-contained polyphonic structure. This
valuable aspect of his theory,19 which is absolutely indispensable to any kind of intelligent
melodic analysis20 , is well substantiated by compositional practice21 . There are many passages in
the literature where polyphonic melodies, implied at one point (often the beginning) are
subsequently realized in full, for example in the first movement of Mozart’s Sonata in A minor, or

17 SLF: Obviously Schenker didn’t consider the text very important. The chord on the second beat of the measure

supports the resolution of the words spriessen (measure 2), werden (measure 6), and klingen (measure 14), all vitally
important words in the understanding of the text. According to his analysis, these considerations are foreground
only and not germane to the ultimate structure of the song.
18 SLF: at the risk of sounding like a tiresome skeptic, I’m not sure how it does ‘lend support’ of the implied

anticipation.
19 A highly interesting application of this concept is to be found in Schenker’s essay, ‘Das Organische der

Fugue” (Das Meisterwerk in der Musik, Munich, 1925-30, Vol. II), where he employs his technique of synthesis, or
reconstruction, to demonstrate that the subject of Bach’s C-minor Fugue (WTC I) implies a compleete, self-
contained contrapuntal structure.
20 SLF: Many Schenkerians are distinguished by a rather nasty, sanctimonious arrogance. That tone alone has

probably done more harm to their ‘cause’ than even the most arcane convolutions of their publications.
21 SLF: But where is it disproven or directly contradicted? That’s how a theory is tested, not by ferreting out

examples that agree with the theory.

in Brahms’s Intermezzo in Bb major, Op. 76/4; and, of course, we find a special development of
this concept in Bach’s compositions for solo violin and for solo cello. Here, in the foreground
sketch of the middle section the diagonal beams show that the vocal melody shifts back and forth
between two lines, the lower of which belongs to the accompaniment22 . It is evidence that this
section contains the most intricate upper-voice prolongation.
It also contains the most elaborate bass motion. The sketch shows how the bass provides
counterpoint to the upper-voice (foreground) prolongation of B, bass and upper voice
comprising the interval succession 5-10-5-10-5, which is enclosed within the middleground
outr-voice succesion, B-C#/E-C#. Observe that the upper voice alternates between an upper
adjacent-tone prolongation of B (marked Nbn.) and the skips into the innter voice which were
explained in the preceding paragraph. The lowest voice in this passage is subordinate to the
voice which lies immediately above it, E-D-C#, the latter succession being the actual bass line
(cf. middleground sketch). Nor does its registral position above the foreground bass lessen its
importance as the main motion-determinant in the lower voices. Therefore, the foreground bass
which displaces or covers it registrally might be termed a “pseudo-bass.”23
One final aspect of the foreground-sketch deserves mention: the form. Schenker indicates this
with the customary letters and exponents. The foreground form therefore corresponds to the
form-generating interruption at the middleground and background levels as follows:
Statement Interruption Statement and closure24
A1 B A2
It should be apparent that an analysis of this kind embraces all the information generally included
under the heading “form and analysis” but that it goes far beyond to interpret the relationships to
the background which are revealed during its initial phases, where the main concern is to achieve
an accurate reading of foreground and middleground.
A summary of this analysis should properly include a classification of the chromatic chords in the
middle section of the piece, and a more precise explanation of the coordination of linear intervals
at the foreground level, the descending thirds and fifths (which latter take the form of diminished

22 SLF: Forte doesn’t identify it as such, but the technical term for this is unfolding.

23 SLF: This statement sounds like another of these Schenkerian massaging of the facts to support the analysis, but

personally I’m rather comfortable with it. (Whether that makes it ‘true’ or not is another matter.)
24 SLF: The analysis places A2 at the upbeat to measure 13. While this clearly correct as to phrase and textual

analysis, Schenker insists on treating the harmony at that point as a tonic, despite its being a clear V7/IV and not a
stable tonality by any means. Schenker needs that chord to be stable, however, because he has assigned to its
soprano C# a fundamental tone of the Urlinie, 3. Allen Forte, in a later commentary, attempts to address this
problem, by placing that all-important fundamental tone C# (3) on the second beat of measure 14. This, however,
implies that this Urlinie tone is on a weak beat (hardly audible as such), and is also the same C# which Schenker
analyzes as a dissonant passing tone in measure 2. It doesn’t quite make sense. To correct that, Arthur Komar also
changes the opening Urlinie tone 3 to the weak beat of measure 2, corresponding to Forte’s change from measure
13 to 14. Neither analysis bears much relation to the perceived experience of the song. In fact, the ambiguity which
marks the opening of the song (given that it is the continuation of a C# dominant seventh chord at the conclusion of
the previous song) is nowhere addressed in Schenker’s analysis, but it seems clear enough that the harmonic
instability of measure 13 reflects the harmonic ambiguity that opened the song. The entire issue of the song’s
location in context is completely overlooked by Schenker, and consequently by both Forte and Komar.
fifths and ascending fourths in the middle section). However, because of space limitations, I shall
not undertake a summary here, but instead to go on to discuss other aspects of Schenker’s work.
If the preceding commentary has succeeded in demonstrating some of Schenker’s more
important ideas, as well as clarifying some of the vocabulary and visual devices which he employs
to express those ideas, it has fulfilled its purpose.