Go to Lecture: One or Two or Three or Four or Five or Six or Eight
Go to Course Outline
Go to Cyber Semiotic Institute Home Page

Musical Gesture

Instructor: Robert S. Hatten

Lecture 7: Gesture and Agency

In the second lecture I offered a series of presuppositions about gesture. The first of these is fundamental to the topic gesture and agency:
  1. Gesture is movement interpretable as a sign, whether intentional or not, and as such it communicates information about the gesturer (or character, or persona the gesturer is impersonating or embodying).
    (Lecture 2)
If one commits to identifying musical events as gestures, then the implication of agency is hard to avoid. The discussion of tropes in the last lecture clearly suggested some kind of persona who, as in the case of the Finale of Beethoven's Op. 101, not only experiences an inner, spiritual victory, but in a performative sense (Searle, 1970), achieves it by affirmative utterance. The enactment of meaning by gesture, however, also carries with it the symbolic significance of more systematic stylistic correlations. It is not enough that we hear the opening of the Finale played "with determination"; we need to hear the imitation that triggers the learned style, and then after the shift to a "softer" contrasting gesture in m. 5, we must recognize the pedal point and continuous swirl of sixteenths that cue the pastoral through the musette topic. In other words, we still need to be able to identify topics, as well as the details of harmony and voice leading, rhythm and meter, etc., that support their cueing in the context of a musical phrase. And we have to recognize the stylistic expressive correlations which these topics bring to the creative synthesis I have called a musical trope. Not all of these elements of musical meaning are necessarily communicated through the Firstness or Secondness of gesture as performative realization, but may depend on our knowledge of style conventions at the level of Thirdness.

Analogously, one may process musical events without taking full account of their gestural agency, and indeed, the history of music theory demonstrates this possibility, with its traditions of formalist analysis. Recall Hanslick's image of the arabesque. One is hard pressed to account for an "arabesquer" in that equation, to say nothing of whether or not an arabesque is expressive of a particular emotional state! Philosophically, we know we cannot assume that the gesturer or agent of a composition is simply the composer, or the performer--but then again, there is the example of Liszt, who appears to have embodied both roles in conceiving his music within the aesthetic of Romantic self-expression. Instead of assuming such a fusion of agency, it may be helpful to propose varieties of agency that may be variously weighted in the interpretation of score and its manifestations in performance. In what follows, I will propose such a working set of categories.

Before offering that outine, however, I would like to follow up on my last lecture's speculations about gestural troping and consider the possibility that juxtaposing contradictory gestures may not always lead to their synthesis as a trope akin to metaphor. A striking example may serve to illustrate one reason why. Alkan's Le Festin d'Ÿsope or "Aesop's Banquet," Op. 39, is the last of twelve etudes in all the minor keys. Written in 1857, it is a set of variations on an eight-bar theme in E minor. Variations 21 and 22 constitute a "double" variation in E major. The hunting fanfare of Variation 21 is continued in the right hand in Variation 22 with the gestural addition of a "riff" in the left hand (5-6-7-8, as in the opening of Mozart's Jupiter Symphony). But the "riff" is marked Abbajante ("barking"), and thus we might already consider that gesture as tropologically developed by a linguistic metaphor ("the riff is a bark"). By twisting the representative connotation of the gesture from a snare-drum or military drum figure to an unflattering baying by a hunting dog, the performance instruction tropes on the correlation of a style type. As the variation proceeds, the barking accelerates, hilariously disrupting the hunting horn fanfare, and ultimately displacing it for a measure before the fanfare's final punctuated cadence.

Whereas I readily interpret a tropological effect of comic irony in this little drama (along the lines of deflation of pomp, such as might be suggested by a uniformed hunt), I am reluctant to consider it an example of gestural troping, since the two gestures remain separate entities in the drama. Unlike the performance gesturing of my Winterreise collaborator (Lecture 6), these gestures do not merge into a single agency but instead maintain their separate roles. The gestures might be said to contribute to a trope at the level of the discourse: their opposite meanings interact dialogically, and rather than fusing metaphorically into a third meaning, they create a trope of ironic wit from their unassimilated friction.

In terms of agency, these conflicting gestural types suggest the roles of protagonist and antagonist in conflict dramas, or more neutrally, actant and negactant, to use the terms popularized by A. J. Greimas and introduced by Eero Tarasti (1994) in his narratological music analyses. I will consider these as well as other types of agency that might be cued by gestures, whether those implied by a score or those generated by a performer while stylistically realizing, or interpreting, that score. Figure 1 outlines four of these types of agency:

I. Interpreted score conceived as drama
A. Level of story (musical events in plausible, logical, stylistic sequence)
1. principal agent (actant, protagonist, persona, subject, voice) = the individual subjectivity with which we identify, whether as performer or listener
2. external agent (negactant, antagonist; or depersonalized external force--Fate, Providence) = that agency which acts upon, or against, the principal agent
B. Level of [built-in] narration (if we can infer a compositional play with musical events or their temporal sequence or relationship, inflecting their significance, or proposing a certain attitude toward them; suggestive of "point of view" or filtered perspective)
3. narrative agent (the creator's persona, or the "teller's" persona)--the sometimes invisible or transparent agency involved in ordering, arranging, and/or commenting upon the (sequence of) events of the story level. (cf. my discussion of levels of discourse in Hatten, 1994: 174ff.)

II. Intepretation as realized in performance
A. Manifestations of types 1-3, presumably without overt intrusion of the performer's personality (beyond the performer's stylistic analysis and sonic reconstruction of the work and its implied agencies) but inevitably further inflected by:
B. Narrativity of presentation
4. performer-as-narrator--directing the listener's attention (possibly over-didactically) to the structure and significance of events, although not changing or reordering the events themselves (exceptions: productions of operas, some 20th-century scores offering the performer options). Commenting upon the events from the perspective of the individual point of view and prejudices of the performer as engaged participant in the "telling" of the story. May (over-) emphasize characterization of actants (1, 2), or especially unusual (narrative) reorderings or disruptions of expected events or event-sequence (3).

Figure 1: Four principal types of agency as (potentially) implied by the work or its performance.

The gestures implied by the score as stylistically interpreted are not simply those explicitly marked in the score, especially with regard to the variable elements of tempo, timing, dynamics, pedaling, and qualities of articulation. A notation may tacitly presuppose certain gestural realizations, such as the characteristic temporal warping of triple meter in a Viennese waltz, or the "swing" of jazz. But beyond the characteristic gestures we might consider appropriately implied by the work in the context of an historical style, a performance may entail additional gestures introduced primarily to clarify, or direct attention to, or emphasize, certain features of the work. In this way, the performer may inject considerable personality into the work, at all levels of agency. It is one way we distinguish what we call performers' personal styles, or temperaments. But we have also had experiences encountering performances which so matched our ideal of a work that we considered them to be pure realizations of the composer's intentions--and our biases may ironically have led us to praise those performers for selfless (!) dedication to the work.

The performer may, even in such seemingly "pure" realizations, be narrativizing by gestural highlighting, which might range from simple deixis (emphasizing a melodic apex or harmonic deflection) to the musical equivalent of abstract pointing (McNeill, 1992: 173) in the gestures that accompany speech. According to psychologist David McNeill of the University of Chicago, abstract pointing is a gesture that highlights the frame, or disjunctive locations and times within the frame, of a story or discourse. An analogous example from performance might involve a pianist who highlights the retrospective character of a tonally-relaxed coda by leaning back, as if suddenly removed from the temporal present and absorbed in the reverie of a memory. This visual gesture may be accompanied by subtle changes of tempo and dynamics that tend to place the section under a veil of reminiscence. Such postures may be seen in photographs and drawings of Liszt and Brahms performing at the piano, and they may suggest a bit of over-romanticizing to modern audiences, but their validity derives from the concept of self-expression, or self-absorption, germane to Romantic aesthetics, as mentioned earlier.

Furthermore, the temporal and dynamic nuances that may result from such performer- narratizing choices may well be implied by the score. For example, in the turn to the coda of the finale of Beethoven's Piano Sonata in A, Op. 101 (mm. 281-91), I perform the return modulation (mm. 290-91) with a rubato that allows it sufficient time to register after the jolting disruption of the fortissimo octave motto. The return of the octave motto (mm. 281-82) parallels the development section's opening and threatens a renewed engagement in the key of F majorIn mm. 290-91, however, a kind of rhetorical questioning leads the listener from F as potential tonic to reintegration as lowered-6 to 5 in a Phrygian half cadence in A. In my performance, I then reset the tempo with the pickup to m. 292, where the thematic return of the coda is initiatory and A major is no longer in doubt. Although no such rubato or taking of time is sanctioned by the notated score in the transition following the fortissimo octaves, any attempt to play through those measures in strict tempo would be, in my judgment, a failure to recreate the implied narrativity of the work as dramatically conceived, and as cued by such factors as harmony, texture, theme, topical contrast, and formal juncture (even the rests need rhetorical emphasis!)

In Bach, we often encounter stylistically encoded rhetorical gestures, such as the Neapolitan bII6 and V4/2 harmonies, arpeggiated in near-cadenza fashion, just before the evaded cadence to iv7-6 in the Prelude in Eb Minor from Book I of the Well-Tempered Clavier. Even if played on the harpischord, subtle emphases of timing and articulation may gesturally enhance the rhetorical, or oratorical, presentation. Czerny, the pianist whose edition is notorious for its added dynamic and other markings, is clearly responding to the significance of these chords with his crescendo and forzato markings. His indications, though crude, exemplify a narrativity of selective emphasis which I suspect is cued by the rhetorical significance of the harmonies. But to what extent is it possible to divide the gestural contribution of the performer between (1) the built-in rhetorical gesture implied by the style, and (2) a performer such as Czerny's own (over-) narratizing of these clearly oratorical flourishes?

Piano music also allows for multiple voices, hence multiple agencies, of which the simplest is the distinction between foregrounded actant and backgrounded environment, as found most obviously in melody and accompaniment textures. But it is also possible to interpret a single agency that is split between two gestural fields, as it were. An elegant and simple example is found in the opening of Debussy's Des pas sur la neige, or "Footsteps in the Snow," the sixth of the PrZludes, Book I. The repeated pattern of the opening represents each footfall crunching into a crusted field of snow, and may be further interpreted (based on the minor mode, limited ambitus, and obsessive repetition) as expressive of grief or sadness. Debussy's three verbal directions leave no doubt as to his intention: "triste" appears twice and "douloureux" once. What makes this prelude peculiarly modern is the fragmentary melody that emerges above the ostinato. By choosing not to begin segments of the melody on structural pitches in the mode of D minor (D and A), or the prevailing harmony (G in m. 5), Debussy creates the effect of a melodic line dissociated tonally from its accompaniment. It is a short interpretive step from dissociated melody to dissociated consciousness, or frozen feeling, if you will. A twentieth-century kind of grief is one compounded of repressed feeling and negated hope, of which this combination of gestures provides an appropriate trope. The contour of the lines in mm. 2-4 and 5-7 suggests a vague yearning in its echo of a Romantic ascent, especially fragile with the one-note sighs in mm. 5-6. The two collapses of the melodic line in mm. 3-4 and 7 suggest that the apex of E was unsatisfactory as a goal or climax, and that the potential yearning toward a warmer emotional expressivity cannot sustain its hopeful energies against the icy weight of frozen grief.

Theoretically, I would assume that the two strands of this example, while implying different gestural realizations, could be understood as two parts of a single agency; the purposeful "split" between monotone repetitive body motions in the ostinato and dissociated emotional thought processes in the upper line contribute two parts of an underlying psychic condition occasioned by a single, thus principal, agent.

I have lightly sketched in the Debussy example a typical chain of interpretants that Western music has previously forged between representation and expression, and I will note only parenthetically here that the performer's gestures in representing any physical part of the natural environment (e.g., snow crunching underfoot) must also, again according to well-established linkages in Western music, reveal the expressive tone of the agent from whose implied perspective we are given the perception. At times, the associative linkage may be forged in the other direction, as when a composer represents the turbulence of a stormy sea and we are led to infer a corresponding inner turbulence of an imagined agent. Eero Tarasti astutely warns us that in seeking for agency in music, "we run the risk that the subject we have found in the music is none other than ourselves" (Tarasti, 1994: 109), but that is indeed the promise and the allure of Romantic aesthetics, well-encoded in such staples as nature-painting. Already Beethoven, in his famous comment on the Sixth Symphony, urges the interpreter to move beyond simple pictorialism to the expression of feelings upon encountering nature in a visit to the countryside.

One of the issues encountered even in this simplified account of four types of agency is determining how a composer distinguishes between or among the first three as implied by the score. Elswhere (Hatten, 1994, Chapter 1) I have demonstrated a case of internal vs. external agency in the slow movement of the Hammerklavier. The "willful" stepwise ascent and search for a cadential resolution in the continuation of the second theme group is frustrated more than once before unexpectedly arriving at the desired cadence by means of an apparently unwilled but providential moment of insight--cued as external to the will of the implied principal agent by the surprise modulation and monolithic texture that ensues. Whatever one interprets that external agency to be, it provokes a reaction in the principal agent that is akin to a psychological moment of insight, if not a mystical epiphany, that appears to have come unexpectedly from outside the principal agency.

There will inevitably be cases where it is not as easy to decide among types of agency. I will expand on an example that Raymond Monelle (1992) presents in his book to illustrate my theory of markedness. The unexpected event is more precisely an instance of strategic markedness, in my theory. In the recapitulation of the first movement of Beethoven's Waldstein piano sonata, Op. 53, the octave arpeggiation that in the exposition led dramatically to the dominant is shifted at the last moment up a half-step to Ab. This provokes a sequential echo on Db, also ending on a "wrong" note, Bb. What follows in mm. 171-73 is a parenthetical three-bar modulation that returns to C major for the counterstatement of the main theme, powerfully punctuated with a subito forte in m. 174.

Note how the three-bar response and recovery to this dramatic moment appears almost to trivialize it, by treating it playfully and resolving its dissonant threat. The problem, however, is whether the previous wrong notes are a fateful injection by an external agency--surprising, disrupting, and briefly deterring the principal agent's forward progress--or whether the principal agent has in some way "willed" the initial disruption, perhaps speculating upon it in the sequence, and then exhibiting a certain power by dismissing the potential threat through a parenthetical modulatory return to C major. Here, a performer's gestures can be quite significant in projecting one or the other possibility. In the first, the principal agent would appear as surprised as the listener is meant to be, whereas in the second, the principal agent acts like a powerful magician, delivering the unexpected Ab and Bb in the role of one who is in on the surprise, then provoking the listener's sense of wonder at the mastery that can transform a threat into a trifle. The second interpretation would be my choice, and I think the composition supports it (note the pianissimo dynamic and near-perfunctory texture in the three-bar return to C, suggesting an ironic dismissal on the part of the principal agent, followed by heroic emphasis upon the recapitulation of the main theme). But the consequence of this interpretation is to introduce our third type of agency, that of the narrative persona which, in telling its own tale, can also direct its own adventures. I mentioned "power" in this respect, and indeed, the subjectivity that we might attribute to the principal agent, and perhaps assume for ourselves as listeners or performers, is that of a super-subjectivity--one which experiences not only immediate power, as in commonplace heroic gestures, but narrative power, as in the ability to determine one's own fate by provoking and then dismissing an imagined threat as harmless. Of course, as easily as we can appropriate such seductive super-subjectivity for our own psychological ends (or needs), we can attribute it to a persona of the composer--here, Beethoven as hero, to echo Scott Burnham's (1995) profound study of the consequences of that attribution. What I call "shifts in level of discourse," which may also imply this third type of agency (internally narrative), are often cued in Beethoven by sudden stylistic changes, use of the "recitative chord" (major 6/3) and/or topic, or other highly marked contrasts (see Hatten, 1994: 174ff.).

I turn now to the consequences of Beethoven's deliberate mixing of gestural agencies in the Finale of his Sonata in C for Piano and 'Cello, Op. 102, no. 1, where the very presence of another performer already introduces further dialogic complications. In m. 1, the opening gesture sets the comic tone; this is the first time in the work that a gesture has ended on the tonic note, yet its flippant release is hardly sufficient to accomplish closure. Unlike the similar motive (5-6-7-8) in the Alkan, this is not a satisfactory drum riff since it doesn't resolve to a metric accent. The cello immediately grabs the released tonic, however, and sustains it, with a gesture that undercuts the witty release as though proposing a more serious, if still speculative, consideration of the tonic degree. That the piano and cello are virtual representatives of kinds of agency, rather than defined and consistent characters in their own right, is obvious from the immediate reversal of roles in mm. 3-4: this time the cello releases and the piano catches the tonic to sustain it.

The movement contains many surprise undercuttings, such that the process of undercutting itself may justifiably be considered thematic and hence part of the plot, not merely the means of manipulating the plot narratively. But most of the undercuttings are mutual--both performers undercut a phrase goal with a sudden drop in dynamics, as in mm. 12 and 20. Thus, both performers participate in an implied narratizing agency, beyond their dialogically-opposed principal agencies. Since the narratizing has been absorbed, as it were, into the very fabric of the thematic discourse, we may experience the kind of super-subjectivity that was only briefly engaged by the Waldstein example.

This kind of gestural interpretation is difficult to unpack, involving as it does a compression of multiple types of agency, and I have yet to treat the two performers' own instinctive additions (our fourth category of agency). But in terms of the implied agencies of the score alone, I would propose gestural and agential complexity as a significant innovation of Beethoven's later period, stemming from the greater motivic or thematic concentration of events that many commentators have emphasized. Beethoven's practice, however, goes beyond mere temporal compression of ideas and their gestural contrasts. The specific density of Beethoven's late style is also a result, I would claim, of the multiply intertwined strands, and levels, of gestural agency, in that so much happens at the level of the story and at the level of the willful manipulation of that story. When such manipulations become integral parts of the thematic process itself, as in this finale, or the openings of the string quartets Op. 130 and 132, then we are describing a technique that is quite ahead of its time musically.

The consequences for the performer in terms of gestural realization can be near-schizophrenic, if one attempts to switch among gestures, and their implied agencies, in such quick succession. Listen to a performance of this work. Can you experience the intertwined strands of gestural agency as contributing to a kind of super-subjectivity which somehow absorbs all the roles? And does that larger, Romantic-ironic embrace lend the comic-heroic style of this Finale a depth comparable to that of the heroic-tragic genres with which we are perhaps more familiar?


1. Beethoven employs a similar device in his Diabelli Variations, Op. 120, with the ironic or parodistic quotation of Mozart's "Notte e giorno faticar," from Leporello's aria in Don Giovanni. Since this variation is also number 22, and the riff motive is the same, perhaps Alkan is also troping on Beethoven! (For an interesting account of the significance of Beethoven's parody of Mozart in the context of the entire variation set, see Kinderman, 1995: 213).
Back to where you left off.

2. It is prefigured in the novels of Romantic contemporaries such as Jean Paul Richter and E. T. A. Hoffmann, who were to have such an obvious impact on Schumann.
Back to where you left off.


Analyze thematic gestures, their developing variation, tropological potential, and implied agencies in the entire Sonata for Piano and Cello, Op. 102, no. 1 by Beethoven. Is the theme of the last movement prefigured in the opening theme of the work? What features of the thematic gesture are developed throughout the piece? How do topics play a role in the transformation of the gesture in the Finale?


Burnham, Scott. 1995. Beethoven Hero. Princeton: Princeton University Press. Hatten, Robert. 1994. Musical Meaning in Beethoven: Markedness, Correlation, and Interpretation. Bloomington: Indiana University Press. Kinderman, William. 1995. Beethoven. Berkeley: University of California Press. McNeill, David. 1992. Hand and Mind: What Gestures Reveal About Thought. Chicago: University of Chicago Press. Monelle, Raymond. 1992. Linguistics and Semiotics in Music. Chur, Switzerland: Harwood Academic Publishers. Searle, John R. 1970. Speech Acts. Cambridge: Cambridge University Press. Tarasti, Eero. 1994. A Theory of Musical Semiotics. Bloomington: Indiana University Press.

copyright 2001, Robert S Hatten.
Go to Lecture: One or Two or Three or Four or Five or Six or Eight
Go to Course Outline
Go to Cyber Semiotic Institute Home Page