SEMIOTIX Review of Books

By David Lidov


Robert S. Hatten

  1. A Theory of Virtual Agency for Western Art Music 

Indiana University Press 


Robert Hatten’s third book on musical meaning, from the Indiana University Press series which he edits, Musical Meaning and Interpretation, is a striking advance for the semiotics of music, specifically, for the semantics of music.  The end of the title, “For Western Art Music” merits some critical reflection, but wait on that for last.  Virtual Agency once more achieves Hatten’s excellent balance of the “how” and the “what” musical compositions mean.  I start with an overview of the problem and, next, of the book itself, not quite chapter by chapter. As an appendix, I note some further semiotics connections.  Many readers of this online magazine don’t have the training in musical studies the book assumes. Nevertheless, that needn’t prevent them from knowing a bit about it.  This review avoids musical technicalities. The book is important for all who seek a coherent notion of semiotics. The semiotics of music offers a key test of the validity of any general semiotic theory because music is so elaborate, so varied and so different from verbal language.  Some of your students may well need a pointer to this finely reasoned theory and its treasure of references.

Virtuality.  The chief task of this book is to show how works of music evoke and constrain images of subjectivity—the virtual experience of virtual subjects.  We deeply appreciate many works of art because they afford us the experience of inhabiting another mind, as if knowing from inside the subjectivity of another person, or what it feels like to be such and such.  Such immersive participation is available in all art media (and elsewhere) but not in the same ways for all the arts. Susanne Langer’s Form and Feeling (a book too little studied, I think) explains her analysis of what is shared and what is different in the virtual images that different arts construct.  In Langer’s schema, dance creates virtual power, music virtual emotion, poetry virtual memory, and so on. She is the first writer I know of who broadly generalized a notion we are most familiar with in literary criticism.  Narrative may let us imagine both the mind of the narrator, and a fictitious world populated by fictitious agents whose own subjective experiences we will sometimes fantasize but probably not fantasize haphazardly. Though our individual memories and fantasies become ingredients, what we read (and, as Hatten shows, what we hear) provides warrant and guidance and form.  According to the effectiveness of our own integration of mind and body, we can feel that a dance takes us into the mind of a danced character. Pictures and sculptures can provide a sense of the mind of a person portrayed or of the mind of the painter or sculptor gazing, and so on. Though not necessarily representing them by formal schemes, criticism for all art media has long appreciated the virtual scenes and actions and characters art creates.  But in the previous century, music’s scholars had become quite shy of affirming equivalent powers for music. Thanks in no small part to his own prior efforts over three decades, Hatten’s newest book is not isolated at this point. As the title announces, it takes up the matter head on, showing how instrumental music establishes a firm basis for imagined virtual, volitional agents, often very abstract, expressing something of their own feelings and situations through sound. (Please keep in mind that when I write “imagine” my meaning is not “visualize.”  The relevant sense is closer to “feel,” “intuit,” “hear something as,” “sense,” and so on.  English is a bit patchy in this department, and I don’t deny that someone may also visualize.) Even for conservative music theorists and musicologists, the reasoning and evidence is solid enough that the analytical and positivistic fashions that discouraged this type of investigation seventy years ago can have no sway.  

Langer’s work is not cited by Hatten, and his musicological elaborations make her proposals, in retrospect, a farsighted but still wishful sketch of a semiotic perspective on music which she didn’t have an adequate tool kit to develop.   In music, we find semiosis deploying signs that, by referring to a fictive object, bring that object into (virtual) existence, not simply modifying—though this can also happen—our conception of something we took to have had a prior reality (Peirce’s immediate object), but causing a new entity to appear ex nihil.   Well, we know that’s one of the ways signs act:  A yellow line down the middle of the road performs lanes, previously invisible, and the musical theme may evoke a willful agent unknown to the world before.

The Structure of the Theory.  Hatten’s book provides both a model of the virtualities that composed or performed musical sounds  evoke and a deeply thoughtful discussion of principles governing the relations between our music hearing and our understanding of the entities modeled.  For the model itself, he acknowledges a debt to the narratology of Greimas (1966) via the studies of Byron Almén (2008), though that debt is not extensive.  Contemporary theories of cognition also offer very decisive influences. Hatten explicates the cognitive act of hearing virtual images in music as progressing by a chain of inferences.  When these inferences are very elementary, the processes of inference in musical listening are necessarily unconscious, but where they become complex, they become available to introspection, even if usually not attended to.

After a winning introductory discussion of virtuality, acknowledging our age of “virtual realities,” the book’s ten chapters progress, with some fruitful detours,  to explications of the accomplishments of imaginative hearing from simple to complex.

At the most elementary level, we hear music as sound in motion in ways that can not be accounted for by physical acoustics.  At this level, where inference and perception are hardly separable, we find “actants,” minimal actions of musical sounds within a tonal and metrical spaces (scales and hierarchies of beats) imposing forces of attraction and resistance on those sounds.  We intuit musical space as a virtual environment. Because motion in this environment implies an expenditure of energy, we readily imagine musical actants as (virtually) gestural and energetic. (The notion of “musical gesture,” distinct from musical representations of bodily gesture, is elaborately developed in Hatten’s previous work.)  From our perception of virtual gestures we infer virtual “agents” that are usually life-like, with hints of the human, though natural subjects like winds and waves can also play and be heard as taking on life-like animation. While actants are minimal units, agents can maintain identity and integrity over a sequence of actants giving rise to our further inference of embodied human agents that will or suffer their own movements and take on (virtual) actorial roles.

Virtual agents, actants and environments in music have character.  Indications in music that warrant our perception (hearing) of character and (qualitative) expression are primarily of two types for Hatten, gestural and topical, both elaborated in his second book (2004).  Musical gestures are shapes in time that may acquire meaning both from biological and cultural sources. What musicologists call “topics”, a strong specialty of Hatten’s and essentially what literary critics call “topoi,” are cultural and depend on intertextual reference.  In two earlier books and in numerous lectures and articles, Hatten has developed a subtle and superbly articulated system that describes relations within a composition between topics (e.g., they may “trope” one another) and their integration with musical gesture, as well as explicating the enchainment of these atomic units as longer forms.  

Musical character, like musical action, is also perceived by inference.  On the chance that it may be a comfort to some semioticians, I would point out that the inferences Hatten attributes to musical (and other artistic) understanding are not different in principle from those we talk about in logic or in cognitive processing generally.  You will find, if you wish (not Hatten’s concern) Peirce’s three categories: There are deductive inferences when musical features iconically link to style categories (for arbitrary examples, to dance types or military music) and allude to the ambience and codes of that style or when an indexical linkage of sound and neuro-muscular patterning lets us hear music’s virtual gestures as sad, angry, loving, etc.   We make inductive inferences when music new to us provokes us to expand our stylistic competency as listeners by generalizing over instances to acquire rules.   But most importantly for the theory of virtual agency, we abduce that certain actants will make sense if we imagine a living actor present as their author and abductively understand musical gestures by attributing the feelings they evoke to that fictional author.  The difference between these inferences and inferences in math and logic are less differences in kind than differences in possibilities for verification. (The normal test for music would be whether or not we can sustain them as the music goes on.)

The initial levels of inference sketched in my previous paragraphs have their exposition in Hatten’s first four chapters, separated from the next five by a brief and, for me, very remarkable Interlude I.  Hatten’s specialized book, demanding a high level of competence in music theory and a sophisticated comfort with musicology is hard work.  The few pages of this Interlude bring a change in tone as if the sun now reached the angle to pour its light through the rose window of a stern cathedral.  The light transforms what has been predominantly, despite emotionally compelling analyses of a variety of musical excerpts, a complicated study of notes.  The book now becomes a study of culture.  

A motivation for the more colorful light is a necessary tussle with the idea of embodiment, a theme which has offered a very productive and appropriately popular tool in semiotics generally, in the “new musicology,” and elsewhere.  We can (as my own writings often do) interpret musical motion via an identification with it wherein music evokes our own bodily movements, gestures and postures, the identifications visible in dancing, marching and conducting but also the identifications fully interiorized.  These bodily movements that “move to” the music might be remembered, imagined or simultaneously enacted, might be fulsome or diminished, subtle and invisible. But Hatten, pointing out that we don’t need to move our lips when we read, argues we might skip that phase altogether, entering at once into the world of music’s thought and feeling.  Embodiment, taken literally, falls short of accounting for our capacity to hear virtual musical agencies with sufficient distance that we can experience sympathy towards or repulsion from their expressed feelings, which are not ours (p136).   For example, I experience such distanced sympathies in instances in Mozart’s operas and with such distanced repulsion from moments in Mahler’s symphonies as well as some of Leonard Cohen’s and Bob Dylan’s songs.  I agree with Hatten that such distancing of the virtual emotion reduces our reliance on arguments from embodiment. In sum, we can get to mind without a detour through body (a thought that makes me curious to revisit Langer).  Our author plays with “endmindment” as a possible substitute for “embodiment,” yet without any pretense of fully resolving the issues. The discussion of embodiment pivots us to an exploration of subjectivity that will turn to Chekhov and Boethius, among others, and then to initiate a perspective uniting the trajectories of virtual actors in a single musical composition as threads of thought, possibly conflicting, in a single virtual, mind.

For readers schooled in music history, this notion may recall Franz Liszt’s explanation of his symphonic tone poems.  (In reality, Liszt et al.  It is well established that Liszt relied on Marie Comtesse d’Agoult and Princess Carolyne zu Sayn-Wittgenstein in transforming his ideas—or ideas emerging in their conversations—to words on the page.)  He proposed that the modern epic, which his composition transposes to the orchestra, differs from the classical epic in that the conflicts and resolutions are struggles in the hero’s mind rather than on the battle field.  

 Artists and philosophers certainly did thematize subjectivity in the 19th century, and it is interesting to encounter this awareness of music as representing an inner dialogue, but for Hatten, music does not need to wait on philosophy or art to develop such representations.  

What I envision. . . is a historically informed yet theoretically and hermeneutically enhanced exploration of what it was possible to hear . . . in music. . . . Indeed, recent scholarship explores how modern subjectivity emerged in the face of crises that may have undermined the very coherence of the subjectivity that I am proposing as a substrate to our understanding. . . .  [I however argue] for a basic level of subjective consciousness that embraces those capacities found already in Boethius. . . . and robust enough to have persisted over the last fifteen hundred years of Western music . . . . (p149)

Music has its own evidence.  Hatten shows how likely it is that richly and personal emotional imagery is central to repertoire of Hildegard von Bingham’s compositions surviving from the 12th century and proposes we have no positive reason to presume it excluded from even the earliest music known only from archeological evidence, evidence that includes hints of music’s spiritual importance to its performers.

From the outset, the expository method in Hatten’s chapters is to indicate a terrain of representation in general terms and then follow that up with detailed analyses of passages of music in score examples to show how such representations are “staged.”   For example, he points out how Beethoven initiates an irruption of unpredictable lyricism at exactly the moment when a conventional continuation of his theme would have turned toward resolution and closure. Of course the listener doesn’t need the analysis to feel the consequence.  We hear a virtual subject manifesting an expression of ecstatic “free will” (e.g. p154). Where does this place the listener’s will in relation to the virtual images constructed? His chapter on virtual subjectivity concludes in addressing a problem that has been a recurring nuisance for music aesthetics, developing a perspective that will be enlarged in the chapter following.  It is a commonplace that music readily accepts the most varied projections of feeling stemming from listeners’ own experiences and fantasies. The problem is to addresses the distinction between what music itself represents and what we bring into the music that is extraneous. Hattten frames the distinction by way of what he terms “aesthetically warranted emotions”. This term miscued me at first.  It means no more than warranted by the composition itself in its cultural and stylistic context. His development of the distinction is positive; that is, he does not elaborate exclusions but marshals evidence for the breadth and integrity of direction provided, first of all, by the notes themselves in notated scores of music and, later on, options in performance that specify and amplify agential and emotional interpretations of musical scores.  

Perhaps the most sophisticated of the virtual agencies Hatten traces is “narrative agency.” Although we don’t quite trace it back to Aristotle’s distinction in the Poetics between narrating and acting a plot, Hatten may be ahead of the crowd in seizing on the complexities acquired by music when it captures a narrative tone of voice, making narration an agential action.  Still, in his fairly brief but expert chapter on Virtual Narrative Agency, Hatten remains, I think, unnoticing of a problem that has besieged the recent turn towards narrative in musical explication.  Musical narrative is generally proposed as the longest (durationally) frame for musical meaning, that is, as an interpretation of a musical work (or movement) as a whole, but I would suggest that narrative has no exclusive claim in this regard and needs a comparative context.  That context should include, as alternatives to narrative structure, argument (in the philosophic, not dramaturgical employment of the term) and perhaps what painters call “composition”, sometimes but not necessarily non-mimetic, and some others that would devour too much space here.  But the central point is that we can often hear music as telling a story, and telling is agential.  Music that strongly evokes narration seems magical in its power for such an evocation, and part of the explanation is here.

Interlude II, Chapter 10, and a Postlude, will present important summations and perspectives, but the climax of the argument developed though the book is Chapter 9, which proposes “An Integrative Agential Interpretation” of Chopin’s Ballade, No. 4 (in F minor).  This is the last of Chopin’s Ballades.  Here, Chopin explores an aesthetic of extreme sophistication.  Hatten’s interpretation shows

“how an agential approach can incorporate a range of analytical evidence—formal functions, genres, topics, tonal structure, rhythmic/metric structure, motivic structure and the expressive implications of melody and harmony—leading to the narrative-dramatic emotional journey of a powerful subjectivity.”  (p244)

He accomplishes all this and more—and yet leaves me a little frustrated.  The analysis, taken on its own terms as a demonstration of what musical academics call music analysis or what literary scholars might call “musical criticism” is magisterial.  Not only are all the sorts of data he lists above marshaled and sensitively coordinated, but the circumstance of dealing in detail with a complete, extended work allows distinctions and subtleties his prior examples could not encompass, allows him, for example to designate, giving very good reasons, the first of the last two recurrences of a theme as an “epiphany” and the second as an “apotheosis.”  He understands the Ballade as establishing an extended, coherent form by using the resources of classical sonata rhetoric without adopting its limitations and anti-narratological constraints.

But there is an elephant in the room.  Chopin’s free constructions in the four Ballades (along with some works of others, e.g., Schumann’s fourth Novelette and various works of 19th century program music) are appreciated as opening up for entire musical compositions the possibility and the ideal of achieving original, sui generis, musical architectures.  In many cases, the technique for such projects was to abstract models from literary texts that could become musical plans.  That was so for the Ballades—almost.  Chopin attributed his inspiration for these four works to four specific poems of his countryman, Adam Mickiewicz.   For the first three, it is easy to find musical allusion to plot, situation, character and dramatic shape. The poem for the fourth, (“The Three Brothers”) has no plot or situations of interest.  Are we misdirected to look at that poem? Or, is there a relation of some other sort, perhaps in tone, perhaps in certain ironies, perhaps in other aspects of the virtual narration? I have seen the poem only in a French translation from which I take no clues.  Neither Dr. Hatten nor I have the requisite expertise in literary Polish to undertake a resolution of this mystery. At the same time, it is very tempting to think he may have given us the tools we need to solve it. On the one hand, Hatten’s election to focus on this possibly marooned, possibly not, Ballade enhances the brilliance of his findings and interpretation.  On the other hand, I feel, in the end, somewhat adrift myself, wondering what is, in summary, the plan and genre of the work?  That may be a very good ambiguity to live with. We have done that in fact for a century and a half. The haze of the elusive form of this Ballade, its extraordinarily untethered musical poetry, is surely one reason for its special place in our affections.

Appendix.  I’d like to note an intersection between current or recent zones of discussion in musicology and general semiotics regarding hierarchy in narrative structure, largely but not entirely independent of each other in their development. 

In recent decades, musicology has shown a resurgence of interest in narrative as a way to model music.  Hatten’s work draws on and contributes to this quest. Dramatic trajectory is a key element in what he considers music’s “expressive meaning.”  More distinctive in his work than the attribution of story or story-like structure to music is his account of ontological status and generative source for these meaningful images of music, with his focus on the chain of inferences that takes us from acoustical to virtual emotional experience, from notes to plots, beginning with “actants”. 

Greimas took the word “actant” from Lucien Tesnière 1893-1954, a French Linguist, who tackled the problem of describing how linguistic coherence emerges when a “structural order” in conception is transformed to a linear order.  The nuclei of his molecules are verbs, and in this context he regarded actants as analogous to morphemes. Taking “argument” not as I did above but in its mathematical sense (as referring to the variables on which functions act) actants are the “arguments” of verbs and among these “arguments” may be agents.  For some verbal concepts, like “sunrise,” it is not readily evident how to separate virtual actions and virtual agents. I would suggest this is generally the case with minimal units of musical motion. In the end, this distinction is not at issue, as Hatten moves from atomic levels of description to bigger elements—themes, expositions, movements, and so on.   

Unknown to Tesniére or, at least until 1958, to Greimas, Vladimir Propp (1928 in Russian, 1958, English translation) developed a system of analysis for narrative that sharply distinguishes between elements of action  and categories of character or agent (hero, villain, etc).  Propp’s purposes have nothing to do with Hatten’s but neither do Greimas’s.  Maybe Tesnière’s more so. In one sense or another, all three authors are very concerned to find a deep minimal vocabulary of minimal units or relations below their surface realization.  This is not Hatten’s priority.

Nevertheless, in this  history, I think we find the solution to a critical question which greeted Hatten’s theory of musical gesture, and which neither he nor I (who both tried) have answered in the most straightforward way.  

In the conventional structural analysis of music, nearly minimal components (groups of a few notes) are frequently called “motives”.  Though it is frequently remarked that motives are the smallest units in music that can carry a distinctive feeling, theories of motivic structure don’t problematize the development of feeling; the relevant theories are syntactic.  For linguistics, what is special about morphemes is that they bridge syntactic and semantic construction, participating in both. The question posed to Hatten’s gesture theory, lurking on many lips I suspect, may have been best elaborated by Arnie Cox (2006 p42-60).   What is the difference between musical motives and musical gestures?  Why do we need a new term? The right answer, or a big part of it, may be that they don’t need to be different.  At a phonemic level, or at the level of notes, gestures and motives may well be the same kinds of thing, but the two terms signal in which of two contexts we are regarding them: one, the paradigmatic groupings within hierarchies of syntax; the other, the flow of interpretable virtual action that constitutes a semantics.  These “units” play roles in very different distributional networks in their two contexts.

A generation before Chomsky provided linguistics with a radical understanding of deep structure, a similar understanding of tonality was developed by Heinrich Schenker.  (The similarities, alas, have been more appreciated than the profound differences, which also obtain between narrative and sentential grammar. See Lidov, 2017, Chapter 13 and 2005, Chapter 6.)  Both had a profound influence but one which, for Chomsky, affected all disciplines then studied in semiotics while for Schenker, affected only music theory. Both of these intellectual tsunamis eventually needed correction, with notions like “thick description” crossing over from anthropology to ethnomusicology.  The range and nature and direction of dependencies of meaning on the sensory surface of a text and its inner, less immediately accessible structures remains ever since a focus of interest—both in thinking about music and thinking about everything else. Hatten shows himself well informed on the matter, not contentious, open, and he defends a strong investment in interpretations of the musical surface.  

In 1984, the closing session of a colloquium on the “Universals of Narrativity” held at Victoria College, University of Toronto, during the Fifth International Summer Institute for Semiotic and Structural Studies took up the problem of surface/deep-structure relations as a debate between Algerian Greimas and Paul Ricoeur.  Permit me a sprinkle of decontexualized citations:

Greimas: I feel that not only in semiotics but also in linguistics more generally, and, again, in the whole of the social sciences, the first major methodological step necessary is the identification of pertinent levels. It is only when a scientific project posits the objects it wishes to describe or construct at a specific level, and not at ten different levels, that it can hold a coherent discourse on these objects. This constitutes, I believe, the superiority of linguistics over the other human sciences.  (p554)

Ricoeur:  My claim here is that surface is more than a kind of reflection of deep structure, it is more than the instantiation of narrative rules that can be construed at the deeper level. Something happens at the level of figuration that makes the dynamism of the processes described possible.  (p552) . . . one could ask if it is not the surface of the text that provides the element of contingency and the series of unpredictable decisions which keep the story moving. (p554) 

(Greimas, 1989).

I have no special wisdom in these matters and only a limited knowledge of the scholarship that takes them up.  Music theory, as an academic discipline, demands a high level of thorough and rigorous accounting for structures, a weight Hatten handles adroitly, and his initial impulse to build a formal, hierarchical theory beginning with actants reflects this.  The one judgment I wish to suggest is simply that in spirit, Hatten appears to me much closer to the humane, discursive and inclusive hermeneutics of Ricoeur than the categorical, axiomatic, reductionist constructions of Greimas.


To end this review, one touchy matter, the last four words of the title.  Nowadays, in Toronto, when a responsible group holds a public meeting, it begins with a ritual called “land acknowledgement.”   The host recites their appreciation of knowing we are meeting on land that has been home to a series of societies for at least 5000 years and that Europeans are new arrivals, a matter many of us failed to contemplate thoroughly until recently.  Acknowledgement by academic scholars that their studies focus on a classical heritage which is a very small part of the whole world’s music is likely to be a politically correct ritual of the same sort and also a public manner of disclaiming elitism, though it is a bit difficult to disclaim elitism if “Art Music” is invoked but undefined.  It is good to see these social objectives advanced, but in this case, it is worth noting that the book may be misrepresented in two directions by this ritualistic element in the title. The misrepresentation of breadth: Most of the book’s attention goes to German and Austrian instrumental concert music from the 18th and 19th centuries, though there are important examples from other places, from vocal music, from earlier and later music, and pedagogical music; still, the biggest shareholders in Western Art Music—church and opera—get only glances.  More significantly, the misrepresentation of focus: Hatten shows himself well aware that the general principles he is advancing may have application to musics of other cultures and to popular music in current cultures. He says he lacks the competence to extend his work in those directions and will be glad if others do.  In fact, a major achievement of this work is to put on the table an articulation of principles that can be studied as candidate universals in two senses, either as representations of universal capacities of human listening or as axes of variation that explicate differences between musical cultures. In such a comparative perspective it might be possible to say positively what is characteristic in the European classical music tradition with regard to its construction of virtual agency, and the title might be justified.   Though that direction is not indicated, the potential contribution towards a truly universalist view is a major accomplishment of this book. The problem of semiotic universals has no inherent political bias.  



ALMÉN, Byron. 2008.  A Theory of Musical Narrative.  (Bloomington:  Indiana University Press.)

COX, Arnie. “Hearing, Feeling, Grasping Gestures” pp 42-60 in A. Gritten and E. King, Music and Gesture, 2006 (Aldershot, Hampshire: Ashgate).  

GREIMAS, Algirdas Julien.  (1966) 1983. Structural Semantics:  An Attempt at a Method.  Translated by Daniele McDowell, Ronald Schleifer, and Alan Velie. (Lincoln:  University of Nebraska Press.)

GREIMAS, Algirdas Julien; Paul Ricoeur; Paul Perron; Frank Collins.  1989. New Literary History, Vol. 20, No. 3, pp. 551-562.

HATTEN,  Robert S.  1994. Musical Meaning in Beethoven: markedness, correlation, and interpretation. (Bloomington: Indiana University Press).

  _ _ _   2004. Interpreting Musical Gestures, Topics, and Tropes:  Mozart, Beethoven, Schubert.  (Bloomington: Indiana University Press).

LANGER, Susanne K. K.  1941. Philosophy in a New Key: A Study in the Symbolism of Reason, Rite, and Art (Cambridge: Harvard University Press).

LIDOV, David,  (1999) 2017. Elements of Semiotics: a neo-structuralist perspective.  PDF at (1999.  New York: St. Martin’s Press).

    —2005A “Is Language a Music?” (Bloomington, Indiana University Press.)

PROPP, Vladimir. (1928) 1968.  Morphology of the Folktale. Laurence Scott, trans. (1st edition, 1958), Louis A. Wagner, revised 2nd edition.  (Austin, University of Texas Press.)


Be the first to comment

Leave a Reply

Your email address will not be published.


This site uses Akismet to reduce spam. Learn how your comment data is processed.