Rom, Uri (2021): Edward Klorman, Mozart’s Music of Friends: Social Interplay in the Chamber Works. Cambridge: Cambridge University Press 2016. ZGMTH 18/1, 187–194.
eingereicht / submitted: 13/06/2021
angenommen / accepted: 13/06/2021
veröffentlicht / first published: 28/06/2021
zuletzt geändert / last updated: 12/07/2021

Edward Klorman, Mozart’s Music of Friends: Social Interplay in the Chamber Works. Cambridge: Cambridge University Press 2016

Uri Rom

Schlagworte/Keywords: agency; chamber music; Kammermusik; music analysis; musikalische Analyse; Wolfgang Amadeus Mozart

This enticing and deeply informed call to regard Mozart’s chamber-music scores as embedding implicit agential scripts – to be brought to life by performing musicians – appeared in 2016, and has since then been the subject of a number of enthusiastic reviews, emblematic of the book’s eager reception and its high appeal for communities of musicians and music scholars alike. The existing reviews – some of which are by notable authorities in the fields of musical signification and late eighteenth-century music, such as Scott Burnham, Esther Cavett, Robert Hatten, Matthew Head, and Giorgio Sanguinetti[1] – effectively summarize the book’s content, its main theses, and methodologies, occasionally taking issue with specific aspects. Given that the book itself is introduced by Patrick McCreless’s review-like foreword (xiv–xx), the task of presenting Klorman’s argument in a nutshell to a potential readership and drawing attention to its multifaceted dialogue with historical and contemporary sources and theories seems to have been accomplished amply and thoroughly. Thus, in my current discussion I refrain from supplying another summary of the book or a comprehensive appraisal, but rather choose to engage in dialogue with some aspects of Klorman’s study from my vantage point as a music theorist and analyst with a specialization in musical corpus studies.

At the center of Klorman’s argument – developed at length in the four chapters of the book’s second part (“Analytical Perspectives”) – is the concept of “multiple agency,” introduced by the author as “a way of experiencing a musical passage or composition as embodying multiple, independent characters – often represented by the individual instruments – who engage in a seemingly spontaneous interaction involving the exchange of roles and/or musical ideas” (122; original italics). Klorman’s theory of multiple agency draws, on the one hand, on the historical conversation metaphor – often applied to chamber music (in particular to the string quartet) in German, French, and Italian sources of the late eighteenth- and early nineteenth-centuries (see ch. 2) – and, on the other hand, on modern approaches to musical agency in the writings of Edward T. Cone, Fred Maus, and Seth Monahan.[2] The impressive, meticulously collected historical evidence in the book’s first part (“Historical Perspectives”) converges with the analytical argument in its second part to yield a number of multiple-agency analyses of shorter and longer passages (mostly taken from Mozart’s chamber-music works), which ultimately culminate in a detailed whole-work analysis of the “Kegelstatt” Trio, K. 498 (ch. 7). In addition to analytical prose, these individual case studies take the form of annotated scores, in which the players’ (= protagonists’) subtexts are written directly into the music (these annotated scores are “brought to life” in supplemental online videos,[3] where they are underlain by Klorman and colleagues’ fine and expressive interpretations).

Klorman’s claim to spontaneous, or “seemingly” spontaneous interaction among the players in the quotation above goes back – at least partly – to a historical reality in which “private and semi-public musical gatherings […] seem often to have been organized spontaneously, with little or no rehearsal […]” (73), a claim borne out by some iconographic and written evidence regarding the ad-hoc circumstances under which Mozart’s music was sight-read in private circles or even presented in public concerts (9–16, 86–104). However, while there was surely a great deal of exalted spontaneity, for instance, during the sight-reading sessions on 15 January and 12 February 1785, in which Mozart’s quartets dedicated to Joseph Haydn allegedly sounded for the first time, there is nothing spontaneous about the composition process which gave birth to these works, famously described in Mozart’s dedication as “il frutto di una lunga, e laboriosa fatica.” Indeed, the concept of “spontaneity” in connection with Klorman’s theory of multiple agency seems to require a closer consideration.

Was Mozart meticulously setting the stage – for instance in his “Haydn” Quartetsfor an unrehearsed sight-reading in which the four players (excepting himself!) would be discovering the works for the first time? And did he anticipate his players to “spontaneously” employ their instrumental utterings to communicate with one another regarding such pressing performance-related matters as “which is the hypermetrically strong downbeat?” (Quartet in G major, K. 387, i, cf. 221–227), or “who is leading right now?” (K. 387, iv, cf. 118–122)? Many of Klorman’s verbalizations appear to deal with situations in which the players are busy guessing the outcome of a yet unfamiliar passage and struggling to come to terms with their own individual role in it. But somehow, I fail to realize why Mozart would invest in witty musical subtexts that are only valid for the first time a group of musicians are tackling a new piece. So, one would possibly better opt for a merely “seemingly spontaneous interaction,” in accordance with Klorman’s wording above. Yet, are then Mozart’s alleged virtual scripts to be understood as messages exchanged among ensemble members simulating spontaneity in spite of previous knowledge of the music played? Consider a cello’s utterance along the lines of “well, as we all know, we are heading towards a deceptive cadence here, but, please, dear violin and viola colleagues, do pretend to be preparing an authentic cadence with great vigor, while I’ll be pretending to blow off the party with my wrong bass note, upon which you will enact surprise!”[4] However, why would musicians who already know the score feign spontaneity in their mutual communications unless this were for others to perceive? So, there is finally the option of chamber musicians enacting roles in a well-rehearsed show of spontaneity meant for listeners/spectators. But in that case, the spontaneous circumstances under which a typical sight-reading of Mozart’s music would take place are of only indirect bearing on the players’ experience in a rehearsed Mozart performance. Moreover, under this assumption the adequate framework for describing the metaphoric transfer is possibly not that of “Chamber Music as Conversation” – notwithstanding the robustness of the conversation metaphor – but rather “Chamber Music” (or, for that matter, any instrumental music) “as Implied Theatrical Action,” whereby this action may be construed to simulate a spontaneous conversation, but also completely different scripts of reciprocated communication and action.

I raise this issue because, although Klorman addresses a variety of interpretation possibilities at various points of his study, I remain confused as to the conceptual framework within which he ultimately construes the metaphoric transfer to take place. Consider, for instance, the flirtatious dialogue between viola and piano that Klorman reads into the first few measures of the “Kegelstatt” Trio (274–275). This script is quite an exception among his other examples, not only in that it involves an implicit theatrical scene – the viola and piano’s lower range representing a gentleman bowing gallantly, but somewhat awkwardly to a lady (mm. 1–2); the piano’s middle range answering coquettishly (mm. 3–4) – but also due to Klorman’s insinuation that the scene in question is enacted not by just any implicit performer personages emanating from the score, but, very specifically, by Mozart himself playing the viola part and one of his favorite female pupils, Franziska von Jacquin, at the piano (268–273). This set of circumstances affords divergent readings ranging from a biographical one, involving a concealed love message by an actual Mozart to an actual Franziska, to an interpretation in which the player personas do not figure at all, and the music enacts a dramatic scene (a whimsical gentleman bowing to a coy lady) by means of more traditional musical agents (motives, gestures, and registral shifts). While this single example already seems to raise several incongruent modes of construing the metaphoric transfer, most of Klorman’s other verbalizations are suggestive of yet another framework in which player personas are discussing with one another structural, even analytical issues of the music they are performing during performance.

And here we touch on the central fascination, but also the main crux of Klorman’s theory. The idea of musicians communicating with one another about music through music is highly suggestive – and recalls, for instance, Hans Keller’s concept of wordless musical analysis.[5] However, these “Dialogues of the Chamber Players” seem to be only for the initiated. They deal with expert structural and compositional issues that concern (or, at least, ought to concern) any performing musician, but are quite opaque to non-musicians. Moreover, they do not necessarily converge with a piece’s or a passage’s expression as emanating from an observation of its other musical markers and, thus, may lead to competing interpretations. An example follows.

Embracing Klorman’s brilliant analysis of the intricate metrical ambiguities in the second movement of the “Kegelstatt” Trio (255–266), I can easily imagine the implied pianist at m. 57 thinking to herself (invariably, Klorman’s players are gendered) regarding the implied violist: “Oh dear, […] he really is lost! […] Maybe I can simplify my part and just beat time to help him out” (261). I can even imagine a bunch of connoisseurs sitting nearby to appreciate this comical sketch, but I fear this compelling little plot about musicians for musicians might be lost on a more distant or music-theoretically uninformed listener. Besides, I am not sure that this metrical imbroglio is the most salient aesthetic feature of this G minor trio section. Throughout this trio, the music exemplifies[6] a number of properties – such as a general sense of tonal obscurity, the viola’s “sneaky” triplets and the overall soft dynamics – which arguably characterize it as schattenhaft, an adjective used by Mahler to mark the shadowy scherzo movement from his Seventh Symphony. I argue that the metrical confusion vividly captured by Klorman’s analysis is construable as yet another marker of this “shadowiness.” While imagining the pianist beating time for the violist’s benefit is, no doubt, a captivating enactment of the passage in question, it brings, at the same time, a rehearsal room’s daylight (or, rather, a chamber-music salon’s artificial light) into what I would otherwise construe as a dim, nocturnal scene.

Whenever there are a variety of possible interpretational frameworks, one may famously resort to some version of Ockham’s razor to select the “simplest” one among them, that is, the one that involves the fewest assumptions and roundabouts. However, Klorman’s theory is far from parsimonious; in fact, it is a “burdened” theory. It calls for assuming not just the normal kind of metaphoric transfer, but also an agential potential of each of the voices/parts involved in a polyphonic composition, on top of all treating individual parts as dramatis personae with their individual characters, wills, whims, and interpersonal communication. In the race for simplicity, the theory of multiple agency possibly won’t take first place. However, if one embraces a post-structuralist stance on music semiotics, there is no fault with a burdened reading, because there is per se no single correct reading. In a variation on the razor principle, I would also be willing to put up with a complicated theory, as long as it yielded fresh and compelling insight into the music analyzed.

This is definitely the case with Klorman’s above-discussed reading of the metrical confusion in K. 498, ii: although I do not share Klorman’s view on what this passage expresses, I cannot help but being impressed by the coherence and penetration of his multiple-agency plot. By contrast, I am at a loss regarding, for instance, his repeated reference to deceptive cadences as embodying multiple agency (cf., e.g., 144, 147, and 172). Deceptive cadences – at least the customary ones involving a move from the dominant to the sixth degree – necessarily have the “deceit” occurring in the lower voice, most typically embodied – in the chamber-musical context – by the cello. Given this, I cannot see what additional insight may be gleaned from opting for a multiple-agency reading of deceptive cadences. The cello gentleman who thus sprouts into being will do just what any Classical bassline is wont to do several times in a piece: climb from the fifth to the sixth scale degree while the other voices follow an authentic cadential voice-leading pattern. The cello’s (= cellist’s) behavior being predictable in this respect, installing him as a persona unnecessarily complicates the explanation of a common harmonic procedure. Unless the cello’s “deceptiveness” can be compellingly embedded into a more comprehensive reading of the piece at hand, I fail to realize its insightfulness. Possibly, a further theoretical advancement on multiple agency would call for developing a set of criteria for determining under what conditions a multiple-agency reading supplies an insightful vantage point to justify the additional methodological baggage.

This brings me to the general question of context. Klorman’s analyses represent an array of isolated instances carefully selected to underline his claims. Among these are several gems – like the two examples from the “Kegelstatt” Trio discussed above. While one can already recognize the analytical potential of multiple agency on the basis of these examples alone, I argue that one thing that Klorman’s approach would also need in order to develop into a theory – in a more rigorous sense of testability and falsifiability – is a broader context for evaluating and corroborating the individual readings.

One such reading that calls for context is Klorman’s analysis of the beginning of the development section in the first movement of the Duo for Violin and Viola in B major, K. 424 (see Ex. 1). Here, Klorman’s analytical prose and his score annotations (shown in the example) assign the violin (= violinist) the initiative for interrupting the triple statement of the cadential gesture in mm. 79–84 by staging a sudden shift to the minor mode (m. 84). According to Klorman, “[t]he dumbfounded viola drops out, while the violin ponders the note A. The pianissimo marking, rare in Mozart’s chamber music, seems like an operatic aside (‘da sé’). With the A isolated and in a melodic register, the violin cleverly treats it as , inviting the viola back in for a new dolce theme in D major” (137–138).


Example 1: “Ex. 4.3” (137); Mozart, Duo for Violin and Viola in B major, K. 424, i, mm. 77–89

Judging by this analysis, which highlights the uniqueness of the passage at hand, one may get the impression that shifts to the minor mode at the development’s beginning represent rare occasions in Mozart’s music in general. This is, however, not the case. In my investigation of digressions to the parallel minor in Mozart’s fast sonata movements,[7] I found that minor-mode shifts at the development’s beginning are very common. The shift to the parallel minor in the passage under analysis is, admittedly, of a more specific type, involving a minor-mode repetition of an element initially stated in the major mode – a procedure to which I refer as “minor-mode echo.”[8] However, I was also able to locate over thirty such development-beginning “echos” across Mozart’s sonata-allegro movements. Over ten of these occur in chamber-music works. Example 2 shows one such instance taken from the first movement of the String Quartet in B major, K. 589.


Example 2: Mozart, String Quartet in B major, K. 589, i, mm. 67–82

The similarity between this quartet passage and the duo passage analyzed by Klorman is striking. Besides the principal key, which, strictly speaking, is not part of the musical structure,[9] the two passages share an identical strategy of “minor-mode echo,” transferring the exposition’s closing gesture from the key of the major-mode dominant to the parallel minor (F minor) at the development’s beginning (in K. 424, i, there is one extra repetition of the motive in major – directly at the double bar – before switching to minor). Intriguingly enough, the first violin is allotted a short solo of two measures in the quartet movement, too, in the course of which it climbs to a high A – in fact, m. 86 of the duo movement and m. 77 of the quartet movement are literally identical. From this point on, both developments modulate to the same key, D major, for the presentation of a lyrical theme in piano (in K. 589, i, this presentation, beginning in m. 81, is based on the movement’s principal theme). There is no way to prove this, but I consider it very likely that Mozart had the duo passage in mind while composing the beginning of the development in the string quartet movement (note also the identical emphatic jumps F4–F5 in the violin/first violin part in K. 424, i, mm. 80, 82, and 84, as well as in K. 589, i, mm. 71 and 76).

But what can we learn from this comparison regarding questions of agency and multiple agency in the two works? In the quartet passage – in a similar vein to the passage from the duo – the motivic repetition effaces the double bar’s formal boundary: this is a corollary of implementing the “minor-mode echo” strategy across the exposition-development border, and occurs in a number of additional works as well. More crucial for our discussion is, however, the question of the treatment of the minor mode: “who” brings it about? – how do the “others” react to this – and with what consequences?

Obviously, in the quartet the initiative for shifting to the minor mode in m. 72 doesn’t come from the first violin: if any single instrument can be made responsible for initiating F minor in this turbulent measure, it is arguably the second violin (= violinist), who is the first to introduce the pitch A. On closer inspection, this pitch is also initially introduced in the duo in an inner voice (m. 84) and not in the melody. While it would seem absurd to assign separate agencies to the violin’s G and D strings, the comparison to the string quartet calls into question Klorman’s assertion that the violin is the “agent” of the shift to minor. In fact, in mm. 79–82 it is the viola (= violist) who has the progression B–A, and the violin’s reasons for taking over at mm. 83–84 seem to have more to do with instrumental pragmatism than with questions of agency.[10] Arguably, listeners will not even notice that the shift to minor at m. 84 – occurring in a middle voice – is brought about by the violin rather than the viola (although the players are, of course, well aware of this); consequently, they will have difficulty recognizing the viola’s (= violist’s) silence in mm. 85–86 as a result of his being “dumbfounded” by the violin’s action. Needless to say, in the quartet’s development section nobody is dumbfounded by the shift to minor – all four players happily and noisily rattle throughout the minor-mode passage, just as they did in the preceding analogous major-mode passage. Although the transitory violin-solo passage in mm. 77–78 in the quartet movement closely resembles the pianissimo transition in the duo movement (mm. 85–86) in its “searching” character and pitch material, I don’t think that the transition to D major in the quartet can be compellingly construed as a corollary of former gestural and dramatic events (in the transitory passage from the duo, by contrast, the violin’s solo organically grows out of the preceding motive).

I am not sure to what extent – if at all – this comparison to K. 589, i necessarily engenders a rethinking of Klorman’s multiple-agency plot for the passage from K. 424, i. There is, of course, the question raised above regarding the violin’s agency in bringing about the shift to minor – but this objection could be easily overcome by maintaining that at least the player personas involved are well aware of the A being introduced by the violin. However, my point here is a different one: even if the validity of none of Klorman’s specific readings should be affected by comparisons to other works, such broader context is nonetheless crucial to enhance the validity of his interpretative approach as such. Provided that we construe the violin-viola’s “plot” at the development’s beginning in K. 424, i to represent a specific “behavior,” a comparison to other “behaviors” occurring under analogous conditions in other compositions is an inevitable step towards corroborating the reading. At least, a concept of context is among the core tenets of any repertoire- or corpus-related approach.

Finally, although musical agency and musical portrayal are arguably two different things, they are nonetheless closely related through the shared concept of musical personification. There are two striking cases of musical portraits in Mozart’s works that Klorman leaves out of account (perhaps because they are not part of the chamber-music repertoire?), to which I would like to draw attention here. One is a piano-sonata andante movement which, in a letter to his father from Mannheim, Mozart reports to have composed “entirely in keeping with the character of Mad:selle Rose.”[11] Rosina Theresia Petronella was the daughter of Christian Cannabich and, for some time in 1777, Mozart’s piano pupil in Mannheim. If the movement in question is, indeed, the middle movement of the C-major Sonata, K. 309, as is generally assumed,[12] this is notably Mozart’s only slow movement employing the varied-repeats technique particularly associated with C. P. E. Bach.[13] As a result, this andante contains an unusually large number of written-out repetitions of the main melody, whereby each repetition embodies a different set of ornaments. It is an intriguing task to try to relate this alleged musical portrait’s peculiar format to Mozart’s description of Rosina Cannabich as “a very beautiful, well brought-up girl. For her age, she has a great deal of sense and a mature nature; she is serious, does not speak much, but what she says – is said with charm and friendliness.”[14]

The other musical portrait is of a very different nature. It is Mozart’s fragment of a Rondo for Horn and Orchestra, K. 514, written for Joseph Leutgeb, a close friend of the Mozart family in Salzburg, who moved to Vienna some years before Mozart and maintained an ongoing professional and personal connection with the composer. The autograph of this fragment contains the nearest thing to an in-time verbalization of an instrumental piece ever supplied by Mozart. Staged as a dialogue between the composer and the horn-player personas, this vivid text, at times verging on the rude, merits a close examination of the way it simultaneously supplies a commentary on the music, reflects the state of mind of the two personas involved and plays out the humorous relationship between these musical friends/adversaries.

In his pioneering study, Klorman compellingly frees Mozart’s chamber musicians from the confinement of the two-dimensional music sheet and has them step as three-dimensional figures into the real (or metaphorical) world. As demonstrated in his well-chosen, insightful case studies, these virtual figures – at times bringing back to life Mozart’s circle of intimate musical friends – have novel and important things to tell us regarding some of the composer’s most significant and most cherished chamber-music works. While the theory-of-multiple-agency component is arguably still in need of refinement and elaboration, Klorman’s opus offers a highly rewarding combination of historical scholarship, analytical penetration, and a practitioner’s straightforwardness and ingenuity. As a profound, original tribute to Mozart’s chamber music, this is a study that already has had, and will undoubtedly continue to have, far-reaching repercussions.



Burnham 2018, Cavett/Head 2017, Hatten 2017, and Sanguinetti 2018.


Cone 1974, Maus 1989, and Monahan 2013.


Accessible through (16 Jun 2021).


This is my re-composed version of Klorman’s description of a deceptive cadence as a witty inganno staged by the cellist, paraphrased in Sanguinetti 2018, 243–244.


Keller 2001.


My proposed reading is informed by the concepts of “exemplification” and “expression” as laid down in Goodman 1968.


See Rom/Rosset forthcoming. I discuss digressions to the parallel minor at length in Rom 2011, 267–569.


Ibid., 400–409.


On the face of it, a piece’s structure is determined by the relations among its pitches and not by its absolute pitch level. However, while modern music theory generally assumes transpositional equivalence among all keys pertaining to the same mode, my research has shown that in Mozart’s music there is a statistically significant correlation between the choice of absolute key on the one hand and attributes of the musical structure on the other; see Rom/Rosset forthcoming.


In mm. 80 and 82, the pitch A is doubled in both instruments: owing to the open A strings, which co-resonate one octave higher, this doubling is acoustically satisfactory and quite reliable in terms of intonation. This is, however, not the case with regard to the pitch A. This is probably the reason why Mozart refrained from doubling it, but this forces him to choose between the violin and the viola. Note that the same instrument which plays the A also has to play the B, because the latter pitch – being a seventh – must be resolved via a step downwards. Had the viola taken over this step B–A in mm. 83–84 – in analogy to mm. 79–80 and 81–82 – this would have meant that the violin, which in that case would have had a G in m. 83 – would not have been able to resolve that G in m. 84, since a doubling of A is – as argued above – undesirable, and the low F doesn’t exist on the instrument.


Mozart’s letter from Mannheim to his father in Salzburg, 6 December 1777. See Mozart Briefe und Dokumente – Online-Edition, edited by the Internationale Stiftung Mozarteum, Salzburg (, 16 Jun 2021).


See Plath/Rehm 1986, XIV–XV.


See Rom 2019, 68–70, 72.


See footnote 11.


Burnham, Scott. 2018. “Review of Mozart’s Music of Friends: Social Interplay in the Chamber Works by Edward Klorman.” Journal of Music Theory 62/2: 339–346.

Cavett, Esther / Matthew Head. 2017. “Review of Mozart’s Music of Friends: Social Interplay in the Chamber Works by Edward Klorman.” Eighteenth Century Music 14/2: 285–290.

Cone, Edward T. 1974. The Composer’s Voice. Berkeley: University of California Press.

Goodman, Nelson. 1968. Languages of Art: An Approach to a Theory of Symbols. Indianapolis: Bobbs-Merrill.

Hatten, Robert S. 2017. “Review of Mozart’s Music of Friends: Social Interplay in the Chamber Works by Edward Klorman.” Music Theory Spectrum 39/2: 261–265.

Keller, Hans. 2001. Functional Analysis: The Unity of Contrasting Themes. Complete Edition of the Analytical Scores. Edited by Gerold W. Gruber and Michael Meixner. Vienna: Lang.

Maus, Fred. 1989. “Edward T. Cone’s The Composer’s Voice: Agency in Instrumental Music and Song.” College Music Symposium 29: 31–43.

Monahan, Seth. 2013. “Action and Agency Revisited.” Journal of Music Theory 57/2: 321–371.

Plath, Wolfgang / Wolfgang Rehm. 1986. “Foreword.” In Neue Mozart-Ausgabe, vol. IX/25/1, Kassel: Bärenreiter, IX–XVII. (16 Jun 2021)

Rom, Uri B. 2011. Tonartbezogenes Denken in Mozarts Werken unter besonderer Berücksichtigung des Instrumentalwerks. PhD diss., Technische Universität Berlin. (16 Jun 2021)

Rom, Uri B. 2019. “‘To Gild Refined Gold,’ or What Mozart Didn’t Want Us to Embellish.” Music Theory and Analysis 6/1: 50–87.

Rom, Uri B. / Saharon Rosset. forthcoming. “Key-Specific Structure in Mozart’s Music: A Peek into his Creative Process?”

Sanguinetti, Giorgio. 2018. “Review of Mozart’s Music of Friends: Social Interplay in the Chamber Works by Edward Klorman.” Music Analysis 37/2: 243–250.

Dieser Text erscheint im Open Access und ist lizenziert unter einer Creative Commons Namensnennung 4.0 International Lizenz.

This is an open access article licensed under a Creative Commons Attribution 4.0 International License.