Reflective Performativity in the Interpretations of Nikolaus Harnoncourt

Read as PDF

Emil Bernhardt

Emil Bernhardt studied the violin at Barratt Due Music Institute and then composition at Norwegian Academy of Music. As part of his Bachelor, he was also an exchange-student in Freiburg, Germany for one semester. Later, he studied Philosophy (Master, 2012) at the University of Oslo.

After finishing his studies, Emil worked as a freelance music critic in several national and international papers and journals. Since 2012 he has regularly been reviewing classical and contemporary music for the Norwegian weekly newspaper Morgenbladet. He has also written for several international journals, given a lecture at the Darmstadt Ferienkurse about the music of Ragnhild Berstad, and made two radio-programmes for Deutschlandradio Kultur in Berlin.

Emil is currently a Research Fellow at the Norwegian Academy of Music, working on the performance practice of the Austrian conductor Nikolaus Harnoncourt.

by Emil Bernhardt

Music & Practice, Volume 5

Artists & Teachers

One major challenge when carrying out research on musical interpretation is to point out what is uniquely characteristic about the performer and the performance practice in question.[1] In my current research on the interpretations of the Austrian conductor Nikolaus Harnoncourt, one obvious point of departure is the concept of historical consciousness and, more generally, a certain reflection that constitutes the basis of his interpretative work. However, Harnoncourt is far from being the only interpreter whose performance practice is based on historical information, and one could hardly think of any conductor of interest who wouldn’t be conscious about, and in some way reflect upon, what he or she is doing. This makes the concepts seem, at first, too general. On the other hand, the problem is probably not the concepts themselves; what is needed is rather a clarification of how, in this case, the concept of reflection is to be understood in Harnoncourt’s practice. The following is an attempt to do so. Thereby I also want to suggest at least one specific characteristic of his work, specified by what I will call a reflective performativity. And, as we shall see, the argument will necessarily include a discussion of the concepts of performance, performativity and aesthetical experience.

My basic hypothesis is that Harnoncourt’s interpretations are characterized by a certain tension. This tension might be connected to the notion of historical consciousness – what is the relationship between the historical information, on the one hand, and our contemporary performance practice – and, more precisely, the actual performative moment – on the other? The tension might also be connected to a phenomenon which, according to Harnoncourt himself, is of crucial importance: the idea of music as language – how may we reconcile the relationship between linguistic concepts such as meaning and understanding on the one hand, and the capabilities or limitations of music, regarded as an abstract medium, on the other?

In this article, I shall discuss the tension through the use of the two concepts: “reflection” and “performativity”, proposing, first, that there is a tension, if not a contradiction, between the act of reflection and the act of performance and, second, that Harnconcourt, in a most interesting way, is able to activate this tension in his interpretations. More particularly, I shall argue that one specific characteristic of Harnoncourt’s interpretation is that reflection is not only located at a preparatory level or in the rehearsal work, i.e. on a level preceding the actual performance; rather, the reflection – or some notion of reflection that needs further explanation – remains present in the actual performance itself. More precisely, it becomes part of what I shall call the performative aspect of the performance. Hence the notion of reflective performativity. And as part of the performative aspect, the reflection also becomes part of my own musical and aesthetical experience as a listener. How reflection becomes an aesthetical parameter is, in my opinion, what makes Harnoncourt’s interpretations extraordinary. But how is this possible?

Generally speaking, one could say that reflection presupposes someone reflecting upon something; a reflecting subject who seeks to develop his or her knowledge about a certain object. Performance, and more precisely the performative aspect of the performance, on the other hand, is something that happens at the nexus of a host of simultaneously occurring events, where separations of subject and object become more blurred and where some factors must always remain, to a certain extent, beyond the performer’s control. Although a performance is something the performer engages in, with a combination of considered premeditation and a lively consciousness in the moment, he or she is not only assailed by a barrage of external factors in each instant but will also inevitably be subject to constant internal reflection upon what just happened, what is happening now and what will happen imminently. The question, then, is, first, how to understand the notion of a reflection that becomes part of the performative aspect of the performance and, thus, is something that in a certain way takes place both within and beyond the performer’s control, and, second, whether this notion (if valid) actually suffices as a specific characterization of Harnoncourt’s art of interpretation. This article deals primarily with the first part of the question.

I should like to underline that my perspective, at least to begin with, is that of the listener; my focus is on the aesthetic experience of listening to Harnoncourt’s interpretations. I am not trying to reconstruct his “technique”, or explain how he acts in order to achieve this or that musical result. Although Harnoncourt is a most eloquent commentator upon his way of interpreting, possessing an impressive rhetoric and discursive power shown in a number of articles and interviews,[2] these accounts won’t play any direct role in this article. Nor do I see it as essential that Harnoncourt himself should have recognized his thoughts in my discussion.  This may seem contentious. However, given that the reflection in question seems to be part of what I call the performative aspect of the performance, and thus takes place partly beyond the performer’s control, this kind of reflection is not identical with the reflection of the performer, understood as a precondition of his or her preparatory work. Therefore, how the performer him- or herself articulates his or her reflection, at least on an analytical level, does not affect the kind of reflection in the performative aspect that I am discussing here.

The notion of reflective performativity that I intend to elaborate is primarily meant to function as a description of the interpretation as it becomes part of the listener’s musical and aesthetical experience. This means that the notion of reflective performativity is connected to a musical quality and, thus, to an aesthetical judgement. When I find Harnoncourt’s interpretations musically attractive and interesting, it is, among other things, because they embody some notion of reflection which communicates itself to me. It is important to stress that when I find his interpretations beautiful, and connect this experience to some notion of reflection, it is not because they force me to reflect upon what is at stake in them. For the moment, this would be to confuse the notions of reflection and experience. The reflective element rather belongs – or seems to belong – to the interpretation itself as a quality that is simultaneously intrinsic to the interpretation and yet not confined by it. It is this dynamic and ambiguous duality which makes the result musically attractive; in some way or another it renders the interpretation beautiful – and individual.

A first attempt to connect the notion of reflection to some actual interpretation could start with the notion of interpretative choice. In Harnoncourt’s output there is of course an almost countless number of examples; here I will mention some quite recent and obvious ones: Harnoncourt’s latest recording of the three last Mozart-symphonies[3] is based on the idea that these three works constitute a whole, in Harnoncourt’s own words an “instrumental oratorio”.[4] Among a number of subtle details, on an immediately accessible level, this means that he minimizes the gap between the finale of Symphony no. 39 – which, according to Harnoncourt, has no real finale and ends abruptly – and the beginning of no. 40 – which, again, according to Harnoncourt, has no real beginning and starts from nothing. Another example is his latest recording of Beethoven’s symphonies no. 4 and 5.[5] Here, the opening theme of the 5th symphony’s second movement has an extraordinary, flexible phrasing, almost parlando-like in character. This probably refers to the, today somewhat dubious, hermeneutic theory of the German music theorist Arnold Schering (1877–1941), who explains the movement as a kind of prayer.[6] Finally, in his last complete-set of Schubert’s symphonies with the Berlin Philharmonic,[7] Harnoncourt makes some remarkable choices regarding tempo, perhaps most obviously in the last movement of the 6th symphony. This movement is clearly divided into different sections and Harnoncourt consequently separates them by use of different tempo layers and fluctuations.

It seems plausible to think of these interpretative choices as based on a reflection that has taken place at the preparatory phase. What Harnoncourt is doing – in the above-mentioned examples: his connecting of the movements and pieces (Mozart); his shaping of the melodic phrase (Beethoven); and his tempo-fluctuation (Schubert) – might be considered as realizations of this reflective work. If these interpretative choices are considered to be such realizations, we may suppose that they somehow bring the reflection (the one preceding the performance) into the performance. However, if this (preceding) reflection becomes dominant in the realization, the performance – to take things to extremes – will end up being nothing but a demonstration. And, in my view, Harnoncourt’s interpretations are undoubtedly more than that; indeed, I believe they surpass even those less banally demonstrative cases where an interpretation is the manifestation of a performer’s desire to realise an a priori reflection by all the musical persuasiveness available to him or her and, thereby, in the most convincing way possible – a kind of advocacy.

I shall return later to the issue of demonstration. For now, the important point is that the kind of reflection I am trying to get a hold of – a reflection in the performative realm, although probably somehow connected to the preparatory phase – should not simply be identified with that preparatory reflection. Furthermore, it is also unsatisfying to think of the interpretative choice merely as a signification of a reflection that has taken place. The interpretative choice is also something that is done – something that is being performed, one might even say that in, many cases, the choice is made on the podium. In other words, the notion of interpretative choice may still be relevant to the notion of reflection in the performative, even though the reflection preceding the performance (as the basis of the choice) does not suffice as an explanation of the reflection in the performative. The opposite extreme would be to think of the performance, and its performative aspect, as something that in itself possesses the capacity to reflect, and thus ascribing to the performance an ability that usually is ascribed to human beings. But this, although tempting as an anthropomorphic fantasy, in my view, would lead to bad irrationalism, even if applied metaphorically. However, let us leave the discussion of the interpretative choice aside, and turn to a closer look at the concepts of reflection, performance and performativity.

According to Lydia Goehr, the emphasis on performance activates what she refers to as the “knowing-doing distinction” (Goehr, 1998, pp. 134, 138). Although both reflection and performance might be called acts – both concepts refer to something that we do – I will argue that these acts differ in their relationship to knowing. Even though the act of musical performance is certainly based on knowledge, an important – if not essential – part of the performance is still spontaneous. What does that mean? Very generally, it has to do with a certain lack of control. And, in my view, this is the very key to the reason why we speak of performance and performativity at all. Of course, the performer, and especially a performer like Harnoncourt, knows, and thus is able to control, what he is doing. Still, however thoroughly based on reflective preparation, technique, education and experience, what happens spontaneously, will always, in a way, escape his, or any performer’s, control.

This becomes clear when we keep in mind the listener’s perspective. What the listener experiences, in a way, goes beyond what the performer is able to control and encapsulate. One may say that the performer knows what he is doing, but he will never be able to control completely how that what he is doing affects the listener. More formally put (and I believe this to be true as a condition for any human being): one is not able to make oneself into an object structurally similar to that which appears to be the object of the other’s (in our case the listener’s) experience. There is, in other words, an aspect or dimension to that which the performer actually does that goes beyond his or her control. And this very aspect is what I call the performative aspect of performance, it emphasizes the involuntary aspect of the performance. This performative aspect of the performance is essential for its musical immediacy. It is, in fact, essential for the musical quality and attraction, but, in a certain way, it refers to something of which the performer’s own knowledge is limited, if possible at all.

The act of reflection, however, seems not only to be based on prior knowledge, it also has the capacity, if not necessarily the specific aim, to evolve new insights. Perhaps the act of reflection could also be spontaneous and therefore independent of prior knowledge. But instead of, so to speak, leaving the sphere of that knowledge behind permanently, the act of reflection constantly returns to the object and to the knowledge about it (cf. re-flect – to turn back), a return that entails a certain resistance: First, the knowledge of the object upon which the reflecting subject reflects will always be incomplete, second, the reflective act will never leave the intentional procedure of the reflecting subject. Still, there seems to be no doubt about the subject-object relationship as the basis of the reflective act. Reflection, as we normally understand it, is done by someone who reflects upon something. Now one may ask how this connects to the perspective of the listener, especially since the act of reflection, instead of going beyond the reflective actor’s sphere, by its quality of returning, rather, remains within it. As already mentioned, one could perfectly well think of a reflection taking place in the listener, but this reflective act will never either constitute or replace the quality of experience. Another possibility that I will discuss later, would be simply to understand the concept of reflection differently, perhaps metaphorically, as a kind of resistance or friction, and now as a characteristic of a performance, i.e. as a listener’s object of experience.

Before that, Goehr’s invoking of the “knowing-doing distinction” deserves some further discussion and I should like to supplement it with an aspect central to the concept of performativity, as formulated by Erika Fischer-Lichte. As far as I can see, her discussion also supports my own thoughts on the concept of performativity, as indicated above. Opposed to the reflective act, which seems to presuppose the subject-object relationship, Fischer-Lichte points out that the notion of performativity seems to challenge the “subject-object” dichotomy (Fischer-Lichte, 2004, pp. 19, 33). Rather than referring to something outside itself, the performative (and, I would like to add: aspect of performance) refers to itself, according to Fischer-Lichte (Fischer-Lichte, 2004, p. 27). What the concept of performativity then refers to, is the fact that the result of a performer’s performance (i.e. its performative aspect) will never become an object of knowledge to the performing subject.[8] Rather, following Fischer-Lichte, the performative constitutes reality (Fischer-Lichte, 2004, pp. 27, 32), in other words: It is something that happens, in a certain way, as discussed, beyond the performer’s control. Hence, we may add to Goehr’s knowing-doing distinction the third concept of happening, which in turn makes the following order: “Knowing-Doing-Happening”. Now this might be supplied with the following, parallel concepts: “Reflection-Performance-Performative”. While “reflection” on the one side of the scheme is mainly connected to “knowing”, “performative” on the other is connected to “happening”, something that to a certain extent is impossible for the performer to grasp. Thus, regarding their connection to knowledge, the conceptual relationship “reflection”-“performative” is contradictory.

Given this background, let us think of the relationship reflection-performance and performativity once again. Since there is a connection between the knowing of the reflection and the doing of the performance, it seems possible to imagine an element of reflection in the performance, in the doing. The performer may also be reflectively aware both of what he or she is doing or performing, and of the fact that they are actually doing and performing (something) in the moment of doing it. However, when it comes to the element of happening – to what I have called the performative aspect of the performance –, it might seem difficult, if not downright absurd, to talk about an element of reflection. Then one could ask: But why is it necessary to locate the element of reflection in the performative aspect of the performance (the happening)? Would it not be sufficient, and by far more plausible, to think of an element of reflection in the performance (the doing) – cf. the discussion above about the interpretative choices as actualizations of a preceding reflection? When it comes to the plausibility, without any doubt the answer is yes. However, as I am trying to show, the notion of reflection in the performance, in the doing, does not cover the notion of reflection I am after here, a notion of reflection in the performative aspect, a reflective performativity. Or in other words: A notion of reflection that seems to be part of my musical experience as a listener, but without simply being my possible reflection upon what I am hearing, while, or presumable after I am hearing it. But why is it so? Why is it necessary to connect the notion of reflection to the performative aspect of the performance, and how could this connection be possible? Let us look at these moments in turn.

I will start with the necessity. As mentioned at the beginning, I am trying to get hold of a specific characteristic of Harnoncourt’s performative practice. I have always had the impression that there was some kind of tension in his interpretations, and also that this tension was somehow connected to a reflective work, a resistance or friction within the interpretation. However, both the idea of a connection between some kind of reflective work and the way someone performs, and the idea of reflection as a basis for the interpretative choices, seem to be valid for almost any interpretation of interest, Harnoncourt’s included. Thus, this way of thinking does not suffice as a description of the specific way in which the element of reflection becomes part of the interpretations of Harnoncourt. Moreover, and more importantly, when the notion of reflection seems to be important in Harnoncourt’s performances, not least by contributing to their beauty and aesthetical interest, it seems to manifest itself in a different way. When the element of reflection is a characteristic specific to Harnoncourt, when it is something that makes his interpretations especially interesting and beautiful, it is not because these interpretations are especially precise or convincing as demonstrations of some kind of knowledge. As mentioned at the beginning of this article, I would suggest that what makes Harnoncourt’s interpretations remarkable, is not the undoubtedly impressive amount of thorough reflective work that certainly lies behind them, it is rather the way this work somehow becomes part of my aesthetic experience of them, one might say: the way the notion of reflection becomes an aesthetical parameter. In Harnoncourt’s interpretations, the reflective element appears to be an aesthetic element, it is part of what makes the interpretations musically and aesthetically attractive; it is part of their beauty. And the element of beauty is essentially connected to the performative aspect of the performance, its spontaneity or immediacy – or to put it the other way: when the spontaneity is essential to the musical performance, it is because it has to do with the aesthetic value of the performance. Since the reflective element contributed by Harnoncourt is part of the beauty of his performances, it must in some way be connected to the performative aspect of the performance, and not only to the performance in general, the doing, as primarily knowledge-based. Harnoncourt’s performances are most certainly knowledge-based, but undoubtedly more than pure demonstrations. However, I will suggest that the doing could also be understood differently, as we shall see.

Now, of course, one might ask whether this idea of a performance as a pure demonstration is at all helpful. Would it be possible to find such performances? Are there at all performers whose only goal is to demonstrate a certain knowledge? Would not rather any performance, good or less good, interesting or less interesting, always transcend the level of pure demonstration? When I use the concept of demonstration here, there are two reasons. First, it is used as an imagined extreme, as a so-called ideal type for heuristic purposes, in order to indicate a certain direction. I fully admit that it would be impossible, objectively to state that this or that interpretation in fact is a demonstration, and nothing more. However, and second, it is important to remember that my perspective is that of the listener. And as a listener, I may very well perceive a certain interpretation as nothing but a demonstration. The term “demonstration” is then used to indicate a certain quality (or lack of such) in some performance, in other words, it is used in order to characterize my aesthetic experience (or lack of such). The notion of demonstration is thus used quite independently from the question of whether the actual performance de facto is such a demonstration or not. Alternatively; it is impossible to say that an interpretation is nothing but a demonstration. However, it may well sound demonstrative.

Let us now move to the question of the possibility of a connection between some notion of reflection and the performative aspect of performance. I would suggest three (combinable) elements here: First, I will stress the perspective of the listener and the concept of aesthetic experience. If we maintain that the relationship between reflection and the performative aspect of performance is contradictory, the performer cannot follow the track all the way to the performative aspect of his performance, and especially not so by use of his reflective capacity. However, our perspective is the listener’s, irrespective of what the performer is actually doing, seen from his, the performer’s, perspective. This does not mean that the reflection is simply passed over to the listener. Rather, it is connected to the performative aspect, but as a characteristic which appears to the listener, in his or her experience.

The second element focuses on the question of whether the reflection in the performative is actually the same as, or even identical with, the knowledge-based reflection preceding the performance. According to what I have said already, it is not – or at least not in a simple way. However, to distinguish between these two notions of reflection by the use of different terms would be unsatisfying, because then the tension would get lost. Instead it seems more convenient to think of some kind of metaphorical transition, in other words that the term “reflection” connected to the performative aspect should function as a metaphor, more specifically: as a metaphor for some kind of resistance or friction. This resistance would then refer partly to the resistance in the knowledge-based reflection, partly to the resistance in the very use of the concept of reflection connected to the performative aspect of performance.

The third and last element again concerns the notion of the interpretative choice. As discussed above, the interpretative choice might be considered to be the actualization of a preceding reflection. And following this, it makes sense to combine the concepts of knowing and doing. However, it seemed unsatisfactory to understand the interpretative choice purely as a realization: The interpretative choice is also something that is done, that is performed. In other words, the element of doing – the very doing of the performer: his actual articulation, pronunciation or even “spelling” of the musical gestures, figures, ideas – might also be connected, in some way, to the happening. Keeping the listener’s perspective in mind, one might say that what the performer does, happens to the listener. The very doing of the interpretative choice could be thought of both as an actualization of a preceding reflection, and as a doing in its own right, a careful articulation that, to the listener, is inevitably connected to the performative aspect of the performance.

In all of the foregoing, there is a challenge when it comes to locating the reflection – in the performer, in me, or somehow in the actual performance itself. However, I am not sure whether there is a clear answer, and I am tempted to argue that it is in the very nature of the complex structure of the aesthetical experience that what seems to be a feature of the object turns out to be something that takes place in me… at least as analytically understood. What makes it tricky is that the feature in question is the notion of reflection; in other words something that we are inclined to connect to some agent (i.e. the person doing the reflecting). Here, however, the agent is ambiguous (either the performer/interpreter or me as a listener) or may even be more abstract than that – an entity that somehow exists in the interstices of the performance, whether on the generative or receptive side, or both. In my view this uncertainty regarding the location of the reflection, far from being unsatisfactory, is integral to the core argument of the article. This is that interpretative choice seems to capture a doing that not only belongs to the performer but also takes the shape of an aesthetic phenomenon experienced by the listener. And it is exactly this thrilling transference – one that, in this context, is articulated as far as possible in the tension between the concepts of reflection and performativity – which seems to me to be a specific characteristic of Harnoncourt’s art of interpretation.

This article is a reworked version based on a paper given at PSN-conference, NMH, Oslo July 5th 2018.


Abbate, C, ‘Music – Drastic or Gnostic?’ in Critical Inquiry, 30(3), 2004, pp. 505–536

Fischer-Lichte, E., Ästhetik des Performativen, vol. 2373, (Frankfurt am Main: Suhrkamp Verlag, 2004)

Fürstauer, J., ed., Nikolaus Harnoncourt: Mozart-Dialoge: Gedanken zur Gegenwart der Musik, (St. Pölten – Salzburg: Residenz Verlag, 2005)

Fürstauer, J., ed., Nikolaus Harnoncourt: “Töne sind höhere Worte”, Gespräche über romantische Musik, (St. Pölten – Salzburg: Residenz Verlag, 2007)

Fürstauer, J., ed., Nikolaus Harnoncourt: »… es ging immer um Musik«: Eine Rückschau in Gesprächen, (St. Pölten – Salzburg – Wien: Residenz Verlag, 2014)

Goehr, L., The Quest for Voice: Music, Politics, and the Limits of Philosophy, (Oxford: Oxford University Press, 1998)

Gülke, P., »Triumph der neuen Tonkunst« Mozarts späte Sinfonien und ihr Umfeld, (Kassel / Weimar: Bärenreiter / Metzler, 1998)

Harnoncourt, N., Der musikalische Dialog: Gedanken zu Monteverdi, Bach und Mozart, (Kassel: Bärenreiter-Verlag Karl Vötterle, 2001)

Harnoncourt, N., Musik als Klangrede: Wege zu einem neuen Musikverständnis, Essays und Vorträge, (Salzburg und Wien: Residenz Verlag, 2004)

Schering, A., ‘Zur Sinndeutung der 4. und 5. Symphonie von Beethoven’, in Zeitschrift für Musikwissenschaft, 2, 1934, pp. 65-83


[1]     I am aware of the possible terminological confusion regarding “interpretation” and/or “performance”. This article takes as its point of departure that, within the musical domain, any performance necessarily presuppose, and in fact is, an interpretation. However, without dealing explicitly with the difference on a terminological level, one could say that the very nuance touches the core of what is at stake in the argument of this article.

[2]    His most important and influential articles are collected in the two volumes Musik als Klangrede (Harnoncourt, 2004) and Der musikalische Dialog (Harnoncourt, 2001), both existing also in translations into English and French. Further volumes of interviews are Mozart-Dialoge (Fürstauer, 2005), “Töne sind höhere Worte” (Fürstauer, 2007) and “…es gin gimmer um Musik” (Fürstauer, 2014), in German only.

[3]   Mozart’s Instrumental Oratorio: The Last Symphonies. Concentus Musicus Wien, Nikolaus Harnoncourt, Sony Classical, 2014, Catalogue No: 88843026352.

[4]    The discussion of whether the three last symphonies of Mozart (KV 453, 550 and 551) constitute a whole goes back to Mozart’s own time. For a thorough discussion on the issue, see Peter Gülke’s book “Triumph der neuen Tonkunst” (Gülke, 1998), an author to whom Harnoncourt himself refers.

[5]     Beethoven, Symphonies Nos. 4 and 5. Concentus Musicus Wien, Nikolaus Harnoncourt, Sony Classical, 2016, Catalogue No: 88875136452.

[6]    See (Schering, 1934, p. 79). In fact, I was personally present at the rehearsals preceding this recording, and although Harnoncourt did not refer directly to Schering in rehearsing the 5th symphony (as far as I can remember), he most certainly did so in rehearsing the 4th. Moreover, Harnoncourt also explicitly refers to Schering’s content-oriented interpretations (inhaltliche Deutungen) in an interview about Beethoven’s non-musical sources of inspiration (Beethovens außermusikalische Inspirationsquellen, (Fürstauer, 2007, p. 137)).

[7]   Schubert, Complete Symphonies, Masses Nos. 5 & 6, Alfonso und Estrella, Berliner Philharmoniker, Nikolaus Harnoncourt, Berliner Philharmoniker Recordings, 2015, Catalogue No: BPHR150061.

[8] Here it would also be worth mentioning the article “Music – Drastic or Gnostic?” by Carolyn Abbate in which she, in a deep reflection on the nature of performance, asks “to reflect, must one in some sense depart? Split a drastic self from a gnostic self?” (Abbate, 2004, p. 511).

DOI: 10.32063/0502