One of the biggest challenges for any performer of jazz, at least as the majority of jazz players I know would say, has to do with rhythm, and specifically the performance of uneven rhythms so as to ‘swing’. Some have tried to notate these, sometimes very approximately, sometimes in much more detail. But the consensus appears to be that codifying such a rhythm then simply executing according to the ‘rules’ thus generated will sound artificial and contrived. Furthermore, it is difficult via such an approach to be flexible in one’s ‘swinging’, as the nature of this can be subject to small variations depending upon the musical moment, or to respond to the particular variants heard from other players in a group setting. So in general, there is thought to be no real substitute for simply playing with others on a regular basis, and absorbing the rhythm through listening, imitation, osmosis.
A parallel situation applies to language learning. While it may be possible to learn a good deal of vocabulary and grammar, even some idiomatic usage, through independent study, from textbooks, etc., many believe there is no real substitute from being immersed amongst native speakers, having to communicate on a regular basis without necessarily having recourse to ‘guides’ when immediate responses are needed. This is not least to do with the process of learning a decent accent (something generally thought rarely to be possible to such an extent as one could pass as a native speaker to other native speakers, unless one learns from a young age), but also absorbing a wide range of idiomatic employment in the form of speech (writing is a different matter).
It would be rash to rule out the possibility that there could ever be found means of learning jazz performance, or languages, in such a way that obviates the necessity for such regular interaction amongst those who have absorbed the idioms. But I am not aware of such means having yet been discovered and comprehensively tested.
This brings me to the question of empirical musicology (by which I mean specifically that using empirical means of measurement of aural data, by software such as Sonic Visualiser or other alternatives) for the analysis of musical performance, something which has been on my mind during a (generally excellent and very stimulating) conference on Performance Studies which I am currently attending (today is the last day). Empirical Musicology is a relatively recent development, about which an important edited volume was published in 2004, edited by Nicholas Cook and Eric Clarke, and for which there is a dedicated academic journal. With roots in the use of quantitative approaches employed in music psychology, and some others relating to such things as bodily motion during performance, a range of musicologists have increasingly used this to analyse various parameters of musical performance, especially tempo and its modification, rhythm, also pitch, timbre (notoriously difficult to analyse using more qualitative means), dynamics and so on.
The value of musical analysis (in the broadest sense) is not uncontested; it is an activity developed within the realms of academia which has not necessarily played a significant role (or a role at all) in the actual work of a great many practising musicians past and present. Cook himself, in his book Beyond the Score: Music as Performance (about which I published an extended review-article which also considers more widely the field of performance studies) was sharply critical of what he described as ‘Analytically-Informed Performance’ (AIP) whereby performers shape their interpretations according to priorities determined by analysts. Focusing on some key writings by music analysts Wallace Berry and Eugene Narmour, Cook suggests this implies a hegemonic imposition of academic ideologies on the world of music-making, not least because the dialogue appears one-way; he believes analysts have as much to learn from listening to performers as vice versa. So far, I can agree that a more two-way approach would be fruitful, but here and elsewhere Cook expresses major scepticism about the possibility that performers could learn anything of value from analysis and analysts, questioning the value at all of a musicological discipline which he has also linked to creating artificial hierarchies of value.
Here I disagree. I have my own scepticism about the value of certain highly systematised modes of analysis, such as grew up especially in the early post-1945 era in the United States, which has been plausibly argued to have been driven both by a positivistic, scientistic academic climate in which research in the arts and humanities was valued largely to the extent it could on the surface mimic the trappings of a ‘science’, but also by the growth of mass education and the concomitant need to find approaches to analysis which could be taught on an industrial scale and learned almost by rote. Much of the analytical work I have read from continental European musicologists has been no less rigorous or thorough, but more ad hoc in nature, freely adapting a variety of tools (or creating new ones) according to needs of the music at hand, and as such avoiding the unhappy situation by which the key criterion appears to be whether the music serves to bolster the theory, rather than vice versa.
But if one embraces a broader conception of analysis, to encompass most approaches to understanding the inner workings of a musical composition (or a performance or other sonic phenomenon), then I think not only can this be invaluable for performers (and listeners) possessing any basic curiosity, but also actually something that many performers do as a matter of course to some extent, as I argued in a keynote lecture entitled ‘In Defence of Analytically-Informed Performance’, delivered in a conference in São Paulo in 2019, which I intend to revise for publication at some point soon.
It is very far from unusual, for example, for a performer to gauge a series of mounting dynamic curves in a series of phrases leading to a mini-climax according to the melodic, harmonic and rhythmic properties of the information contained in the score. This may be done in a relatively intuitive and non-systematic manner, but nonetheless does constitute a form of analysis, a discernment of aspects of the score which reveal something about the musical processes, which then inform aspects of the realisation of the score in performance. Whether this perspective is generated simply by playing the music to oneself and listening, or some more studied process of identifying chord progressions, melodic properties, etc., does not make it any more or less ‘analysis’, in my view.
I was doing this myself (essentially intuitively) when practising the opening passage of the last movement of Schumann’s Fantasy in C, op. 17, yesterday, trying out different ways of shaping the two-bar phrases in bars 5-14, in terms of relative dynamic peaks, my choices informed by such factors as the extent to which the music modulates away from the home key before returning to a dominant harmony on the relative minor in bar 10, or the transformation of the bass line from a descending progression starting in whole tones towards a chromatic one in bars 11-14. Much is at stake in these bars in terms of the expressive and emotional trajectory prior to a return to a dominant pedal point in the home key in bar 15. Similarly, the shift to the major submediant in bar 2 is striking in the context of the wider style. Even after having played this work for over 30 years and having heard many others perform it, it still makes an impact, and so I wonder about such questions as whether to register this sentiment by a small holding back of the pulse and tenuto at the beginning of bar 2.
Even if I am describing aspects of the score above using very basic technical terms, I do not believe that I am in any sense unusual in grappling with such interpretive questions; the use of that language is simply a way to articulate some information discerned essentially intuitively so as to be comprehensible to one reading this piece of writing.
Empirical musicology is on another level in terms of the technical language and illustrations employed. Over the course of the conference, I have been looking at highly detailed charts, multi-coloured shapes (containing many hues) to illustrate timbral properties, animated graphics to trace the variation in basic pulse during a performance. Much of this yields interesting insights and conclusions, some of which appear to be established on a more secure basis than might have been possible otherwise.
But then I do ask the basic question: what purpose does this research serve? In one case, the researcher suggested that they were attempting to codify the rhythmic practices found to be common in a particular regionality when performing some music identified with that regionality, so that those performing such works in more remote locations could do a type of ’empirical musicology in reverse’ (my term, not theirs) and translate these back into performance.
But this is where I become more sceptical, and wonder if the results of such a process would ever really convince, for the same reasons as those outlined above with respect to jazz or language learning. How much music-making can really be reduced to a finite set of stylistic principles which can be codified and reproduced so as to create plausible imitation of the ‘real thing’? I would need to hear a range of realisations of this model to judge that, but do not have a huge amount of faith. Musical style is not, I believe, something so easily reified, but a flexible and continuously developing thing, driven as much (in repertoire involving more than one musician) by regular interactions between players as by collective adherence to a set of rules.
The context of this had to do with global music-making, in particular performance of a certain repertoire at localities a long distance from the music’s point of origin, also the place where the most regular and distinctive performance tradition has occurred over an extended period. But one possible conclusion from the above – that it is no more possible to perform in a relatively ‘unaccented’ style without regular interactions with the music’s ‘native speakers’ than is the case for speaking a language – may be unacceptable in terms of global aspirations. I should emphasise here that the analogy of a ‘native speaker’ does not really have anything to do with particular upbringing, nationality and certainly not ethnicity; it is about whether one has had the opportunities over an extended period for acculturation through musical interaction with others well-versed in the style. This ought in theory to be perfectly possible amongst diaspora communities; a group of Georgian folk singers who relocate in Brazil could in some sense continue a tradition of such singing in the new location, and incorporate into it others not originally from Georgia (whether other social and political factors might influence the extent to which this actually happens is of course a whole other set of questions).
Empirical musicology applied to performance attempts to provide an alternative to this, a means of learning a style through applying a set of principles obtained through systematic empirical analysis of performance. Those employing those are of course also free to listen to the original performances too, but what they may not be able to do is play together with other musicians versed in it.
Is this a viable alternative? I am not sure, and would need much more evidence to be convinced. But there is another possibility to consider, related to the considerations of the post-1945 United States academic environment I described above. Whether or not this research has a wider ‘impact’ (to use a buzzword familiar to all those in UK academia), could such empirical musicology be notable primarily as a means of trying to turn a lot of what musicians do anyhow, without necessarily having any input from academics, so as to appear like a type of quasi-scientific ‘research’ which is thought to be the best way to secure the status and funding of a particular academic sub-disciplines?