Part 1 - Style


Evidence, Taste and "Romantic" Style

It hardly needs saying that these days, an unstylistic performance of baroque music leaves many of us listeners feeling a bit short-changed. On the other hand, a clean and rubato-free modern performance of, say, the Elgar String Quartet, is perfectly acceptable, and would rarely result in anything even close to a raised eyebrow. But it is every bit as historically 'inaccurate' as heavy early music, and perhaps even more so, as we have real recordings of Elgar and a much clearer knowledge of his performance tendencies. In practice, then, the idea of historical context works in all sorts of different ways. In the case of Elgar, historical context often means a general sense of the composer's emotional world and society, covering Edwardian culture and the Great War, but staying away from the more specific features of contemporary string playing. Contrast this with the primarily treatise-dependent approach to baroque music and it is clear that historical information about music seems to function differently in different repertoire. How this varies probably has a great deal to do with current ideologies and tastes.

Of course this doesn't pose too much of a problem, especially if you agree with the view that communication and expressiveness is the real heart of music. If done well, historical research into performance is more about curiosity than correctness, and aims to enliven rather than limit communication. But these sorts of paradoxes and inconsistencies make performers like us think quite carefully about how evidence actually functions in the modern world of classical music, and about the often unspoken assumptions that lie behind our own practices.

The Elgar example makes for a good starting point, as we have quite good evidence for his performance style. Yet the whole idea of evidence somehow seems less important to Elgar's music than it does to older repertories. One difference is in the format of the sources, which is important because different types of information seem to work in slightly different ways as evidence for performance practice. In the case of Elgar, our evidence includes recordings, which - and this is a bit of a broad generalisation - are quite often regarded as interesting, but dispensable. By contrast, (written) evidence of baroque performance practice is generally interesting and indispensable.

Perhaps one reason that the recordings are not vital is simply that they sound strange and unfamiliar, and so don't we tend to like them. This may be simplistic, but it actually leads to a rather more sophisticated point: sources whose expressivity we reject are likely to be treated differently to 'real' evidence. Although this is starting to change, old recordings can quite easily seem like nothing more than examples of poor taste - performance done badly. Real music-making leaves a space for artistic reactions, and this fact creates huge complications for the status of recordings as proper sources for historical performance. This is one of many sensible reasons why HIP has traditionally focused on written sources; treatises leave far less of this space for such reactions, and words offer just the sort of confident and prescriptive tone - in a word, authority - that performances cannot.

Not only that; Romantic playing is especially unpalatable because most modern ideologies of performance, with their moralistic emphasis on obligation to scores and to composers, are built on its rejection. Partly thanks to the tone of HIP's (early) rhetoric, practices such as sliding and extreme rubato can sometimes elicit responses in modern listeners that point to disingenuousness, or even arrogance, on the part of the performer. And so we are left with a potent combination: years of rhetorical training against these practices, allied with the fact that treating real performances involves engaging emotional as well as critical faculties. Clearly, taste has a significant role in determining the role and status of 'musical evidence'.

Of course, there is nothing wrong with taste; the point here is simply that there is a double standard at work. Take the confident proclamation of the Musical Times in 1849:

"How differently is the Messiah now performed - how much is it improved from the work which the Composer left!"

It would be a surprise to find anybody willing to make such a brazen claim about Handel today. But in essence, it is not all that different to modern approaches to Romantic performance - something along the lines of "we do it better now". Again, maybe this attitude is so rarely challenged because unfashionable style simply doesn't feel like the same thing as a historical style - it is objectionable, not historical.

The big question is: if we did have recordings from much earlier historical periods, to which we could have our own reactions, would they be treated as real sources to the same extent as treatises have been? Maybe they would, maybe they wouldn't; it is only hypothetical. But judging by the way old recordings have been used in the past, we might actually need to like what we hear, and this seems a bit unlikely.

It is yet another sign of the success of twentieth century rhetoric that Romantic style - which actually provides the best evidence we have of what real musicians did - seems by its very nature to 'get in the way' of the music. By contrast, 'historical' style is shown as providing the tools to 'reveal' it, despite tending to deal entirely in non-sounding sources. The irony of this situation is that by starting with earlier music and working forwards, 'Historically Informed Performance' was eventually going to embrace the very attitudes that it was meant to eradicate. Now this is happening, and early recordings are beginning to be treated as real evidence, rather than as unworthy distortions of an ideal past.

Making the leap and taking late Romantic musicianship seriously is incredibly exciting, because these recordings are full to bursting with practices that are as intriguing as they are baffling. And because the sources include real performances by real people, this evidence hits us in a visceral way that is just not possible with written description, no matter how vivid. We need to tread a little carefully, as early recordings are far from value-free, but they give us a captivating snapshot of an alien musical world.

Partly because of HIP, most listeners are very amenable to unfamiliar styles of music-making when there is a real historical link to a composer's authority. What we find is often surprising: take, for example, the recording of Elgar's Cello Concerto made by soloist Beatrice Harrison, conducted by the composer. A lazy, preformed opinion of old performance styles might be that they tend to be sluggish, sentimental, and heavy. This hardly applies to the accelerando passage in the slow movement (starting around 9:38), which reaches a faster tempo than almost all modern versions! (Here's the link:

And there are plenty of examples of when the modern idea of 'how the music goes' doesn't quite match how it has actually sounded in the past. This idea, and its revisionist implication, will be very familiar from the Introduction. The only real differences here with earlier HIP approaches are the repertoire and the types of sources available, which by the end of the nineteenth century obviously start to include recordings.

It was in this spirit that pianist Sigurd Slåttebrekk and producer Tony Harrison copied Edvard Grieg's 1903 recordings of his own piano music, a project that yielded some absolutely fascinating results ( One of the many insights to come out of their pioneering work is that Grieg's phrase structures - the written regularity of which is often seen as a compositional deficiency - actually sound far less predictable when the composer plays them. Instead, they are frequently blended together and elided, creating far more complex patterns than the ones on the printed page. Slåttebrekk and Harrison also showed that Grieg's performance style is actually very internally consistent, despite giving a strong impression of spontaneity and improvisation. His repertoire of performing gestures - things like rushing, lingering, eliding, voicing - present layers of complexity in the sound of the music that are simply not in the written text. This also suggests a fundamental problem in the relationship between convention and notation: conventions, almost by definition, are the sorts of things that don't need to be written down, because they are part of a living, performance-based culture: everybody knows what to do. For instance, before prescriptive attitudes to notation became commonplace, composers writing for strings had no need to notate portamento, because it was assumed they would do it anyway; the same goes for many other practices heard on early recordings.

On that subject, an excellent example for string playing is the Bohemian Quartet (later known as the Czech Quartet). The group was in existence from 1891 until 1934, and made several recordings in the 1920s, including of Dvorak's Op.96, Smetana's First, and a Quartet Op.12 by their own second violinist, Joseph Suk. They are clearly historically interesting, but at first hearing their playing sounds incredibly sloppy, lacking in any of the discipline we have come to expect from string quartets. Perhaps more than the previous examples, these performances do not come out favourably if judged by modern criteria. Listen to the Bohemians play the first movement of Dvorak's "American" Quartet Op.96:

It can be quite unsettling to encounter a quartet that is not especially worried about playing together, not to mention their sliding, rushing and irregular 'swing'. But this sort of performance has so much detail in it that, after the initial shock has subsided, their playing starts to become more and more compelling. As with Grieg's recordings, the musicians seem rather more interested in creating a constantly shifting and ambiguous surface than projecting structural clarity. Their bubbling rubato gives the impression that the music could turn in any direction at any moment, and they rarely telegraph exactly where a phrase might be heading. When they do, it has a special impact. (Slåttebrekk and Harrison call this sort of thing 'marked' and 'unmarked' elements of style, which are very useful categories for thinking about old performances.)

By the time they made their recordings, the quartet had existed for over 30 years, and this particular combination of personnel had been together for at least a decade. They were one of the most famous groups in the world, and it was performances of this Czech repertoire with which they forged their reputation. Given this fame, it is surely misguided to dismiss their performances as a poorly executed version of the style which we are used to; it is surely more likely that extreme flexibility and unsynchronised ensemble are fundamental parts of their expressive vocabulary. It obviously isn't possible to say with any certainty to what extent these features were worked out in rehearsal, but it is interesting that a similar technique, known as dislocation, was also a prominent feature of contemporary piano playing. The Bohemian Quartet was a really fascinating ensemble, and the next chapter examines them (and quartet playing) in much more detail.

It might seem obvious that society's perception of what is 'musical' changes over time. All listeners to classical music are probably aware of it on some level. It is quite easy to forget, though, because it happens over relatively long timespans, and early recordings do the important job of reminding us just how significant the changes can be. Expressive devices that were second nature to musicians in the past have been replaced by entirely different ones, which are gradually integrated into listeners' expectations and performers' aims. Over time, musicianship once regarded as the height of good taste can come to sound careless, uncontrolled and fundamentally unmusical. How these changes happen and what drives them are the focus of a lot of current research, and there are certainly no simple explanations. But one other important aspect of working with early recordings has been to show that a historical approach to musical performance has created some extremely complicated tensions between the conventions of the present and those of the past.

So far, these examples have been chosen because they represent historically relevant information - the Bohemians knew Dvorak, and both Grieg and Elgar were performing their own music. The sounds themselves are unsettling and intriguing in subtly different ways, offering all sorts of opportunities and curiosities. Now we're going to examine in a bit more detail at what old recordings can tell us about some of the most deeply embedded of modern conventions: playing together.



© Chris Terepin 2015