Understanding the Clavichord

The following talk was given by Peter Bavington at St John's Church, Keswick, on 8 March 2003 as part of the British Clavichord Society's Educational Initiative 'Clavichords in Cumbria'.

I suppose that in the first forty years of my life I heard the clavichord twice. The first occasion was at Fenton House, that wonderful National Trust property in North London which houses a fine collection of antique keyboard instruments, most of them in playing order. In those days - it is very different now - supervision was lax, and visitors, even teenagers like me, were allowed to wander round playing the instruments. I can remember reaching out with one finger and drawing a sound from the Christoffer Rackwitz clavichord (an instrument, incidentally, which is now in the Russell Collection, Edinburgh, but which was at that time on loan to Fenton House).

That was a single note - quite an exciting sound, I remember, unlike anything I had heard before, but as an experience of the clavichord, not much to go on. The next occasion on which I heard the clavichord was when, still a teenager, I belonged to a kind of musical youth club which met in private homes on Sunday afternoons. A group of us were invited to a flat in Eaton Square - which, I should explain for those who don’t know London, is a wealthy district, too select even to appear on the Monopoly Board alongside Mayfair and Park Lane, yet on the fringes of bohemian Chelsea. Our hostess played very skilfully to us on her new large Thomas Goff clavichord. I was curious about its action, of course, like any mechanically minded boy, but musically what struck me this time was its extreme quietness: a silver whisper at the limit of audibility.

These two experiences neatly summarise the role of the clavichord in the middle of the last century: on the one hand a museum piece, primarily of historical interest; on the other, an exquisite secret pleasure enjoyed in private by a few cultivated and well-to-do individuals. On the general musical life of the time, it made virtually no impression at all. A long time was to pass before I heard it again.

I took no further interest in the clavichord for a while, even when I decided to take up harpsichord making as a career. After College, I worked as assistant to the harpsichord maker John Rawson. One day, we took in a clavichord for some repair work which I undertook under John’s guidance. Having fixed the instrument, I tried to play it. Now, as a keyboard player, I was - am - a fumbling amateur, but I thought I could make a reasonable stab at a few pieces on the piano and harpsichord. I was surprised to find that I could not manage them on the clavichord at all. I had to start learning to play all over again from scratch.

In due course I realised certain basic things about playing the clavichord:

All these expressive skills, which the clavichord player gradually learns, are concerned with shaping individual melodic lines, rather than with passage-work or figuration; which is why the clavichord is, in my view, quintessentially the instrument of contrapuntal music. It is also, I think, the instrument of improvisation, or just messing around at the keyboard: and here, I admit, one of its advantages is that your fumblings are not transmitted to the whole neighbourhood.

There is something both pleasing and challenging about finding all these possibilities in an instrument so apparently simple. Somewhat to my surprise, once I had begun to learn its special technique, I found it hard to leave the instrument alone, and I was sorry to see it go.

The next eye-openers were the Bruges exhibition in 1992 and the first Magnano clavichord conference in 1993: here I encountered some really good instruments and realised what the clavichord was capable of. Some of these clavichords had a remarkably bold sound: one completely forgot that the instrument was supposed to be quiet.

Now, some might be inclined to say that loudness is not the point of a clavichord, and I would agree that an attractive ‘voice’ is also important, and of course in absolute volume of sound it cannot compete with other instruments. Nonetheless, in my view loudness, or perhaps I should say responsiveness, is definitely a virtue. On an unresponsive instrument it is hard work for the player simply to produce an audible sound, and subtleties tend to go by the board. A responsive one not only is more enjoyable to play, but also makes it possible to play more expressively.

Achieving this kind of responsiveness is the core of the maker’s task. I think it is basically a matter of energy conservation. The way in which the strings of a clavichord are made to vibrate, by striking them with a tangent which remains in contact and defines one end of the sounding length, is comparatively inefficient. Plucking, for example, as in a harpsichord, delivers far more energy to the strings for the same effort by the player. Now this means that in a clavichord the amount of energy in the strings is comparatively small. What the maker has to do is to make sure that as much as possible gets transmitted, one way or another, to the air as audible sound, and as little as possible of it is wasted. In practice this is an extremely complex task, and I shall consider some aspects of it in a moment, but first I want to mention one other thing about clavichords that I became conscious of as my understanding developed.

Some of them are fretted. What does this mean? The term comes from the bars across the neck of instruments like the guitar or lute. These are called frets, and their purpose is to enable the player to get several notes from a single string by stopping it at different points along its length. In the same way, on a fretted clavichord, more than one note may be obtained from the same pair of strings by making the tangents strike at different distances along from the bridge. To achieve this, the keylevers may have to be angled to the left or to the right, or as we say ‘cranked’.

You can tell at a glance whether a clavichord is fretted by looking at it from above. If it is a fretted instrument, you will see the cranked keylevers creating irregular gaps at the back; and if you imagine a line joining the tops of the tangents, it will be jagged and irregular, whereas on an unfretted clavichord it will be a straight line or a nice smooth curve. Usually the clavichord will be fretted for only part of its compass: fretting is generally impractical for the lowest notes, because the levers would have to be extremely cranked, and could not all fit in the space available.

One thing that fretting does is to reduce the total number of strings required. This makes for a smaller, and in particular, a narrower instrument. But it presents certain challenges (let’s not say difficulties) to the player.

Firstly, you cannot have two notes from the same fretted group sounding at the same time; and secondly, you must exercise care when passing from one note of the group to another, otherwise you make an ugly noise. So on a fretted instrument you must lift up your fingers and practise a clean touch. This is a good corrective to the kind of continuous legato which piano students are encouraged to learn, and a good training for the ‘standard’ touch used on all keyboard instruments in the seventeenth and eighteenth centuries, which we know was based on the principle that each note should be finished just before the next begins. Legato is a special effect: it is quite possible on a fretted instrument, with care, but what you cannot do is to overlap two notes when both are obtained from the same pair of strings.

Are there any advantages to fretting, apart from reducing the physical size of the instrument, and forcing the student to practise a hygienic touch? Well, the reduction in the total number of strings reduces the loading on the soundboard, which perhaps makes it easier for the maker to achieve a lively acoustic response. There is also a more subtle mechanical advantage: fretting makes for keylevers which are more equal in length over most of the compass, or at least, a more equal distance from keyfront to tangent. On an unfretted instrument, as you move down the keyboard from treble to bass, each tangent has to be a little nearer the front than the last. On a fretted instrument, because several tangents share the same pair of strings, and these strings are normally sloped from back left to front right, as you move down through a fretted group each successive tangent will actually be placed further back. This largely compensates for the need to move the tangent forward when you move on to the next pair of strings. The average line of the tangent tops will be more nearly parallel to the back wall of the case than on an unfretted instrument.

This makes possible a more even leverage and hence a more even touch, and it also makes the most of the space available. In an unfretted instrument there is unavoidably a quite large triangle at the rear left-hand corner which is basically dead, unused space: this triangle will be smaller in a fretted clavichord of similar compass.

Now, the clavichord seems to have been invented around 1400, and for the first 300 years or so of its existence it was always fretted. We now think that the first completely unfretted instruments were made in the late seventeenth century; nonetheless, fretted clavichords continued to be made and played alongside the unfretted ones right up until the instrument became temporarily obsolete in the mid-nineteenth century.

There was then a gap of fifty years or so before the instrument’s revival in the 1890’s. At the time of the revival, and for many years afterwards, fretted clavichords were almost completely disregarded. The fact that the player of a fretted instrument does not have quite the same freedom as the pianist or harpsichordist to play notes in any combination that the fingers can reach was seen as an irritating obstacle to self-expression. If fretted clavichords were made at all, they were definitely ‘second best’.

Around 1970 there seems to have been a change in attitude, neatly expressed by the title of an article by the late Edwin M. Ripin published in the Galpin Society Journal that year: ‘A Reassessment of the Fretted Clavichord’. Ripin identified a number of technical advantages of fretting (one of them being the geometrical one I referred to just now). Comparing fretted and unfretted clavichords of the 1780’s, he went so far as to say that the fretted instruments were [quote] ‘far more efficient, sensitive, and finer-toned musical instruments than their larger rivals’.

Actually I find that a rather sweeping statement, but it illustrates the enthusiasm with which some clavichord players and makers set about reviving the fretted instrument. As well as the perceived technical advantages, there was the curiosity factor, something which has repeatedly inspired the revival of Early Music, both repertoire and historical performance practice. People find they are no longer satisfied just with the question ‘What is it like for us?’, but also begin to ask ‘What was it like for them?’.

Nowadays both fretted and unfretted clavichords are made, but it is a fact that most makers seem to prefer the fretted instruments, perhaps because it is easier to endow them with a lively acoustic response. When I began making clavichords, I was, of course, keenly interested in how to achieve this lively response. Unlike an aspiring maker in the eighteenth century, I could not simply rely on an accumulated fund of wisdom passed down through the centuries from master to apprentice: that chain was broken in the 1800’s when the instrument ceased to be made. When it was revived, such traditional knowledge as there was came through the piano industry, and was in many respects misleading when applied to the clavichord.

What I wanted was to understand how the clavichord actually worked. Although it was re-born into a scientific and technologically sophisticated age, hardly anything has been written until very recently about its technical aspect. I think there are two reasons for this:

Someone who took a great interest in this was the late John Barnes, for many years assistant curator of the Russell Collection in Edinburgh, and one of the founder members of our Society. He had a life-long interest in all kinds of keyboard instrument, but in his later years he focused particularly on the clavichord. John was generous in sharing his ideas and experience with anyone who was seriously interested. It was my good fortune to have several long discussions with him, and continual help and advice when I began eventually to make experiments, rather amateurishly, in an attempt to understand the clavichord better.

Eventually I came to realise that the design of every part of the instrument was important, that apparently small changes could have quite big effects, and that everything was interconnected.

Take the keylevers, for example. They appear to be fairly straightforward first-order levers. The finger presses on the front end, the lever rotates about its pivot, and the tangent, somewhere near the rear end, rises to strike the strings. The speed of the tangent - the suddenness of the attack on the strings - is the vital factor in producing the note. It depends, of course, on the pressure applied by the player’s finger, but for any given finger-force it will be affected by several things:

All of these factors have to be correct for each individual keylever, in order to produce a reasonably good note. But it is also necessary to achieve some uniformity across the compass, despite the fact that each lever necessarily has its own unique shape and no two are exactly alike. Each key should feel fairly similar , when played, to its neighbours, otherwise the touch will seem irregular and the instrument will be hard to control.

That’s not the end of it. The keylever-and-tangent, as a unit, must have sufficient mass to sustain the vibration of the strings after they are struck. But the amount of mass required depends, to some extent, on the strings themselves: thicker wires will be at a higher tension, and will require somewhat heavier keylevers. But the choice of wire diameters needs to be in balance with the mass of the bridge and the thickness of the soundboard. It also needs to be appropriate to the sounding length and pitch. These last two factors, along with the nature of the string material itself, determine how high in the elastic range of the wires the working tension will be, which has a dramatic effect on the nature of the sound they produce.

You can see, I think, how interconnected all these factors are. And this is before we even start to consider the listing, which everyone agrees is vitally important, but which, so far, has defied any attempts at a systematic explanation.

I hope I have said enough to convince you that the clavichord, which seems so simple, is actually a very complex instrument. In my opinion the maker needs to understand these complexities, but the player - you will perhaps be relieved to hear this - the player need not be conscious of them. The aim of all the maker’s care is to produce an instrument that follows the player’s thought, that almost becomes one with him or her, like a second voice. (This does not, of course, mean that it is necessarily easy for a beginner to play).

Some - both musicians and makers - are suspicious of the kind of technical analysis that I have been advocating. All you need to do, they say, is faithfully copy the best old instruments: don’t put too much effort into trying to understand them.

My answer to this is firstly to agree that we should be extremely cautious before making any change to the designs we have inherited from the makers of the past. Just because the instrument is such a complex coupled system, a seemingly small change can have large and unpredictable effects. We need to be very sure indeed that any change we might think of making will not end up spoiling the instrument.

However, in my opinion, copying without understanding is doomed to failure, unless the maker is remarkably lucky. The problem is that the factors which make the difference are extremely subtle. It is quite difficult to make a precise copy of any object made of wood, but even if it is dimensionally precise, in sound and in playing characteristics it may be very far from anything the original maker would have recognised. This is confirmed by the fact that so-called ‘copies’ of the same instrument made by different modern makers sometimes sound completely different because of subtly different choices made during the construction and setting-up.

The modern maker, just like the maker of the past, cannot avoid using his or her own skill and judgment, and that judgment will be sounder if it is based on an understanding of how the instrument actually works.

return to top of page

back to home

updated 20 February 2009