Watching different fine cellists play, we might be surprised to see that there are such an extraordinary variety of ways to hold the bow and still play very well ……. and sometimes it even seems that the very best cellists have the strangest bowholds (Yo Yo Ma for example)! We can’t fail also to observe that these fine cellists don’t normally have only “one” bowhold but rather that they will often vary their bowhold,  according to the musical and technical demands of the passages being played. In fact, even just going from one end of the bow to the other can produce a significant change in bowhold.

When playing passages in the upper half of the bow and on the string (with no part of the trajectory in the air), we can basically use any bowhold because the weight of the bow is supported by the string. In this situation, our hand is really only guiding the bow and applying pressure to it rather than actually “holding” it. In contrast to this, there are two situations in which we definitely need a good bowhold:

  1. when playing at the frog, and
  2. when the bow is in the air.

In both of these two situations we need that good bowhold because here we have to control (hold) the entire weight of the bow and also manage the complex leverage effect (which is actually much more demanding than simply carrying the weight of the bow  (see at the bottom of the page).

To summarize: as with so many aspects of cello playing (and life itself!), we want maximum flexibility, without however losing strength or control. Possibly the best way to find the best mix of these contradictory elements is through “play” in its literal infantile sense: in other words, simply “fooling around”, experimenting with the bow and the many different ways to hold it in different musical contexts. The most useful “play” will be at the frog, with many placements from the air (see Bow Trajectory In, To, and From the Air)

For me personally, the most helpful ideas that I have come across with respect to holding the bow are:


Holding the bow with our fingertips, as though it were a pencil or an artist’s paintbrush may seem like a good, intuitive way to achieve delicacy and fine control. But a cello bow weighs approximately ten times more than the pencil and is many times longer, all of which means that the bow is simply too heavy, cumbersome and unwieldy for us to really be able to control it with such a delicate “pencil” hold. What’s more, we often need to apply much more pressure with the bow than we ever need to with a pencil or artists paintbrush. Even with a toothbrush, the pressure we need to apply is such that we tend not to hold it in our fingertips, and we would never imagine a house-painter using a fingertip hold on their big heavy paintbrush. Probably the same logic is behind the double-bass bowhold(s): the french, with their desire for lightness, delicacy, bounce and air use the overhand cello bowhold while the germanics, with their desire for weight, sostenuto and volume use the underhand bowhold, perfectly flexible and loose even in very loud, heavy situations.

Having the fingers tightly curled (the “claw” posture) might give us a tight firm grip on the bow, like that of a bird of prey holding onto its struggling prey, but the weight and length of the bow are such that we are obliged  -especially at the frog, for pianissimo playing, and when the bow is in the air – to grip it so tightly that our hand can become tense and stiff. This tension and stiffness has serious drawbacks for almost all of our different bow techniques in causing a loss of the hand flexibility that is so absolutely essential not only for our bowchanges but also for our string crossings and our control of the bounce, our bowstarts and our placements from the air. Letting (holding) the bow deeper into the hand (the “paw”) rather than holding it with our fingertips (the “claw”) may seem counterintuitive but, by allowing the hand and fingers to relax, will probably give us ultimately more flexibility and control.

The difference between the “pawhold” and the “clawhold” is not just related to the depth at which the bow is held in the hand but also to the degree of separation between the fingers. With the fingers closer together (the “paw” posture), the hand is liberated from a lot of tension and becomes much more flexible. With the fingers spread out, the hand becomes more stiff and rigid. If however, we find that we are lacking control when playing at the frog, this might indicate that we need to open the fingers out a little, in order to be able to securely support (hold) and control the full weight of the bow.


It is generally considered that having the last joint of the thumb bent helps enormously, because the “bend” acts as a shock absorber, giving flexibility. Jeffrey Solow has another idea, undoubtedly very valid also. He believes that many of the greatest cellists historically, including Rostropovitch and Emmanuel Feurmann, played with a straight thumb, wedged into or alongside the frog. Watch Jeffrey talking about this here.

Possibly, the length of the thumb in relation to the fingers might have something to do with which posture each cellist uses. Certainly, if our thumb is bent, its tip pushes with considerable force against the point of contact with the frog/bowstick. The exact location of this point of contact between thumb and bow is a subject that merits some study. Is this point usually at the left end of the frog, where the frog meets the bowstick? In any case, with a bent thumb, all the pressure of our thumb on the bow is applied via the relatively small surface area of the thumb’s tip. It takes time for the end of the thumb to get used to being jammed permanently into the bow in this way and it will certainly need to develop a thick tough skin callous on the area of contact. Some cellists (Tortelier) put medical rubber tubing over that part of the bow to avoid this problem (and also to give a more secure hand/bow contact).

No doubt with a straight thumb, the pressure is transmitted more via the thumb’s pad, which is a larger, softer, “meatier” surface than the tip. For this reason, perhaps the pad is better suited to give sensitivity and flexibility in the bowhold.


If we hook the first finger around the stick of the bow then we can easily tend to “the claw” posture (not recommended). It may be preferable – especially in legato passages – to use the first finger principally as a “modulator of pressure” rather than as an element that actually “holds” the bow. When the bow is in the air however, the first finger does need to contribute to the actual “holding” of the bow.



Let’s start by finding the centre of gravity of the bow. When the bow is perfectly balanced on a finger inserted between the hair and the stick, that finger is at the bow’s “balance point, or in other words, at its centre of gravity. Because the frog is the heaviest part of the bow, the bow’s centre of gravity is much closer to the frog than to the middle of the bow.

If we were to hold our bow exactly at this balance point, the job of our bowhold would be vastly simpler, especially when the bow is in the air. This is because, when holding it at the centre of gravity, we only need to control the bow’s weight without having to manage any of the destabilising forces of “leverage” that are created by holding the bow at the frog (relatively far from its balance point). Just try it, to see how light the bow feels when we hold it at its balance point.

These forces of leverage are proportional (exponentially?) to the distance away from the centre of gravity at which we hold the bow. When chopping with an axe or hitting with a hammer, we use the same principle of holding them far from their centre of gravity in order to multiply our power. For the bow, this leveraging power can also be very useful, especially in high-intensity Romantic and 20th Century music, but it does (of course) have its downside. It not only requires much greater muscular effort but also greatly increases the risks, multiplying the difficulties of control – especially of fine control. It’s much easier to hit the nail cleanly on its head when we hold the hammer closer to its centre of gravity ……… but we have much less force in our hit!

We can find a beautiful analogy for this phenomenon in the world of economics and finance, where the term “leverage” refers to the possibility to make huge amounts of money from small initial outlays. Yes, the power is there and some people get very rich very fast, but the risks are enormous and many just get wiped out at the first sneeze. Leverage (great power with only small efforts) is not only a double-edged sword but is also a phenomenon that modern society has developed to an extraordinary degree. Look at weapons: in ancient history, they used rocks and catapults, then we invented guns, then simple explosives, and now with the click of a button (or a mouse) the whole world can be blown up. Look also at transport: first, there was walking, then the wheel, then bicycles, then cars, and now with a tiny pressure on a pedal or lever we can get up to mind-bending speeds in the space of a few seconds. That is “leverage”.


In fact, the presence or not of this “leverage” effect is the source of one of the major differences between Pre-Romantic (without leverage) and Post-Romantic (with great leverage) bowing. The Tourte bow (invented around 1790) is a “leveraged” bow, with the bowhold displaced much further away from the centre of gravity than the Baroque bow. What’s more, the spike, which allowed the cello to be held more horizontally, was invented around 1830, not long after the Tourte bow. The spike is also an “increased leverage” tool because it allows us to hold the cello more horizontally. This is significant because the more horizontal the cello’s position is, the greater the effect gravity will have on the bow, and thus the greater is the effect of leverage.  ….. and thus the more complicated becomes the art of holding and controlling the bow!

The simple fact of moving our bowhold “up” the bow (away from the frog = in the direction of the balance point) gives our playing an immediate lightness and “Baroque touch” because it removes the leverage effect. If we simultaneously place the cello more vertically and sit forward on our chair, the leverage effect is even more diminished (see Baroque Style and Interpretation and Classical Period: Style and Interpretation)