Bow Trajectory In, From and To The Air

The bow’s “air time” – its silent choreography in the air – is as vital to our bowing technique as silence is vital to music. However, because the bow’s movements in the air make no noise, this subject tends to be neglected in cello teaching. There are no instructions about aerial bow choreograpy written into our music parts, which is one of the reasons why learner string-players can get literally quite stuck on the string (see “The Retake“).

The bow’s aerial dance is not only an important element of bow technique, but is also an important element of emotional and musical communication. When we listen to a recording, these movements are not only inaudible but also invisible, and are thus only important to the player(s). But when we are watching a performance, or playing music with others, the bow’s aerial choreography takes on a whole new world of importance.

The cello is one of the few instruments that absolutely must be played from a sitting position. While playing, the only limb of the cellist’s body that is normally absolutely free to move is our right arm (and bow). When in contact with the string, the bow is our voice and our principal means of musical expression. But when the bow is freed from its contact with the string, even though it is no longer making any noise, it continues being a powerful means of expression. In the air, thanks to its now hugely amplified range of movement, it becomes a potent element of body language and remains still one of our greatest means of communication with both audience and other musicians, abandoning now its role as a voice, but adopting instead new and important silent roles as both dancer (accompanying the music) and conductor (leading the music).

In this article we will not only talk about the bows silent, aerial choreography but also about how we cross that delicate, dangerous frontier between the air and the string – in other words, about how we place the bow on the string (from the air) at the beginning of the bow stroke and how we take it off the string at the end of the bow stroke. We will not however talk here about Spiccato or Ricochet. This is because, in these bowings, we don’t actually “place” the bow on the string (except perhaps for the first note) nor do we “take it off” the string. Instead, we “throw” the bow at the string in order to produce the bounce, and once it is bouncing we just allow this bounce to naturally make and break the contact with the string.


In music, as in life, the most interesting (but also the most fragile, dangerous and unstable) areas are the transition zones where different worlds meet. An example from the natural world could be the seashore, where the ocean meets the land and the breaking waves reflect the border agitation. Sometimes the transition between the two worlds is dramatic and abrupt, as with a cliff or a mountain rising sharply out of a wild ocean (sfz). Other times the transition is almost imperceptible, as with a gently sloping sandy beach (al niente, da niente). Political frontier zones between different countries and cultures show the same characteristics: across some borders the cultures merge imperceptibly while at others the change is sudden and radical.

The string player’s bow also has its own delicate, agitated, dangerous frontier zone that it has to learn to safely cross. This frontier is not however between land and water nor between beliefs or cultures, but rather between the air and the string.


Both planes and cello bows divide their time between the air and the land so let’s compare string playing with aviation to see if we can find any more interesting comparisons. Firstly let’s look at a major difference: bows, unlike airplanes, do their “job” while they are on the ground (on the string) and have their “time off” while they are in the air. Secondly, an important similarity: most airplane crashes happen at landing or takeoff.

For the bow also, the moment of making contact (or of breaking that contact) with the string, is a critical time. But whereas a pilot does only one landing and take off in each “job”, we cellists, by comparison, do hundreds in each piece – probably thousands in a long opera – so we are obliged to become expert “bow pilots” for these delicate manoeuvres!! Another difference between pilots and cellists is that, whereas the pilot’s main job is to avoid crashes, the cellist does not have safety as the absolute overriding objective: our “job” is to paint a wonderful picture (tell a story) with our plane (bow) as if it were a paint brush. Because music portrays every aspect of life, violent crashes are as much an essential part of our repertoire as are perfectly imperceptible landings and takeoffs.


Unlike landing a plane on a runway (touching it to the earth), the way we touch any living or inanimate object, communicates, in a way more powerful than words could ever do, a whole universe of emotions. A violent punch (sfz), a firm deep massage stroke (forte sostenuto), a tickle (light spiccato), a gentle nudge (portato), a tender caress (pianissimo dolce), a delicate sensual brush with the fingertips (flautando) etc: every emotion ever felt by living creatures can be transformed into (and transmitted by) touch. Touch can transmit infinite delicacy, absolute brutality …… and everything in between. Playing music – especially string instruments – and especially bowing – IS touch.


Touch is a powerful communicator, but what is important for our bow landings and takeoffs is that a very large part of the emotion, feeling, character transmitted by touch, is actually determined by the way in which we start and finish each contact. In other words, the difference between a punch, a slap, a massage, a stroke, a caress and a tickle etc is not only determined by its duration and force but also, and very much so, by the way in which we start and finish each respective “stroke”.

This becomes immediately obvious when somebody – anybody – touches us: the way they start and finish the touch (stroke, contact) tells us more about how they feel about themselves (and how they feel about us) than the stroke itself does. This of course has its exact parallels in the many different characters we can give to a note according to how we bring the bow into contact with the string (and the way we remove it from the string), in other words, how we start and finish the notes. While “touch” for a pianist refers to the way they bring their fingers to and away from the piano keys, “touch” for a string-player refers to the way we establish and break our bow’s contact with the string.


Violinists and violists are the luckiest string-players because their instrument is attached only to them (and not to the floor). This allows them to accompany the placement of the bow onto the string with a cushioning, downward movement of the instrument, and to accompany the removal of the bow from the string with an accompanying movement of the violin (which can be either up towards the bow or down away from the bow). These are a beautiful, dance-like gestures, in which all the participants (bow, musician and instrument) are engaged in free, interactive and harmonious movement. But these gestures are not only expressive and beautiful, they are also also extremely useful and practical ergonomic techniques.

Unfortunately, we cellists cannot imitate these movements because of the fact that the cello is attached to the floor and thus has no possibility to move in such a way as to accompany (cushion) the placement of the bow on the strings (or its removal). This means that our skill in placing the bow on the string from the air and removing it again (in the infinite variety of ways that the music demands), depends almost exclusively on the sensitivity and control of our bowing arm and hand, with virtually no possibility of help from the rest of the body or the instrument.

We can help ourselves in this delicate task by holding the cello more vertically, as is usually the case in Baroque-style playing. In this posture, the weight of the bow does not fall so directly onto the strings, and we can thus control the bow’s landings and take-offs more easily (see Bow Starts).

Let’s now look in more detail at these placements and removals.


Placing the bow on the string at the tip is very different to placing it at the frog. As we go out towards the tip, we can gradually use more and more of the leverage motion (rotation of the hand) to place the bow on the string, whereas at the frog we cannot use this movement at all. At the frog, the movement that our hand needs to do in order to place the bow on the string is 100% perpendicular to the string, whereas at the tip we can use any combination we like between 100% rotation and 100% perpendicular movement. The leverage motion gives us very good fine control of the weight we apply after landing, which gives us a huge advantage for doing gentle, imperceptible bowstarts in the upper part of the bow. But at the same time, bow placements (landings) in the upper half have the drawback that we need quite a lot of time to control each landing because:

  • the distance between the hand and the part of the bow that is making contact with the string is so great
  • the leverage effect means that a tiny movement of the hand makes a big movement at the tip of the bow
  • the tendency of the bow to bounce in its middle area

Therefore, bow landings in the upper half will be used mainly in situations in which we want a gentle start, and have plenty of time to place the bow on the string.


The closer we are to the frog, the more we need to activate all our shock-absorbing equipment – the joints of the fingers, wrist and elbow – to acheive gentle landings and starts. The frog is perfectly adapted to sfz, loud violent beginnings, such as the opening of the Dvorak Cello Concerto, but is much less suited to pianissimo gentle starts/landings. We have to be able to do these though, because very often they are unavoidable, especially after pizzicato notes when we have not enough time to get out into the upper half, where we really would like to be.

We can help our fine-landing control by turning the bowhair so that less hair is in contact with the string. When we do this, the hair offers less resistant to the string and thus acts as a more gentle “shock-absorber” than when we use the full hair.



Do we wait till the bow has stopped, to remove it from the string? Normally not, because stopping the bow on the string kills the strings vibration and also makes our body come to a total stop. Allowing the note to ring on almost always gives a beautiful musical effect, and allowing our body to keep moving through the silence keeps the musical intention alive as well as being a beautiful visual effect.


Removing our bow from the string is not quite as simple as it sounds. Theoretically, the same two basic movements that we can use in bow landings can be used in reverse for our takeoffs:

  • the rotation (leverage) movement of the hand
  • a movement of the hand perpendicular to the string


In practice, the rotation movement of the hand is not only impractical but can even be positively harmful. Because the bow is in motion as we remove it, using hand rotation can make it hit the neighbouring strings. Only on the outside strings is the use of rotation a possibility: on the C-string we could rotate the hand clockwise without the risk of hitting a lower string (because there is none), whereas on the A-string we could do the opposite rotation (anti-clockwise) without the risk of hitting a higher string (because there is none).

Even on the outside strings however, the use of rotation seems artificial and unnecessary. In fact, rotation in general appears to be one of the worst possible ways to remove the bow from the string. Understanding why this is so will need more work, but it probably has to do with the fact that no anticipatory preparation movement nor follow-through is possible because it is a movement that only involves the hand and fingers. It seems to completely break the flow between hand, wrist, elbow and shoulder.


Let’s discard hand rotation now and do our removals exclusively with a perpendicular movement of the bow away from the string. For smooth, flowing removals, we need to involve as much of the arm as possible. Why? This might, at first sight, seem unnecessary because we only need to lift our bow off the string by a few millimeters, but it is this arm involvement that makes all the difference to our ease of movement and our control. This is a bit like a system of cogs and gears whereby a large movement of the rope is transformed into a small movement of the target object. This type of transmission of forces gives us excellent fine control, because it much easier to dosify and control a relaxed large movement than a tiny tight one. This means that we will usually try and take our bow off the string with a movement that initiates in the elbow, if not the shoulder, and in which the whole arm is involved, in both the preparatory and follow-through phases. For example, a gentle removal during an upbow in the middle of the bow is made by simultaneously lowering the elbow and flexing the wrist (which makes the wrist joint “stick up”).The moment of removal of the bow from the string is just the follow-through of the large movement initiated by the arm.

Bow removals on a crescendo are usually totally unproblematic as they seem to incorporate very naturally the essential preparation and follow-through phases, rather like a running jump, a tennis stroke or a golf swing. The most problematic removals are those that come on an upbow with a diminuendo a niente towards the frog. Here, if we are not careful to keep our whole arm moving, we can easily freeze up. Downbow diminuendo removals are not usually problematic, no doubt because the upbow tends towards increasing pressure whereas the downbow is the exact opposite.


Apart from our bow landings and removals at the beginning and end of a phrase/passage, the most commonly used bow aerial technique is the retake.


In the following example we will retake our bow (bring it back in the direction of the frog) during the rests (where the arrows are marked).


If we try this same example but this time without retaking the bow, then we will immediately feel, hear, see and understand why we need to do the retakes. If we don’t retake our bow, then the assymmetry of the rhythms pushes our bow out to the tip and creates large accents on the shorter notes as we try to fight our way back into the bow’s confort zone (in the middle). Apart from reatkaing the bow, the only other way to avoid this happening is by doing hooked bowings (see Dotted Rhythms and Choosing Bowings)

In the above example, Mozart was kind enough to write rests just at the moments in which when we need to retake our bow. Very often this is not the case, and we will need to “steal” the time for the retake by cutting the longer note (just before our retake) short. Fortunately our instrument continues to resonate even when we lift the bow off the string, so this cutting short of the longer note can often be made almost imperceptible. And in other casdes, the slight gap that may be produced is actually quite a musical effect – like a little breath – which breaks up what could otherwise be a legato sostenuto monotony.