To Play Parallel or Not: A Controversy of Technique, Part 2


Continuing the epic question: to play parallel or not?

Sad to say, there is not enough communication between musicians and scientists on string acoustics. These scientists can offer us an incredible amount of knowledge and understanding. All we have to do is ask. This relates to our primary goals here at String Visions – to inspire musicians (especially string players) to communicate more, to help them attain better answers from scientists, and to assist them in generating better questions that these scientists can research and provide answers to.

In the second article of this series, I would like to introduce a brilliant young scientist from Sweden to the readers of String Visions: Erwin Schoonderwaldt, a post-doctoral fellow at the Institute of Music Physiology and Musicians’ Medicine in Hannover Germany. He received his PhD from the Department of Speech, Music and Hearing at the Royal Institute of Technology in Stockholm, Sweden, and has completed a number of important studies on string acoustics including his PhD thesis “Mechanics and acoustics of violin bowing.”

In this article I would like to talk about and quote from Dr. Schoonderwaldt’s scientific paper “On the use of Skewness in Violin Bowing: Should the bow be straight or not?” [1]

I highly recommend that you all spend some quality time studying this wonderful paper so that you really understand the science behind this topic.

Before getting started I would like to tell an anecdote from my days as a student:

As a young cello student I was always taught to play parallel to the bridge. After returning from a concert featuring the French cellist Paul Tortelier, I said to my teacher, “The concert was fantastic; such a wonderful sound full of expression and colors…but Mr. Tortellier hardly ever played parallel to the bridge.”

My teacher responded, “Tortellier is so good that he can get away with it.”

His answer was very surprising to me and it wasn’t until I became a teacher myself that I fully realized why he had given me this answer. I think my teacher knew that advanced string players hardly ever play parallel to the bridge. He, however, wanted to stress the importance of what playing parallel to the bridge felt like so that one may truly develop the feeling for the vibrating string.

Here I will quote some of the important sections from Dr. Schoonderwaldt’s paper. (As stated above you must read and study the whole paper in order to really understand this topic).


Bowing parallel to the bridge is by many players considered as the golden standard. However, in practice straight bow strokes are rarely observed, and the bow can be considerably slanted even in the performance of renowned players. In any case, the angle of the bow with the violin (skewness) is likely to form an important control parameter, which has hardly been addressed in scientific studies of violin performance. In the current study measurements of skewness in violin and viola performance are presented, and possible explanations of the observed behavior are offered. The results provide strong indications that skewness fulfills an important function in controlling bow-bridge distance, integrated in players’ performance strategies.


During the last two decades, several studies of violin bowing in performance related to the control of the sound have been conducted using sensor systems, motion capture techniques or combinations of those. In these studies the focus has been on the main bowing parameters at the bow-string contact with a direct influence on the string vibrations, in particular bow velocity and acceleration, relative bow-bridge distance and bow force. These bowing parameters are well established in bowed-string theory and simulations dealing with playability….

As yet, scientific studies of the use of the bow angles are scare, and not much is known about their function in bowed-string performance. A common opinion expressed in classical pedagogical works is that skewness of the bow inevitably leads to inferior sound quality. However in practice the bow is only rarely drawn exactly perpendicular to the string, without noticeably influencing the quality of the tone.

Aims of the Study

The major goal of this study is to clarify the use of skewness in violin and viola playing, and to identify possible factors on which it depends. This is done using empirical data from a performance study with advanced violin and viola players, whose bowing actions were recorded using a method developed by Schoonderwaldt and Demoucron. It could be hypothesized that skewness mainly has a secondary control function,i.e., that it is used by players to exert control of intentional changes in one or more main bowing parameters, without having a significant direct influence on the sound quality by itself. In particular, the possibility that players use skewness in a functional way to control bow-bridge distance is investigated The mechanism for the natural drift of the bow-string contact under influence of skewness will be explained. An analysis of the use of skewness in performances of crescendo and diminuendo bow strokes will be presented, demonstrating the role of skewness in the control of bow-bridge distance.


A number of evidence were found in the analyses, that support the notion that skewness was used by the players to control bow-bridge distance.


A simplified mechanism was presented for the natural drift of the bow-string contact point along the string when the string is bowed at an angle, and an equation was derived to calculate the drift velocity as a function of skewness and bow velocity…. Skewness can therefore be considered as a secondary bowing parameter, playing an important role in the interaction between the player and the instrument.

Given this function of skewness in playing it is hard to maintain the position that the bow by rule should be parallel to the bridge. This does not imply that a good player should not be able to perform straight bow strokes. However, it is probably most important that the player is able to exert a conscious control of skewness depending on the musical demands during performance.

This is a lot of information. Please take some time to absorb and think about this. The next article will be interactive as we look closely at different examples of professional musicians playing on YouTube. We look forward to hearing your thoughts and ideas on this subject!

Tags: ,


12 Responses to To Play Parallel or Not: A Controversy of Technique, Part 2

  1. He Sihao June 16, 2011 at 9:51 am #

    i claim that parallel is necessary and important for string students that is play parallel in practise will help us to make the sound wonderful and improve the ability of right hand.
    When you totally control your bow ,then you can do anything you want.

    You have to practise Parallel perfect.

  2. Blake Oliver June 16, 2011 at 11:33 am #

    Great article, Hans!

    I agree that skewness is a secondary bowing parameter. It is the only way to change the contact point–the distance between the bow and the bridge–while maintaining good tone. Here’s an example: Try playing down-bow at the fingerboard and then drawing the bow to the bridge in without angling it. It is impossible.

    It is important to play parallel to the bridge if you don’t want to change the contact point, but if you DO want to change the contact point, then it is necessary to angle the bow in relation to the bridge to either draw or push the bow toward the fingerboard or bridge as necessary. Changing the contact point changes tone, so it is important to know how to skew the bow properly if you want to be able to make more than one type of sound.

    The reason that students are taught to always play parallel to the bridge is that they usually have trouble maintaining the proper contact point, and therefore their tone suffers. What is important, however, is not always playing parallel. The important thing is to always have control over the contact point. So students should be taught not only to play parallel, but how to control the skew of the bow to achieve the desired contact point.

  3. Ryan Sweeney June 16, 2011 at 2:37 pm #

    I agree with Blake that parallel bow angle is taught to help students gain control over the bow. The ability to control bow angle/skewness, and through that maintain a desired bow-string contact point, seems to be an aspect of the bow technique that is very difficult to master.

    One thing that doesn’t seem to be discussed yet is the importance of bow distribution in determining the three fundamental parameters of bow usage (I haven’t yet read Schoonderwaldt’s paper though, and it may be covered there). Bow pressure/force, speed, and bow-string contact point are inextricably linked when it comes to tone production. If you change only one parameter, the tone will change, but it is possible to change a combination of these principal bowing parameters in order to maintain a relatively consistent tone quality. This is important when considering any change in bow distribution, small or large.

    For instance, if one plays a slow whole bow on a down-bow close to the bridge, but must then use the whole bow again on the up-bow in 1/3 the amount of time to get back to the frog, changes in bow pressure, speed, and contact point will be required so that a drastic change in tone can be avoided. Bow speed and pressure are controlled in a more obvious way – direct changes in muscle/arm use – but the only way to get the bow-string contact point closer to the fingerboard without an unpleasant sound will be to adjust the skewness of the bow in order to make the change.

    These types of changes in bow distribution are also happening in less obvious ways and on a very small scale throughout every piece of music. Consequently, we must compensate for these required changes in bow distribution by frequently adjusting our three principal bowing parameters, of which only bow-string contact point is affected by this secondary parameter – bow angle/skewness.

    String length is also an important factor in determining a desirable bow-string contact point. As our fingers move up and down the fingerboard, string length is constantly shortening and lengthening; Consequently the resistance in the string is changing and we must adjust our bow-string contact point to maintain a good sound. I think it’s safe to say that most music indicates frequent changes changes in pitch – often covering a wide range of registers – and therefore we are constantly adjusting our skewness to find that bow-string contact point that will produce the sound we are looking for.

    Given that our bow-string contact point is so heavily influenced by these two frequently changing parameters – bow distribution and string length – I think it makes perfect sense that it is rare to see any string player using a bow angle/skewness that is consistently parallel to the bridge.

    I’m looking forward to more ideas and thoughts on this subject. Thanks, Hans!

  4. Jeffrey Solow June 16, 2011 at 6:28 pm #

    One thing that should be noted is that ‘parallel to the bridge’ and ‘perpendicular to the string’ are not the same thing–especially on the cello. Skewness–bowing not perpendicular to the string–is a useful technique for adjusting the contact point (or in some cases, for keeping the same contact point as one reaches the tip). However, there are so many variables involved in controlling the bow and its contact point that it is difficult to make a categorical pronouncement about this. I would say categorically that the bow should not drift up and down the string unintentionally.

    • Hans Jensen June 17, 2011 at 5:17 pm #

      I like very much all these comments. Is it possible for you Jeff to elaborate a little more on this?
      Can you take us one or two steps further into your thinking about this:

      ‘parallel to the bridge’ and ‘perpendicular to the string’ are not the same thing–especially on the cello

      • Erwin Schoonderwaldt June 18, 2011 at 2:19 am #

        Just to return to the parallel-perpendicular question, you are quite right that there is a slight distinction. However, it is very small, and I used not to consider it. What I should have said between the parenthesis in my earlier post is “bowing parallel to the bridge is *approximately* the same as bowing perpendicular to the string”.

        Let me quantify the angles we are talking about for the violin (at the moment I don’t have a cello within reach to measure this on). At the bridge the distance between the string is about 11 mm, and at the nut it is about 5 mm. For a string length of about 30 cm the angle between adjacent strings is then slightly less than 1 degree, and the angle of the outer strings with the bridge deviates by about 1.5 degrees from perpendicular. To put this in a context: an angle of 1 degree corresponds to a displacement made by the minute hand on a clock during 10 seconds, which is hardly visible. The skewness angles we are talking about can be 10 degrees or more (displacement on the clock of 1 minute and 40 seconds).

        I expect that this is not much different on the cello. Even if the difference in the distance between the strings at the bridge and at the nut is supposedly larger, the string length is also much greater.

        I hope this clears it up 😉

  5. Erwin Schoonderwaldt June 17, 2011 at 7:42 am #

    Thanks for the interesting discussion about this, I am very excited that it is taken to this level. As additional information, I uploaded a short video on Youtube illuminating the mechanism of the drift of the bow under influence of skewness. Hope that it will make clear how it works. See

    Some reactions to the above comments:
    @He Sihao: I do agree that “the ability” to bow parallel to the bridge (which btw. IS the same as bowing perpendicular to the string) is an important skill. However, it is not the whole truth, and it can be helpful for players to realize that the skewness can help them achieve a change in bow-bridge distance. There is no reason for this not to be taught (e.g. Simon Fischer offers some nice exercises for this in his book). The most important point of this is that players should be made aware of how they can use skewness.

    @Ryan Sweeney: Thanks, these are very interesting points, and I totally agree on the importance of bow distribution. I haven’t really looked into mutual relations between skewness, bow distribution and other bowing parameters, but this can certainly be studied in the performance data I have recorded. I did do a study of how players vary the combination of bow velocity (speed), bow force and bow-bridge distance (contact point) for different dynamics and note durations. The original publication is “The player and the bowed string, published in the Journal of the Acoustical Society”. A summary of can be found in the introduction of my thesis, which you can find at

  6. Colin Cronin June 18, 2011 at 10:09 am #

    Thank you Mr. Schoonderwaldt for taking the time to post and respond to comments here. We are very humbled and grateful to be able to use your research to advance discussion and awareness of such ideas.

  7. William Cernota June 22, 2011 at 2:03 pm #

    I believe the bow should not travel in parallel to the bridge. My approach on the cello is based on the following points which, as usual, require modification moving from cello and bass to violin and viola:

    1. Definitions: current hair is that currently in direct contact with the string; new hair is the hair about to contact the string as the bow is moved in either direction, up or down. These simple definitions allow one to eliminate confusion between down and up bows and develop a simple general principle.

    2. Drift: (A) If the new hair is aimed below the current hair (toward the floor), the bow will naturally drift toward the bridge, maintaining a core of sound. Control of the drift will prevent the bow from entering the ponticello region. Drift is best tested by the player to experience it. (B) Conversely, if the new hair (again in either direction, up or down bow from the current hair) is aimed above the current hair, the bow will naturally drift toward the sul tasto region, diminishing the core of sound and also easily confirmed by testing. Considering the two directions of up and down bow and the two ‘aims’, there are four possibilities to test.

    3. Frown vs. smile: When viewed in a mirror or by an observer standing in front of the cellist, the bow path of (A) above, in either direction in its purest form resembles an asymmetrical frown; (B) above resembles an asymmetrical smile. Note that at the bow changes, the direction of hair ‘aim’ changes. If it is not already evident, I will state that in my opinion, a frown is the better basic path for the bow because it continually guides the bow naturally toward the bridge rather than toward the finger board. It provides grist for the mill of sound.

    4. The corners of the frown: Rather than make a sudden change of hair direction on the bow change, the aim is adjusted slightly before the change. At the tip on a down bow, for example, the direction of the new hair is adjusted so just before the change it is already prepared for the up bow direction in a caressing manner. It is also important to note that the frown is not symmetrical like an upside down half-circle; that image would imply the adjustment of hair direction is happening far too soon.

    5. ‘Figure eights’ and confirmation of the above: While I developed the above images on my own, it was encouraging to find confirmation in the words of Leonard Rose and his much admired colleague and violinist, Oscar Shumsky. ‘Figure eights’, referred to by both of these musician/teachers summarizes the ideas of new hair being maintained below the current hair, anticipating the bow change by adjusting the direction of the hair prior to the change, and continuing with that direction until just prior to the next bow change.

    Best Regards,

    William Cernota
    Member, Chicago Lyric Opera Orchestra Cello Section
    Faculty, Northwestern University and Loyola University Chicago

    • Erwin Schoonderwaldt June 22, 2011 at 3:10 pm #

      Thanks, I like the metaphors you are using. This will certainly help people to understand this mechanism, while remaining physically correct. Of course the shape of the bowing trajectory (by which I mean, the trajectory followed by the bowing hand guiding the bow) will depend on the musical needs (e.g., crescendo, diminuendo or constant dynamics).

      Regarding point 3, I did observe the “frown” in my measurements of violin bowing in long notes with constant dynamics, where players change the aiming in the vicinity of the bow change. This is described in the skewness paper referred to above. Concerning the timing of the change I observed changes of aiming both before and after the bow change, but I can’t really say what is optimal. The exact shape of the bowing trajectory depended much on the individual, and it could also different for bow changes at the tip and at the frog.

      Regarding point 5, I sincerely wonder why this type of bowing path is generally referred to as “figure eight”. In fact, I find that quite misleading: if you consider the movement of the hand (which in my opinion is closest to the intuition of the player) a complete up-and-down bow stroke looks rather like a single loop, not a double loop as in an 8. I would be grateful if somebody could explain why this “figure eight” concept is so established in this context. Who came up with it first? Is perhaps something else meant by it, or is it just wrong?

  8. Hans Jørgen Jensen June 24, 2011 at 6:45 am #

    Here is also a terrific video where Todd Ehle explains the figure eight bowing:

  9. Joel May 28, 2014 at 12:42 pm #

    Erwin, I believe the figure 8 describes the movement you may observe if you watch both ends of the bow while someone plays in the manner that Mr. Cernota mentions.
    Thanks for the great article and discussion.

Leave a Reply