deBreved
The More You Bow

There are many cool French and Italian terms for different bowing techniques. The texts have explanations of each of the different bow strokes, but these explanations range from excellent to confusing to plain contradictory (just google détaché). These terms refer to things the section does ‘under the hood.’ As an orchestrator, it is important to understand these, as they are the rudiments of string playing. Remember that in the end, though, the important thing is to know how to notate your intent; the section will take care of the technique.

When discussing bow technique it is important to note that several terms players use to describe a stroke are also terms used in notation; however, they have a slightly different meaning in this case. The two main ones to watch out for are staccato and legato, which will be discussed in this section.

The default for bowing is to play on the string and change direction with each note.

Long Notes
Long notes are played on the string, or shorthand on, meaning that between notes the bow does not leave the string, but rather maintains contact. The bow will change direction for each new note. If you want more than one note per stroke, use a slur to designate how many notes should fall under each bow.

How many notes you can write per bow depends on several variables. Some are non-negotiable; for example, at loud dynamics the bow must move faster to maintain a full tone, so fewer notes are possible. When the dynamic is soft, many more notes can be played and the number of changes and the speed of the bow can make some nice colors. Watch a section bow; the speed changes all the time and this is what gives the phrase life. There is a third variable: placement. Players will put the bow in slightly different positions on the string to produce different sounds. The extremes are sul ponticello and sul tasto, but there are many little variations in the middle that when combined with speed and pressure give every note a unique sound. They can also use different parts of the bow and take advantage of physics. for example it is much easier to play softer at or towards the tip as the bow is lighter there. I talk about this sort of thing in Swell Enough and Tremolo and the Abstract Truth.

The reason why string samples have no ‘life’ is that every note is recorded at similar speeds and pressures, so there is no relationship between technique and musical context.

The ‘L’ word
I avoid using the word ‘legato’ when writing for or talking to string sections. Some people use it to mean smooth and connected, sometimes mixing it with slurs. The problem is that in some countries, the term ‘legato’ literally means to slur. So, don’t use it as a descriptive word in the score, part, or when talking, unless you are in one of the countries where it means to slur and you are making a change and telling them to slur!
Additionally, the word is not needed if you get everything else correct. Remember, the default is to connect and play legato, so unless you have told them to play it another way, or there is some other context that would lead them not to play that way, the word is unnecessary.

Short Notes
The default is still to play on the string. However, context and physics start to affect what is going to happen and if you learn these things you can use them to your advantage by predicting what will happen and then notating accordingly. 

The alternative to bowing on the string is, of course, bowing off the string. This is where the bow leaves the string between notes. The most common example of an off-the-string bowing style is spiccato. One common misconception is that spiccato itself is the way you label a passage as off the string. It is not. It is just one particular style of off-the-string bowing. However, if you do use that word and it is perhaps not the correct stroke for the phrase, players will do what they do best and adapt their bow stroke to the current musical context and ignore your indications. But, in general, I don’t tell them what stroke to use, give them the correct notation and they will do the correct stroke.

I believe the misunderstanding and misuse of the terms marcato and spiccato is due to certain string libraries using these names for their short note patches. In the real world, it is impossible to play spiccato as loud and heavy as it sounds in these patches, and as mentioned above, historically marcato in general is a type of attack, not duration.

‘Off-ish’
Spiccato is a very useful and common bow stroke, but if you look at the physics of it, it is not useful for heavier attacks and louder dynamics. For these we need more contact with the string and bouncing the bow off does not allow this. But we still want the bow to come off between notes, even if it is very slight. This is a stroke that is between a spiccato and a marcato. It is not discussed often but if you watch a section, you will see it is what naturally happens when they have to play short but loud notes. I have also noticed that even when asked to play on the string, when the music gets fast and loud, the same thing happens and they gravitate to this same stroke, the bow will lift a little as it changes direction. A combination of physics and performance practice are at play. They also often do this for clarity. In cello and bass, playing fast on the string does not project or come through as well as it does on the violin. Note that this is not really an official term!

What really happens?
Ever wondered why a repeated note figure played by a section sounds so much better than one played by samples? Considering you can have lots of round robins and no issues with connections or legato transitions, one would think it would sound pretty similar. I think I have figured it out, or at least identified one of the big reasons why.

If you watch a section playing these sorts of short note figures, common in film scores, they are all playing it slightly differently. This variation leads to a blend that is unique to that passage of music, giving it a life that samples never achieve. We can predict how it will sound and what sorts of strokes will be used, but when 12 players do it and blend based on their ears, rather than following strict instructions in a sampling session, you get subtle variations in the stroke used, which is a good thing!

When they make samples, the musicians all play the same bow stroke and produce a bunch of identical recordings that are edited to create a predictable and programmable sample library. This results in a very homogeneous and sterile sound.

 

Under Pressure: Looking at the Bow(ie)

Notice how I have bowed the above phrase. Even though I have not marked a crescendo, there will be a slight one as the bow speed and emphasis change and the sound gets more intense. If I did mark this as a crescendo I would also want the same bowing. The louder the sound, the faster the bow has to travel to maintain a quality tone. If it were a big crescendo and I had not marked it like this they would automatically go to at least two bows per bar as the music gets louder.

There are so many subtle differences to the sound you can make when you think about the bowing.

Here is Endre Granate demonstrating these strokes.

Bowing does not always follow the phrase, but how you bow makes the phrase.

1. There is a school of thought that says you should use a long slur as a phrase mark and the section will sort it out. Yes, they can work something out, but you will get the result YOU want if you take the time to bow the passage yourself. In this case, the principal will decide how to bow it. But they do not know the music or how intense I would like each note.
2. This is how they would probably play it, starting on an up-bow of course. They may also slur C to Ab. Each bar uses the same amount of bow, and the sound will be consistent.
3. How I might do it. This will give the phrase much more life than the even bowing of #2. The more notes, the slower the bow but more pressure is applied. Fewer notes or single ones lead the bow to move faster with less pressure, giving a different intensity and some nice phrasing. So while the section will use a whole bow to cover the six beats of the Ab and G, they will use half to three quarters just for the Eb. This is what makes the music magical, and why samples have a hard time replacing the emotion of a real section.

Keep in mind that even though the notes may be slurred, there are often places where the players will have to change strings. These notes will not have the ‘slurred’ sound as the new string will not be vibrating before the next note starts. Players can work around this in many cases by playing the line in a position that avoids the string change, but if the interval is wide, this may not be possible. It is a very subtle effect, and good players will almost be able to mask it, but it is still something to think about.

In reality, a good section can change bow direction without you knowing. This is very easy on the violin, but gets harder on the larger instruments. On the violin, the string is not moving very far as it vibrates, but on the bass it is moving a lot, so stopping and starting the string as the bow changes is more perceptible. On long notes the players will automatically stagger the bowing and you won’t hear any changes. No special notation or instructions are needed for this, just write the duration. I always have a bit of a chuckle when I see a score and on the long sustained notes the orchestrator has written something like ‘add lib bow changes’ or ‘stagger bow changes’!

Sometimes you will not want to hear any bow changes as the notes change on a line, just a completely smooth and even line. In order to get this effect you will need the section to change bow at different times. This can be dictated by writing two overlapping bowing patterns, one set of slurs above and one below. This will mask the changes and is sometimes referred to as ‘divisi bowing.’ In the studios we often mark the phrase with a long slur and ask the section to ‘free bow’. They will then change bow at different times, not as a group. Unfortunately ‘free bow’ is not a universal term. I have not found a common way to describe this practice, so some clarification is often needed to keep the section from bowing the line all together. If you just write a phrase mark with no other indication, the players will follow the principal.

As previously mentioned, there are many things that effect how many notes can (physically) or should (musically) go under each bow, and even the most experienced orchestrator cannot get it right all the time. Lucky for us, we are writing for humans, not computers. They can make judgement calls and adapt the bowing to suit the musical situation. So if you do screw it up, don’t worry, you are still giving the section invaluable clues as to what you are looking for.

Check out my Error Code 7 article for some great score and audio examples of how slurs work and effect the intensity of the music.

Watching and listening is the best way to learn how to bow.

Useful articulations

a. No emphasis, no space between notes.
b. Slight emphasis, very small space due to the bow recharging for the emphasis.
c. Detached, bow stops between strokes.
d. Heavy attack with a drop-off to a little space as the bow recharges for the next attack. You can add a tenuto to limit the space, but the accent will be less.
e. Louré, all notes under one bow. The bow pulses the notes by slowing, releasing the pressure then applying and speeding up again.
f. As above but the bow pauses between notes.

If you want an extra smooth bow change, mark it sostenuto. This is the string equivalent to legato tonguing on a wind instrument.

Notation
As stated, the default is to play on the string, so no special notation is required if that is what you want. But now we get into territory where the string terms start to mean one thing to the players, but another in notation. Staccato, when used in the score, refers to the length of the note and it does not necessarily mean to play the staccato bow stroke.

In my scores, if I want off the string I mark it staccato (with dots or the word stacc.). I prefer to use the word not the dots. It is cleaner, easier to change if we want to play it differently, and it keeps the notes clear for when you want to add accents. See how much less cluttered b. looks and how it could be changed to ‘on the string’ by just taking out the word stacc.

If strings are playing eighth notes at light dynamics they will play spiccato and as the dynamic gets higher or more intensity is required they will go into the ‘off-ish’ style I described above. The players adapt the bow style to the dynamic and intensity of the music, and I do not need to micro-manage them.

If I want it heavy I label it stacc. e marc. If you accent every note you have no way of getting an accent when you really want one. This is the notational equivalent to ‘crying wolf.’

[2024 Update]
While I wrote like the above for many years, I realized a few years ago that the ‘stacc’ part is redundant. To achieve the marcato effect, there will naturally be a separation of the notes, making them short. So these days, I would just write ‘marc.’

Another realization is that, in many cases, you don’t even need the staccato text or dots. This somewhat contradicts my original statement that the default with no other information is to play ‘on’ the string. When a section sees a bunch of quavers or semiquavers, they will play them short and thus ‘off’ the string. The faster the tempo, the more obvious and understandable this becomes. However, if I have a passage that is slower and could be played either way, it will still be played ‘off,’ so I need to do something to break that default. You could use tenutos, but I usually just add the text ‘on.’

Here are some examples from recent projects, both studio and live, where I did not feel the need to add any info. In all cases they came out short. There are a few other things worth noting in these examples too.

Let’s look at the alternatives for this rhythm.

1. How most people would do it. The staccatos are redundant on the quavers and double redundant on the semiquavers as they are already short.
2. At least they know the semi’s will be short, an improvement!
3. I see this often, breaking it all down to the shortest duration. Some people think it will make sure every note will come out the same length, it won’t! It just makes it harder to read. And of course the dots are redundant.
4. A slight improvement but still harder to read.
5. What I did, the cleanest and easiest to read.

As I mentioned before, some people think all the notes will be the same length, but they won’t be, and it has nothing to do with the notation. The reason the quavers will be a tiny bit longer, and the note on beat 4 a little longer still, is that when there is space after the note, the players will give it more bow. This is really obvious if you have a stream of quavers or semiquavers that just stop; the last note will be longer. It is natural, and we are so used to hearing it that most people will not even notice it. Of course, I do, and I often get blank ‘what are you talking about?’ looks from the players who don’t even notice they do it when I bring it up. It is similar to how they will taper off a long note at the end of a phrase; that is just what happens when they get towards the tip of the bow and there is nothing after it. Another time string players do something and do not notice or think it is weird is they will leave a slight gap when playing a descending fifth up high. This is because they need to shift and it is easier with a sligh break. When I point it out I get the stare, I will offer to play it back to them if they don’t believe me. The next take will be fine, as they will push through it even though it is harder, but they got busted taking the easy route.

---

Accents
Accents are performed by adding pressure, speed, and duration to the contact of the bow and string. In the example below, the accented notes will be a little longer than the unaccented ones.

This sounds very natural and is easy to play at slow and medium tempos, up to 130 bpm or so. But when it gets faster, or if this were sixteenth notes, the effort needed to make the accents can make it difficult to keep the unaccented notes smooth and metric. When I hear this problem in the studio, I have the players only think about the accents. But it is always good to think ahead and orchestrate around this problem.

One would not think that a pattern like this would cause any trouble, and maybe to some it does not, but being a drummer, I want everything as even and metric as possible, and when I hear the violins play this the time will slightly push and pull around the accents. There are several solutions to this problem. In this case, as the accented notes are a different pitch, they will stick out anyway, so recently when this came up at a session, I just had them ignore the accents. But let’s imagine that everything is the same pitch, or the composer really wants big accents. Here is what you do –

I have taken the accents out of the first part and the seconds are just playing the accented notes. You will not miss the seconds on all the notes, and the firsts will actually be able to play more heavily now as they do not have to allow for the accents. Finally, the timing will also be perfect. The finished product will sound much better and be easier to play. If there are more than strings in the orchestra one could also add some woodwinds, perhaps an oboe with the seconds. Percussion can also help. In fact if you have percussion and brass hitting the accents, you do not need them in the strings at all, as the overall result will sound accented.

I talk more about timing issues here in my conducting article.

Labeling Bow Direction and Bow Position
99% of the the time this is not needed. String players have been bowing since they were four years old, and choosing to start with an up- or down-bow is second nature. In general if a melodic phrase starts on a weak beat (a pickup, for example) that will be an up-bow. If it starts on a strong beat, that will be down, simple as that. Long notes at soft dynamics start on an up-bow as that is the easiest way to start softly. The one time it is common to label is when you want a series of heavy attacks, in which case you mark all down-bows. In reality, players are pretty good at a heavy attack in either direction, but what a series of down-bows does is ensures that there is a gap between notes as the player has to retake between each one. They will also play at the frog. Many times I have seen a series of down-bows written and it has been ignored by the players though so only do it if it really makes musical and technical sense.

The bow is held at the frog, so this is the easiest place for the player to exert the most pressure on the string. The tip is the furthest point from the hand, so consequently it is the easiest place to play with the least pressure, producing the lightest tone. Some orchestrators mark to play at the tip or frog. Just like direction, I have found the players tend to pick the best position with no input other than a dynamic. They will play a quiet tremolo at the tip, and a ff stab at the frog, because it is second nature. But should you wish to break this default, then mark it otherwise.

So while I don’t mark the directions, I have a pretty good idea of what they will do. Another thing to consider when working out what your want, and what they will do is to go backwards. If the phrase ends with crescendo, they will do that, or at least the last part of that on an up-bow, as that is where they can easily make the most sound, at the frog. If that note end with a ‘button’ or hit, then then easily then pull down for the big finish. Conversely, if a phrase ends with a dim, or just normally, they will try to end with a down-bow as that is the easiest way to release musically.

Another reminder to check out part 2 of the Error Code 7 article for lots more information and examples of how the bow works and how to notate what you want.

Styles
Ordinario (ord.) – The default, use ord. to cancel any of the following. The bow is in a neutral position, the ‘sweet spot’ between the fingerboard and the bridge. This is not a set position as it shifts depending on the string length. Some people label this as Normale (norm.)

Sul Tasto – On the fingerboard. The bow plays over the end of the fingerboard, producing a softer tone with a less edgy attack and less high harmonics. See below for more discussion on this technique.

Sul Ponticello (Sul Pont.) – On the bridge. Opposite sound of Sul Tasto. The bow moves closer to the bridge producing a raspy, edgy tone. There is less fundamental and more high harmonics to the sound. If you go extremely close to the bridge you can completely lose the fundamental and will only get harmonics.

Flautando – The bow moves faster than normal with very light pressure, aiming to produce a pure, flute-like tone. Some players describe it this way, but in practice, there are varying interpretations and approaches. Often, a fast bow stroke isn’t practical. I view it more as a mindset or concept rather than a strict technique. Personally, I avoid using the term, similar to how I handle legato or détaché.

Col Legno – ‘With the wood‘. There are two types of col legno. Tratto is where you bow the string with the wooden part of the bow. Battuto is where the wood strikes the string. The default is battuto. Keep in mind, the sound of col legno samples is never replicated in the real world. Players do not like smashing their $10K+ bow into their $200K+ instrument. Sometimes the players might get out their 2nd, not so precious bow or a pencil. I have heard of people asking them to play with rulers or pieces of dowel, just don’t! It does not matter, they will not hit the instrument ‘that’ hard with anything. In scores where the composer wants loud, unrealistic col legno, we usually just use the samples and don’t even try to play it. If the passage is only some isolated notes, you can substitute snap pizz. When it comes to Tratto, it hardly makes any noise, and what noise it does is not very pitched. To make this work, players usually let a little bit of bow hair contact the string. That recording where you thought you heard col legno tratto, maybe not!

Ricochet and jeté are off the string strokes where the bow plays multiple notes in one direction by bouncing. I have found these impractical for studio work unless it is just an effect. Multiple bounces at medium and fast tempi are just not precise enough for what we need in modern recording. I sometimes hear a player suggesting it as a way to play a fast, light figure and will nicely ask them to play it with separate strokes, it takes more effort but it will be much more precise.

Circular Bowing
In recent years this has become a very popular technique in the scoring world. Instead of the bow staying in one position in relation to the bridge and moving from side to side (whether normal, sul pont, or sul tasto), it circles, covering all of those spots. I have found that players tend to do this quite fast but I much prefer the sound of them slowly circling. I use this technique in my piece Error Code 7 and explain it here and there is also a demonstration. The best notation for this is just simple text but I know some of you won’t be able to resist adding some pretty picture of a bow circling!

Circular bowing is great for a ‘wire’ or subtle texture, but there is a lot of randomness in the texture and it is not very ‘dynamic’. If I need something to pop out, I will often have them move from sul tasto to sul ponticello and back as a section. You can combine this with a swell to great effect. I have notated this in various ways. In Error Code 7 I used the instruction “tasto to extreme pont. and back with each swell.” However, I have found that simply saying “to pont. and back w/swell” achieves the same result. Some people feel the need to add extra lines and arrows, but if you consider it carefully, the swell marking already conveys this, making the additional notations unnecessary—except, perhaps, for an Instagram flex.

This effect also works really well with tremolo and trills. A trick I often use to add a bit of life to this kind of voicing, especially where there are minor seconds, is to have the lower note trill. Regarding notation, I only provide detailed instructions on the first swell. The note lengths and swell are the same in the second phrase, so it is obvious that the same instructions apply.

The Left Hand
Apart from the lowest notes on the lowest string, and the highest notes on the highest string, all other notes can be played in more than one position. The higher up the string the position is (within reason), the darker and less edgy the timbre will be. This can be notated: Sul G (‘on the G string’) passages are very common on the violin, but in many other cases this is the sort of judgement call a player will make based on the musical situation. While this has nothing to do with the bow, it affects the sound as higher positions mean shorter strings.

Sul Tasto – Not So Fasto
I have noticed from the podium that when the violins have been asked to play sul tasto, they often do not appear to be playing over the fingerboard. It was not until researching this article that I learned what was really going on. There are some technical and sonic factors to consider when playing over the fingerboard. I have heard different takes on what is happening when performers see this indication, but one thing is for sure: sul tasto is not as cut and dried as any of the books would have you think.

In the case of the violin, the strings are relatively short, so the ‘sweet spot’ where the bow makes the best sound is quite small. It is also constantly shifting depending on the hand position. As the fingers shorten the string, the sweet spot moves closer to the bridge. In fact when playing in the highest range of the violin, in order to maintain the correct ratio and thus the sweet spot, the bow is actually at the bridge. In higher positions if you really play sul tasto, the strings does not vibrate well. You are now bowing too close to the middle of the string and it is impossible to produce a good tone.

But there is more. At the bridge or in the normal position, the strings are in a nice arc that allows the player to play any string without touching the other ones with the bow. If you are playing in anything other than first position and the bow is close to or on the fingerboard, pressing the strings flattens this arc and it becomes much harder (sometimes impossible) to avoid sounding the other strings. On some violins, and the viola too, it is also possible for the bow to hit the body of the instrument when playing on the outer strings. In the normal position, the bow falls right into the cutout of the body, but when you move over the fingerboard, the shape of the instrument no longer works in your favor.

The true sound of sul tasto on the violin is a wispy, almost fuzzy tone. In many cases, what we are really asking for with sul tasto is a softer, less edgy tone. Some would call it muted, but the actual sound of con sordino playing and sul tasto are not that close. There are a number of ways to create this sound. The violinists accomplish the sul tasto effect by using a combination of light bow pressure and slow bow speed. They may indeed move closer to the fingerboard depending on the pitch, but not always. As the instruments get bigger, so too does the sweet spot, and actually playing on the fingerboard becomes no problem at all. Viola has some issues similar to the violin but there are none on the cello or bass. Other techniques players use to make a softer tone include playing in a higher position on a lower string and using less hair on the string.

In the studio, players may make a judgment call in order to create a fuller sound that blends with the other instruments and projects well for the microphones, all while catching the essence of sul tasto and avoiding the technical difficulties.

Talk to players, and you’ll discover a range of opinions about this technique. In the concert hall, violinists will try for a more literal sul tasto than they do in a recording studio. As one player put it, “I would try to play a little bit over the fingerboard, just not as extreme as a conductor may want to see.” The same player also noted that everyone’s instrument is a little different, so their approach may need to be different too. The situation here seems to be similar to that of double stops or divisi, where what actually goes on has not quite filtered back to the authors of the text books.

Based on the musical surroundings, players may play sul tasto (or fake it) without being told. In fact, they are always making little adjustments, as depending on pitch, volume and context, the sweet spot may be in a different place for every note. Remember, the higher the note on the string, the closer to the bridge that sweet spot will be.

Instead of the technique being called ‘on the fingerboard’ a better definition would be ‘toward the fingerboard’. With all of this in mind, you will see that sul tasto is a type of color or effect, and not necessarily just the technique of bowing over the fingerboard.

Here is Endre Granat discussing and demonstrating the ‘sweet spot’, Sul Tasto and Sul Ponticello and Con Sordino.

---

---