True legato is a difficult effect to produce on the marimba. It involves trying to match the attack level of a note to the sustain level of a previously played note. On the marimba, this is easier when descending in pitch as the natural sustain time (and the perception of sustain level) increases. While playing an ascending figure, the player must decrescendo to create the illusion of legato. In either case, if the sticks are hard enough to produce too many upper partials (which have much faster decay time), it is a poor representation of legato. The natural shape of the ADSR envelope (attack, decay, sustain, release) of the bars of a marimba represents staccato in the upper range, a large dose of tenuto in the middle range, and sort of legato in the lower range. The player can only control the attack level which in turn influences the other parts of the envelope with the exception of the release, which the player cannot influence without physically stopping the vibration of the bar by dead sticking, stick dampening or using parts of the hand. That's why I reserve the terms "legato" and "staccato" for another application altogether, which will be the subject of this article. I'm also going to forego any discussion of artificial alteration of the natural release time of the bar such as muffling or muting.
Let's think about the legato effects available to the violinist. Using a single bow motion, the violist can change pitch by simply placing his/her finger in another location on the string. If the player is careful not to "hammer" this fingering, there will be hardly any noticeable articulation. The player can also make use of glissando or portamento to change pitches within a single bow motion and then there will be even less articulation. The marimbist cannot begin to make these sounds. If a player could accurately play a note with an attack level equal to a preceding note's sustain level, the louder first note's continuing sustain would overtake the second note's sustain level immediately and the listener wouldn't be able to hear it. Our sound is more like the pianist's in that we must create a different kind of legato. But even still, the sound envelope of almost any note played on the piano has a much more gradual decay and a considerably greater sustain than a note played on the marimba. Also, the pianist can control the release time of a note by holding the key down or using the damper pedal. The very best we can do is to use sticks that are sufficiently soft so the timber of the attack is similar to the timber of the sustain. As the decay, sustain, and release portions of the envelope decreases as the pitches rise, even softer sticks will fail to produce a true legato articulation. At louder volume levels the attack will be so much greater than the sustain level that legato is impossible.
So how does the marimbist create legato? How do we do any articulation? I am not referring to obvious things such as rolling long notes. If that were the only way to achieve legato, then we could dismiss the discussion altogether. It can get cumbersome- think of playing the passage below (Fig. 1) with a series of short rolls on the slurred notes and single strokes on the staccato notes... Nah. You could also just use a soft stick for the long notes and a hard one for the short notes. Yeah, right. You could also take advantage of the relative sustain times offered by playing the long notes in the center of the bar and the short notes near the string. That will work for small groupings, or at some dynamic levels for larger passages. You have to deal with the fact that playing in alternate playing areas also produces a different timbre, which may or may not enhance the effect. This technique is only really noticeable if used in contrast with a normal full sound produced by playing in the sweet spot. Play in either spot for any length of time, and we're back to the not-quite-legato, not-quite-staccato sound.
One thing will definitely not work. I am referring to the exaggerated slow motion stroke that Leigh Stevens refers to as the "resurrection stroke." A player is not going to lengthen the sustain of the note or accurately control its attack this way. "Play with a looser grip for legato," isn't going to yield any perceivable result either.
There is another solution to solving articulation issues that I found long ago, and quite by accident. I had been listening to a pianist performing a series of appoggiatura: an eighth note slurred to a staccato eighth note and successive repetitions of the same articulation:
Figure 1
I was struck by the impression that there seemed to be a slight pulsation of the notes on the strong beats, as though a similar effect could be created by dynamics alone:
Figure 2
In listening to string player's interpretation
of this same figure I noticed that each pair of notes received
a change of bow- down bow on the first beat, up bow on the second
beat, and so on. This creates a natural pulse regardless of how
short the even numbered notes are. The same phenomenon can be
observed in the tonguing used by wind and brass players. It's
not just the length of the notes, but the natural stresses of
different dynamics.
Try this experiment: find your favorite gigue from one of the unaccompanied violin or cello suites by J. S. Bach. (If you can find one in 6/8 or 12/8 this might work better than one in 3/8). Play with a slight pulse on the first note of each major grouping, i.e. each strong beat or dotted-quarter. Now play the same piece with even dynamics on every note. Doesn't the second rendering sound more legato? Granted, either way if there isn't some amount of variation, it will get boring. But this is probably the best method we have at our disposal for creating different articulations: relative dynamic stress. Some performers would say that playing with pulse creates a lighter feel. I understand this concept but would qualify it by adding that it depends to some extent on the overall volume of the passage. What we have accomplished in this exercise is a varying degree of legato. But I don't think we have made a clear distinction between legato and staccato. Staccato without artificially stopping the sustain of each note is only really possible in the upper register of the instrument. I think the best imitation of staccato we can do is to play on the bar nearer to the string where the sustain is shorter. But then, there is the problem of timbre...
Articulation is one of the aspects of phrasing that contributes to expressive performance. Our articulation has a different definition than those of other instruments. Some of our articulations are less effective than others. For example, rolling is a series of re-articulations of either a single note or as many as four at a time. To many listeners, this is a poor attempt to create a smooth sustained sound. Well, that's the sound we make, no apology necessary. It's not as severe as the tremolo on the mandolin, and it's at least as effective as what the harpsichord player has to contend with. Our phrasing is accomplished with dynamics and the creative use of time and rhythm. I put "phrasing" in italics because I want to make a distinction between the phrase and phrasing.
Too many players are not aware that too much use of dynamics and manipulation of time can break the phrase itself. For example, in the music of Bach the phrases are usually at least two measures long and often as long as eight measures. There are plenty of instances where one could make a case for creating a phrase in a shorter context than another player might, such as the common practice of "terracing" the dynamics when a measure is repeated, or during sequences. I prefer to look for longer lines which means taking care that my use of phrasing serves the length and structure of the phrase. Look at the Cello Sonata in G major of J. S. Bach, first movement: Praeludium.
Figure 3
In the first four measures how many phrases are there? Some players delineate eight distinct statements. Others four, some two. I see only one phrase, and it doesn't have a definitive end but rather a transition into the next phrase. I may take a little liberty with the length with the very first note of the piece, and I may give each succeeding bass note a little more weight than the other notes of the measure. But if in every eight note figure I give the bass note too much attention (preparing for it by stretching the time and then playing it as though it were the beginning of another phrase) I will not achieve any continuity. Any dynamic scheme that identifies a climax or a resolution can be obscured by too much short term phrasing. How a player treats each note in the eight note figure should be a matter of articulation- there are different places for more or less stress. My own interpretation would be tedious to describe, and I always leave a little room for "playing in the moment." Suffice it to say that I think articulation and phrasing should be done in the context of the structure of the entire phrase. In the example of this particular piece of music, I try to create the effect of legato, that is, play the notes as evenly as possible, with only the slightest of nuance in articulation.
Now consider another dilemma. In the same movement observe measures nine and ten- a real roller coaster:
Figure 4
Look at the third full beat of measure nine.
The "G" is an octave displacement from the direction
beat two was appearing to be heading. It happens on a strong beat
so a little extra volume isn't going to hurt anything. But just
try to play it at the same volume as the last note of beat two
(the "A") and it becomes very difficult to tell whether
you accented the "G" or not. The phenomenon is even
more exposed at the end of beat three. Here, the high "D"
is also an octave displacement from the preceding three, it is
the highest note in the measure and is placed in the weakest part
of the beat. Try to not accent that note! (It almost accents itself).
It's not hard if you think about putting more weight on the following
"C#." The same will work in the following measure with
all those "Ds" on weak parts of the beat: remember where
the pulse should be. This articulation problem is caused by a
phenomenon is what I call the "breakaway effect." Sometimes
it is caused by having a greater distance to travel and neglecting
to take into account that the larger trajectory can translate
into increased stick velocity, and you know what that means: more
volume. Other times the effect is caused by the natural acoustics
of the marimba. For the most part, it is a psychoacoustic phenomenon:
is it louder or does it just seem that way? In any case, I don't
think this particular passage has inherent syncopation implied,
so careful consideration of the articulation is called for. There
are better places to get funky with Bach.
Another way some players handle this passage is by giving extra weight to the first three notes of beat three: G-F#-E and taking a little extra time for the E. Then they echo that effect on the first beat of measure ten: (F#-E-D) as though it was a short sequence worth pointing out. Not a bad idea, but you have to be careful not to overdo it and disrupt the flow into the resolution: F#-E-D. Besides, too much attention to this "sequence" leaves the remaining three beats of measure ten to just dawdle around.
By the way, the excerpt above is the end of a phrase, and consists of measures nine and ten. If measures one through four constitute one phrase, how many phrases are there in the six measures between five and ten? What do you think? The only wrong answers involve odd numbers of measures.
OK, so I moved from the discovery of a working definition of legato to the much more complex and subjective topic of phrasing. The point to remember is that for the most part, articulation for the marimbist is more about dynamics than varying the length of notes. Since dynamic variation is the most obvious way to delineate the phrase, the two subjects are naturally related.
Remember that the title of this article is Just What Do You Mean, Legato? We still don't have a working definition of staccato. Know what? Right now I don't think I even want one.
Copyright © 2002 by Tyler-Rounds Productions