Skip to main content
Loading Explore guides…
First Species Counterpoint · Acedra
Home
Explore
Music
First Species Counterpoint
Sign in to save
Chat
If YouTube asks you to sign in, open the video on YouTube.
Watch on YouTube
Reload player
Transcript
67 segments
Copy
Hide video
0:09
Hello. My name is Dr. Jacob Gran, and in this video I will teach you how to compose one-to-one counterpoint also known as "first species" counterpoint. This kind of composition exercise is pretty ancient, and most of the great classical composers were taught using this method. Haydn, Mozart, and Beethoven were all taught species counterpoint, and when they had students of their own, this was the first thing they taught them. Even when
0:37
Arnold Schoenberg came to the United States in the 1930s, and his students were very excited to get to learn all of the cutting-edge atonal stuff that he was most known for, instead he disappointed them by teaching semester after semester of counterpoint. Just like a physical exercise will build up certain muscles, these counterpoint exercises will build up the skills involved in composing simultaneous melodies. Western music is full of instances where
1:07
this skill is useful: whether you are trying to compose in the style of Mozart or to compose in the style of Schoenberg or you're trying to improvise a guitar solo in popular music.
1:20
This video will be broken into three main parts. First we will learn the rules of one-to-one counterpoint by looking closely at an example by Schubert and his teacher Salieri. Then we'll do an exercise together and I'll talk you through my thought process. One thing I like about counterpoint exercises is that there's always more than one right answer, and so there's room for creativity even in the easiest exercises. Lastly, I'll leave you with some homework.
1:56
Here's an example of a given melody which you'll often hear referred to as the "Cantus firmus," which just means "given melody" in Latin. These tend to be pretty short -- somewhere between eight and twelve notes. The given melody is provided by the teacher and the student must compose a new melody to go with it. This one was written by the Viennese classical composer Antonio Salieri, who some of you might know better as "the guy who did NOT kill Mozart." I like to think of
2:33
the Cantus firmus a little bit like what a haiku is in poetry. This is the fewest possible notes that still give the impression of a complete composition. It avoids anything unnecessary; so for instance you'll notice that there are no directly repeated tones. We also don't repeat groups of tones, like melodic motives, since we want the given melody to be a complete whole that can't be broken up into smaller parts. So,
3:03
there's no question-and-answer, or anything like that. When we compose polyphonic music, we want the listener to be able to split their attention between multiple melodies, and to make that as easy as possible for the listener, we want each of our melodies to be melodically fluent. Melodic fluency doesn't mean that the melody is the prettiest one ever written or that it's the catchiest or anything like that. It means that the melody
3:31
takes very little mental effort to keep track of. Here's what Salieri's given melody sounds like.
3:51
The first thing to notice is that the Cantus is mostly stepwise. The melody is more predictable and easier to keep track of if the listener knows that it is most likely to go somewhere nearby - one of the stepwise adjacent tones of the scale. If the melody instead leapt all over the place, it would be hard to predict its motion and that would force the listener to put more effort into paying
4:23
close attention, which is not what we want. When the Cantus does leap, it is always by a consonant interval. Salieri includes only two leaps and they are both perfect fourths. On the white notes of the piano, the consonant leaps are thirds, fourths (except for F to B), fifths (except for B to F), sixths, and octaves. We are going to avoid using dissonant sevenths, augmented or diminished
4:58
intervals, and leaps larger than an octave. Leaps are also typically recovered by stepwise motion in the opposite direction, as happens after both of Salieri's leaps. This rule is only important for large leaps (let's say leaps larger than a third), and it's more important for upward leaps than for leaping downward. I think of it like throwing a basketball in the air: what goes up must come
5:31
down. When we throw the melody up, we can give it an imaginary sense of weight by following the leap with a downward step that also makes the melody easier for the listener to predict. The Cantus firmus always begins and ends on scale degree 1: in this case middle C. All of our melodies are going to end on scale degree 1 (or "do" in solfege syllables) because that stable tone gives the melody its strongest sense of closure. The Cantus needs to begin on scale degree 1 because,
6:10
before we've written the other voices, we don't know if it's going to be used as a bassline, or as an inner voice, or maybe as the top voice. If it is going to serve as the bassline, then it needs to begin on scale degree 1 in order to define our sense of key at the beginning. We want this haiku-like composition to have a good beginning, middle, and end, and scale degree 1 can be used to
6:38
create a sense of stability for our beginning and our ending. Since our melody is going to begin and end in the same place on scale degree 1, we need an intermediary goal of melodic motion that's far away from scale degree 1 that will clearly separate the state of rest at the beginning from the state of rest at the end. For that reason, a melodically fluent melody has a high point somewhere in the middle that is a consonant interval away from scale degree 1. Salieri's high
7:15
point is A, which is a major sixth above C. The Cantus firmus now has a sense of goal directed motion in the first part as it ascends by leaps up to the high point, increasing its potential energy as it ascends, sort of like rolling a boulder up a hill. In the second half of the melody, that energy is released through fluent stepwise motion back down to scale degree 1. Here is a solution to this exercise written by Salieri's student: the fifteen-year-old Franz Schubert. The
7:56
new melody is called the counterpoint. The first thing to notice is that Schubert's counterpoint is also melodically fluent: it follows all of the same rules about steps and leaps that govern Salieri's given melody. The counterpoint only does two things that the given melody would not have done. First, it begins on scale degree 3. When a counterpoint is written above the given melody,
8:23
as we have here, it can begin on scale degrees 1 3 or 5. But when you are composing a counterpoint below the given melody, so that it has to serve as a bassline, then it must begin on scale degree 1. Notice also how Schubert's beginning note is the melody's high point, so that the overall shape of the counterpoint is to start on its high point and then gradually descend. But, it scoops a little too low and it has to approach scale degree 1 from below. The counterpoint therefore
9:00
has a very different overall shape compared to the given melody. One of the reasons we wanted our melodies to be so smooth and easy to follow is because the secret to a good counterpoint exercise is that the musical interest comes from the way that the two melodic curves interact, not from either of the melodies on their own. It's perfectly fine if both melodies are a little bit boring on their own, as long as they create interesting voice leading. Let's listen
9:32
to Schubert's counterpoint, first on its own, and then listen to the two melodies together.
10:06
The second thing to notice is Schubert's repeated Cs. It would be pointless for the given melody to repeat tones like this, but our counterpoint melody is allowed to repeat tones every once in a while, if there's a good reason. Here, Schubert's repeated Cs allow the listener to pay more attention to the given melody at the moment when it has it's interesting leap of a fourth up to the high point A. Too many repeated tones, though, will encourage the listener to only
10:37
pay attention to the voice that's moving, so we should keep tone repetitions to a minimum. There are two rules about how the given melody and the counterpoint can interact. The first rule is that the exercise can contain no dissonant harmonic intervals. A harmonic interval is simply the interval between the two voices, and there are three kinds. Unisons, octaves, and fifths are called perfect consonances. These intervals are the most stable, but they are actually a little
11:11
bit too stable. The voices appear to be glued together and they don't seem to have any need to continue on to the next sound. Perfect consonances put a drag on the forward flow of your melodies if they are overused, so, for that reason, you should avoid using perfect consonances twice in a row. Unisons are the strongest perfect consonance, and so they may only be used as the first or last interval of an exercise. Thirds, sixths, and tenths are called imperfect consonances.
11:49
Although the word "imperfect" makes these intervals sound like they're bad in some way, these are actually the most useful for our purposes. The two voices sound in agreement, but without sounding like they're stuck together. If the perfect consonances are like glue, then the imperfect consonances are like motor oil: they make it very easy for the two melodies to flow past one another smoothly, which is exactly what we're going for. Seconds, fourths,
12:25
and sevenths are dissonant harmonic intervals, as are any augmented or diminished intervals. Dissonances are unstable and they create a sense that the two voices are in disagreement with one another -- repelling one another, kind of like magnets of the same polarity. For that reason, dissonances are actually the best kind of interval for creating voice leading independence. The two melodies will sound very independent if they are able to clash with one another. But,
12:56
unfortunately, the first rule of one-to-one counterpoint states that we may not use dissonant harmonic intervals in these exercises. You have to unlock them by getting to the next level.
13:10
If we look at the intervals that Schubert used in his exercise, we'll see that all of the intervals are imperfect consonances (which, again, are the best for creating a sense of forward flow) all except for that final perfect octave, which is going to happen every time both melodies end on scale degree 1. Schubert also mixes up which kind of interval he uses to create a sense of variety. In order to understand the second rule of one-to-one counterpoint we need to define the
13:43
three possible ways that the two melodic lines can move relative to one another. The best kind of motion for creating a sense of independence is "contrary motion." Here, when one melody moves up, the other moves down, and vice versa. When the voices move in the same direction, they give the impression that they are moving together, and so this isn't as good for creating a sense of two independently moving melodies. We call this "direct motion." We can distinguish two
14:19
different kinds of direct motion. Parallel motion happens when the voices move in the same direction by the same interval. Here we have two sixths in a row, which we call parallel sixths, followed by a pair of parallel thirds. "Similar motion" is when the voices move in the same direction, but by a different interval. So here, for instance, we have the top voice descending by a leap of a fifth
14:47
while the lower voice also descends, but only by a step. Of the two kinds of direct motion, parallel motion is the one that we should avoid the most, since it's the least independent way that the two voices could move with respect to one another. If we look at this passage, for instance, and imagined taking Schubert's note D and replacing it instead with an A then the exercise could have had four consecutive parallel thirds. There wouldn't have been anything incorrect about
15:21
Schubert's counterpoint melody if he had done this, but the voice leading between the melodies would have become very boring and predictable. The two voices would have appeared glued together, not because they formed perfect consonances, but because of the way they moved in parallel motion. We should avoid long stretches of parallel intervals. The last kind of voice leading motion is called "oblique motion." This is when one melody moves in one direction or the other,
15:53
and the other melody remains where it is. This will only happen in our one-to-one exercises when we compose a repeated tone in our counterpoint. Oblique motion definitely creates a sense of voice leading independence, and so it is better than direct motion in that sense, but if we overuse it in our counterpoint voice then the other voice will be much more melodically interesting and the listener will stop paying equal attention between the two voices. So, again, only repeat tones
16:26
when you really feel you need to. I've said that contrary motion is the preferred kind of motion, but in reality there can be too much of a good thing. We want to have an unpredictable variety of motions in order to create the most interesting possible voice leading. Now we're ready to understand that the second rule that Schubert obeyed while composing his counterpoint melody: we may not use direct motion to approach a perfect consonance. This rule makes perfect sense, since
17:02
direct motion is the most stuck-together kind of voice leading motion, and perfect consonances are the most stuck-together kind of harmonic interval. So this direct motion to a fifth and this direct motion to an octave are going to sound kind of clunky and inelegant. Even worse would be parallel fifths or parallel octaves; the melodies would sound completely fused together. Looking at Schubert's exercise, the only perfect consonance that he includes is the octave at the end,
17:37
and it is safely approached by contrary motion. Schubert uses the sense of independence that comes from the contrary approach to counterbalance the sense of "stuck-togetheredness" of the perfect octave. The task of composing a counterpoint below a given melody is almost exactly the same as writing one above. Students for some reason assume this is harder, but it really isn't. Here is Schubert's exercise for Salieri's Cantus firmus. The
18:12
most important thing to remember is that the counterpoint must start on scale degree 1, and very often you will be aiming towards a low point instead of a high point. In Schubert's case, the low point is the note A. Schubert's exercise includes two sets of repeated tones that I find to be a little bit suboptimal. Now I hesitate to criticize Schubert, since, after all, he was taught counterpoint by a mediocre lunatic (*joke). I wonder if you, the viewer, can find a
18:46
solution that improves upon Schubert's. Composing an exercise in the minor mode also doesn't change very much. Here is a Cantus firmus that Salieri provided for Schubert that treats the note D as scale degree 1. Notice that the second-to-last note approaches scale degree 1 from scale degree 7 below, and that Salieri added an accidental to raise the note C to a C-sharp. This is called "raising the leading tone," and it is customary at the end of a cadence in the minor mode.
19:30
And here is Schubert's counterpoint melody.
20:08
First, we can see that Schubert also raises the leading tone throughout the exercise, although here it leads to a pretty significant voice leading problem. The leap from F to C-sharp is a dissonant diminished fourth, which most teachers would have considered an outright error, but Salieri apparently allowed it. Having listened to this excerpt, I think you'll agree that the leap doesn't sound unnatural or difficult, so maybe Salieri was right. Schubert could have
20:41
easily avoided the dissonant leap by leaving the middle pair of C's as C-naturals instead of C-sharps (and by the way you would need to change both of them, since a C followed by a C-sharp would be technically an Augmented unison, which is also a dissonant melodic interval).
21:02
And, for the sake of being thorough, here is Schubert's counterpoint melody written below Salieri's minor mode given melody:
21:32
Sound familiar? This is almost an identical counterpoint to the one that Schubert wrote above, except that here, he avoids the leap of a diminished fourth by composing in the note D. D wasn't an option earlier, because it would have formed a dissonant harmonic fourth. Now, I don't have any proof, but I suspect what happened here was that Schubert was trying to show off by composing a solution in what's called "invertible counterpoint." That means a counterpoint voice
22:07
that works just as well above a given melody as below it. Once Salieri overcame his profound jealousy and urge to murder the prodigy (*another joke), he made Schubert correct this C-sharp to a D, because there's really no justification for it when it happens in the bass voice.
22:32
Now let's walk through the process how to compose our own counterpoints against a given melody. This Cantus firmus was written by the Belgian musicologist Francois-Joseph Fetis. It's a little bit on the longer side at 15 notes, but it makes for an excellent teaching example. Let's listen.
23:12
The first piece of advice I can give is that you don't have to compose chronologically from the beginning to the end. We know that our counterpoint needs to end on scale degree 1, so we can start by putting that in. We also know that this final octave is a perfect interval, and so it has to be approached by contrary motion. The given melody descends by step to scale degree 1, so our counterpoint will have to ascend, preferably by step to scale degree 1. This ending,
23:47
with a 6th moving to an octave in contrary motion, is the most standard melodic cadence, and it's going to be how we finish all of our 1 to 1 counterpoint exercises. We have 3 choices about how to begin our exercise: we can start with scale degree 1, C, forming an octave; scale degree 5, G, forming a 5th; or scale degree 3, E, forming a 10th. We wouldn't want to start on the lower E, a third from the given note, because the voices would be too close together and we want to leave
24:24
some space for when the given melody ascends up to its high point. Starting on anything higher than a 10th would also not be ideal, since we don't want our melodies to be spaced too far apart either. Generally, we should think of our exercises as being written for human voice ranges, so in this case, the given melody is in the Alto range, and our counterpoint will be written for the Soprano range. The temptation is just to pick a starting note and try to compose one note at a
24:55
time. Instead, we should always try to compose in medium-sized chunks. My choice for starting note will depend on how I want to counterpoint this short stepwise segment of the given melody, and also on more global considerations that I'll mention in a moment. One possibility would be to start with octave C's, move in contrary motion to B, forming a sixth, and then move in parallel motion to C, forming another sixth. This is a very good way to begin, but I slightly prefer
25:28
this option. Here, we begin on G, forming a fifth, move in contrary motion to F, forming a third, and then in parallel motion back to G, forming another third. I like this option slightly better because we want our counterpoint to have a nice overall shape: ascending to a high point and descending back to scale degree 1, just like the given melody did. By starting on the lower G, we can accentuate that ascent up to a high point and create a more obvious goal-directed
26:02
shape. Speaking of high points, we should plan ahead to make sure that our high point doesn't happen at the same time as the high point of the given melody. If they happen at the same time, they would sort of "steal one another's thunder," and distract from one another. Also, their overall shapes would coincide, and they would lose some of their large-scale independence. I'm imagining a high point for our counterpoint that comes slightly later than the given melody's A,
26:34
something like this. Here we have a C, forming a third above the given Melody's high point A, moving in contrary motion to a fifth, D, above the given G, and again contrary motion to our counterpoint's high point E, as the given melody leaps down to scale degree 1, forming a spacious 10th. I like this solution quite a bit, because of its overall contrary motion, and because the two melodies both increase in energy up until this point, but they don't peak at exactly the same
27:10
time. This is the working method that I would suggest when you compose your own exercises. Begin by composing the ending, since it's going to be the same for all one-to-one exercises, and you don't want end up painting yourself into a corner and not be able to reach this cadence formula. Then pick a starting point that will allow both melodies room to develop. Next, compose a high point while you still have the freedom to place it wherever it will be most effective. Then,
27:44
all that's left is to fill in the gaps. Let's start by filling in this later gap. The contour of the given melody for this passage goes step up, leap up, step down, step down. We want to compose a counterpoint that fills in the gap from E to B while also complimenting the contour of the given melody. I came up with this: from the tenth, C to E, we move in contrary motion to the perfect octave, D to D, leap in contrary motion to a third, F to A, leap in contrary motion to
28:25
the sixth, E to C, and step in parallel motion to the sixth, D to B. We do have two leaps in a row, and the leap from D to A is not recovered by step in the opposite direction. But this is okay because it was a downward leap and the gap between D and A is filled in by the next few notes. The earlier gap has a given contour of leap down, leap up, step down, step down, and leap up. This
28:58
one's a little bit more challenging, and there are multiple good solutions. One option could be this:
29:06
Here we leap in contrary motion to octave Cs, step in contrary motion to a third, G to B, step in parallel motion to a third, F to A, step in contrary motion to a fifth, E to B, and step in similar motion to a third, A to C. This is a perfectly suitable counterpoint, but I think we can do better. First, the approach to the given melody's high point A by similar motion deserves to be accentuated by a stronger approach. I happen to know from experience that
29:44
the preceding fifth between E and B is somewhat weak: it implies the harmony of a III chord, which you will know if you've studied harmony is a somewhat ambiguous chord. Scale degree 3 and scale degree 7 are both very active scale degrees, but they're kind of slowed down when they're fused into a perfect fifth like this. Again, there's nothing technically wrong with this solution, but I prefer this. In this solution, we step in contrary motion to a sixth, C to A,
30:18
step in similar motion to a third, G to B, leap in contrary motion to a sixth, F to D, step in parallel motion to another sixth, E to C, and repeat the tone C, allowing the given melody's high point A to be approached by oblique motion, forming a third from A to C. When we step back and look at the entire counterpoint melody, we see that it has a coherent goal-directed motion from its starting tone to its high point, and a smooth release of
30:53
tension down to scale degree 1 for the ending. The two melodies form an interesting voice leading and are highly independent and contrasting. You can gather a lot of that information by looking at the exercise written out, but the ultimate test is, of course, whether these things are audible. Let's listen to the counterpoint melody, and then to both melodies combined.
31:58
There are three exercises that you can do as a kind of homework that will improve your understanding of voice leading in one-to-one counterpoint. The first is just to do what we've been studying and compose a counterpoint melody against a given Cantus firmus. Here are two Cantus firmi, one in major and one in minor. Compose four counterpoint melodies, one above and one below each of these Cantus firmi, following all of the rules and guidelines discussed earlier in the video.
32:33
The second skill you can develop is your auditory imagination through an audition exercise. Memorize the Cantus firmus so that you can play it back in your head easily without looking at the score. First, imagine it performed by your own voice singing in solfege syllables, then imagine the sound of it being performed by a cello, or some woodwind instrument. Try to hear it as specifically as you can, but don't sing it out loud or play it on an
33:04
instrument. Then go through the exercise that you've composed and try to hear both voices together in your head. I don't want you to just imagine any old perfect fifth, or minor third, or whatever the interval is, I want you to try to hear the specific combination of those scale degrees that make up that particular interval before moving on to include the next interval. Go back and forth between the first two intervals, making sure you can hear them as
33:34
precisely as you can, then start adding in more of the sounds into longer strings. Eventually, you will be able to play back the entire exercise in your head. Don't worry at all about rhythm just yet, we care only about hearing the combinations of scale degrees.
33:53
This exercise is not nearly as easy as it might seem at first, and it will get much more difficult as we progress into more advanced species of counterpoint and when we eventually add a third voice. Audiation was a skill that the classical masters developed very thoughtfully, and I think the reason we don't typically teach this kind of exercise in music schools is because there's no way to grade it: you're the only person that knows what's happening in your head, so you'll have to
34:22
grade yourself on this skill. Finally, the third exercise is to prepare analytic diagrams of your composition exercises. When I was in math class in high school my teachers always reminded me that it was important not just to get the right answer but also to show my work. Analytic diagrams are a way for you to "show your work," so to speak. If we were to prepare an analytic diagram of Schubert's
34:49
one-to-one exercise, there are only three steps. First, use a dashed slur to connect any repeated tones together. This is meant to draw attention to all of the instances of oblique motion in the exercise. Next we're going to use solid slurs to indicate all of the consonant leaps in our exercise. This is a good opportunity to double check that all of your leaps are indeed by a consonant interval. Visually, now, we can estimate the proportion of consonant leaps and
35:24
repeated tones to stepwise motions (which are all of the motions that have neither a solid or dashed slur). Finally, we're going to label the harmonic intervals between the voices. Double check that all of the perfect consonances are approached by either contrary or oblique motion. You can circle all the perfect consonances if that helps. The analytic diagrams for one-to-one counterpoint
35:51
will seem pretty easy, but that is because the compositional relationships are pretty simple at this point. By the time we get to multiple rhythmically free melodies written against the Cantus firmus, these analytic diagrams will become very important. That's all for this video. Stay tuned for the next one, when we'll be talking about two-to-one counterpoint
Here to help you learn
Sign in to keep a private conversation with your course tutor.
Ask me anything about the material…
Sign in to chat