That said, there is a way to know that you are listening to dorian and not some other relative modal area. After all, I've been saying that you can't just rely on the note names to tell you everything - you need to look at the relationship between one note and the next and the next and the next. As I said back on the first page, we need to establish a tonic. The way that we do that is with cadences and by harmonic progression. The diatonic modes, unfortunately, make this difficult, so we'll go with the major scale for the moment. As you may be aware, there are two half-steps in any major scale (or any other diatonic scale, for that matter): one between steps 3 and 4, and another between steps 7 and 1. As it turns out, those half steps are involved in important harmonic tendencies. 7, or the "leading tone" wants to ascend by half step to arrive at the tonic. 4, or the "subdominant" wants to descend by half step to arrive at 3, or the "mediant". If you put together the tendency tones (4 and 7) in a diatonic scale, their intervallic distance is a tritone, which is a dissonant interval. When these tendency tones go where they want to go (3 and 1), then the dissonance is resolved, leaving a consonance. And you can either spell that tritone as an augmented fourth (leading tone on top) which expands to a minor sixth, or as a diminished fifth (leading tone on bottom), which collapses to a major third. You might notice that this resolution gives us what is basically the makings of a tonic triad (1 3 5), minus the fifth of the chord (1 3). This is why the major scale is quantitatively easier to deal with than the other diatonic modes: the tritone is already in there, pushing our ear toward the resolution to the tonic. Even if the tritone isn't so explicit, if you're running up and down any mode of the major scale, your ear is going to gravitate toward the relative major key area. It's the tritone's fault. The tonic is always going to be our point of resolution, so in a single voice texture, things won't sound resolved until get to 1. Let's have a look at all the ways that we can approach 1 with diatonic intervals. I've laid out approaches from above and below by second, third, fourth, and fifth. Sixths and sevenths are not included because they're too wide of an interval. You can try, but it doesn't sound right (even though Bach does it, but even he approaches it as an octave displacement). You see our friend the leading tone in the first measure. Stepwise motion is always good, and the half step resolution is all the nicer. The second measure is approaching from a second above. 2 is called the "supertonic", because it's above the tonic. Woo, Latin. Next is 5 1 from above - the dominant to tonic. This is a very harmonically stable interval, and our ear likes hearing it. We can also approach it from below, where the interval becomes a perfect fourth - 5 to 1 is still dominant to tonic, no matter how you flip it. Let's take a moment to look at what we have so far - 5, 7, and 2, all going to 1. If you play 5, 7, and 2 at the same time, you get A C# E, which is a major triad. More specifically, it's V. You'll often hear about the dominant-tonic relationship, and you might know that V-I is the fundamental chord progression. Consider that V consists entirely of tones that want to go to the tonic, either because of stepwise motion or something to do with the overtone series, and that 7 is the strongest tendency in there. Proceeding, we get to relationships that are not quite as strong as those found in the previous measures. We have 4 1, both from above and below, and that might be involved in a plagal cadence, but 4 3 is more natural, because of that tritone resolution thing. 3 1 works quite well actually, but those are both members of the same chord (in other words, no harmonic motion is happening in 3 1), so it doesn't possess the same urgency as one of the members of V proceeding to 1. The last one, approaching the tonic from a third below, is not a common approach, because 6 1 seems to outline a vi or IV triad more than anything else. Not an effective cadential figure. When you add 4 (G) on top of 5, 7, and 2, (A C# E) you get V7 (A C# E G). The seventh chord is a lot more dissonant than the triad, because it possesses that tritone that so firmly defines tonality. What we're looking for in order to establish tonality is the V I relationship, preferably with V7. But, there's a problem: the major mode is the only one that has this built in. One of the ways that we cope with this issue is by making temporary alterations to our pitch palette. This is the most famous example of getting around the tritone's tendency to pull us into the relative major key. First, we start with the natural minor scale on D, D E F G A B♭ C D (or 1 2 ♭3 4 5 ♭6 ♭7, if you want key neutrality). There is a tritone in there, between E and B♭. That's cool, but it's between 2 and ♭6. That's not going to get us to 1 and 3 very easily, we really need a tritone between 4 and 7. As it stands, that D minor scale risks becoming an F major scale. What we can do to fix this is raise that C (♭7) to C# (7), which will give us that 4 7 tritone we wanted. This gives us the half step between the leading tone and tonic, but a whole step between 4 and ♭3. The ♭3 is something we can live with, because resolution to 1 is the more important part. If ever you wondered why harmonic minor is called harmonic minor, wonder no further: harmonic minor facilitates the harmonic progression of V7 i. In this way, you can get V7 in a mode that does not normally have it. One thing that troubled the guys that were founding these practices was the augmented second interval between ♭6 and 7. When played in order, 1 2 ♭3 4 5 ♭6 7 1 sounds, to be blunt, Middle Eastern. Western music does not typically contain augmented seconds prior to 1900, and there were those Crusades things, so to further the distance between Christian Europe and the Muslim world, we dropped the augmented seconds. We still liked the leading tone that harmonic minor made for us, though, so raising the sixth degree of the minor scale to turn that augmented second into a major second was a good compromise. This gives us the melodic minor (1 2 ♭3 4 5 6 7 1). Melodic minor is used whenever you want scalar motion and have something going on between 7 and 6. 5 6 7 1 is common, as is 6 7 1, and 7 6 7 1 (where 6 is a neighbor tone), and so on. Also, rarely something like 1 7 6 5. In any case, the melodic minor is not meant to have harmonic implication. 6 is usually a melodic non-chord tone. Hence, you know, being melodic minor and not harmonic minor. The next bit of information that you need to know is how chord progressions work. Our basic goal is to go from V to I, so we can abstract any functional progression as [whatever] -> V -> I. It gets more delicate than that, of course, but that's the gist of it. If you want a more methodical approach, it's subdominant functions (IV and ii) -> dominant functions (V and vii°) -> tonic functions (I, occasionally vi for deceptive cadences). Notice that I am using the numerals for the major mode. If you want to do minor or any other mode, different things are capitalized or lower case. ======== Excuse the long preamble. We have to have some idea of the problems we encounter in other modes before revisiting our analysis. But now that we've done that, look at what's going on here: (I am extrapolating the chord progression by the notes that we hear, as well as what sounds correct to my ear. For the most part, the melody outlines a triad in every measure. The only ambiguous part was measures 9 and 13, where we might have Bm, but I decided that G sounded more correct.) We know that this is somehow related to E minor, because we hear a very clear outline of the Em triad at the beginning, and we have harmonic reinforcement of Em as the tonic. The first two chords, Em D, do not give us a lot to go on, but as soon as we get to the end of the antecedent phrase, we have D# a couple of times, which is the leading tone of E. The last chord of this phrase is V, meaning that the antecedent ends in a half cadence. The consequent phrase starts the same as the antecedent, making the first eight measures a parallel period. Fancy words. For our purposes, this is reinforcement of the harmonic context, since it's the exact same thing that we just heard. Measures 7 and 8 are where we get a cadential figure in E minor. That i6,4 V pairing is the dead giveaway. We call that the "cadential 6/4". I6,4 (or i6,4, in our case) acts as a prolongation of the dominant harmony, and is a signal that we are approaching a cadence. You can hear a little more on that from this video. He uses "c" to indicate a second inversion triad. I prefer the figured bass symbol. Look over the next phrase really quick. Do you notice that it, too, is a parallel period? Both measure 9 and 13 start the same way. The way that these phrases start suggest a visit to the key of G major, because we have what could be I V in G, but since there's no V I, it never takes off. Instead, we know that we are still in E minor because we get a few instances of V i in E. Also check out measures 3-4 & 11-12, then 7-8 & 15-16. Same thing, huh? Tight little tune. To address the dorian parts of the song, notice that the sixth scale degree, whether it be 6 or ♭6, is never part of the functional harmony. We see C (♭6) first as a neighbor tone to B in measure 1, and between that and the next measure, that suggests natural minor (1 ♭3 4 5 ♭6 in mm.1, 1 2 4 ♭7 in mm.2). We see C# (6) in measure 7 as a neighbor tone to D#, and C# D# E should spell out E melodic minor to you. We get C# again in measure 9, this time as a passing tone between D and B. The presence of D and C# is what gives us the E dorian feel. Leading tones are great for establishing tonality, but we do not always want to use leading tones. In fact, that's the main problem we deal with in modal composition, so we have to find ways to cadence without tritone resolution. Koji Kondo - Song Of Time Another piece from Ocarina of Time. There are a number of dorian pieces from that game, owing to the compositional limitations that Koji Kondo had to conform to. Anyway, let's look at what's going on here. There is a lot of arpeggiated Dm, so much so that there is almost nothing else. We have to assume that D is the tonic by default. He does use that 5 1 relationship from the get go, so there is something resembling a V i progression. The dorian part comes in on the downbeat of measure 3. The modal color tones always come in after some amount of preparation. The cadence at the end is what we want now: C E D, no raised seventh, no half steps. What we do have is stepwise motion from both directions, surrounding the finalis. A lot of what I see in these modal pieces is this pattern: use the tonic triad profusely at the beginning, have very simple harmony that primarily serves to expand the tonic, use the color tones as non-chord tones, rely on degrees 5, 7, and 2 (altered or unaltered) to approach the tonic at cadences. I've written a couple short melodies with this scheme, and it seems to work so far. This is probably better for short forms. As soon as you get to longer music, the ear yearns for harmonic variety. But, as they say, start small.