What do a drunken sailor and an old English fair have in common? Well, of course, both are the subject of traditional Anglo-Saxon folk songs that made a successful entrance in contemporary rock music. But there is more to their family resemblance, as both these tunes are based on the Dorian mode. That is why these songs have the feel of shifting between two keys. Introducing the concept of the Dorian twin tone system, Ger Tillekens here discusses the family traits of "Drunken Sailor" and "Scarborough Fair".
Playing with the key. Popular music, as Peter van der Merwe (1989) argues in his book Origins of the popular style, first and for all is modal music. By saying that, he alerts us to the fact, that many popular songs tend to play with their key. In contemporary pop and rock songs this is mostly done by mixing in chords derived from relative and parallel keys. For instance in a song in C Major, one can often find that the dominant chord (G) is traded for its relative Minor (Em) or its parallel Minor (Gm). The first of these stand-in chords is the regular dominant of the relative key of A Minor, the last one that of the parallel key of C Minor. This substitution of chords, however, does not imply that the key of each and every pop song really modulates or shifts to these sister keys. There are hardly any solid rock progressions without at least some incidental chords inserted between the basic ones, and almost all songs contain short parts, leaning towards a related key — especially in the middle eight. The average song, though, using the musical idiom of pop and rock music, firmly holds to one and the same key.
Primary and secondary chords. Scales with a fixed key will provide the composer or performing artist with only a restricted array of tone material. Overall, the seven notes of the diatonic scales — the series of consecutive white keys on the keyboard or one of its transpositions — are just good for forming three consonant triads, the so-called basic chords: the tonic, the dominant and the subdominant (Tillekens, 1999). Mixing in relative and parallel chords has the advantage of extending the available tone material of a composition or accentuating interesting intervals, which will provide the characteristic "unexpected notes" of popular music (Becker, 1994). Pop musicologists tend to see these chords as "borrowed" from the related key. That is why they call them "secondary" chords and — accordingly — the basic chords the primary ones. Nowadays composers and performers alike have grown used to this kind of chord replacement. In the world of rock music it has become a rule of thumb for guitarists and songwriters alike.
A songwriter's advice. In a song in a Major key even relative and parallel Minor chords themselves can be traded for their parallel and relative Major counterparts. Andy Widders-Ellis and Jesse Gress (1994: 99), for instance, offer their readers the following songwriter's advice in the Guitar Player magazine:
"Want to lift or brighten a progression? Change one or more of the diatonic Minor chords (ii, iii vi) to Majors or dominant sevenths. In C, for example compare the diatonic C -» Em -» Am -» Dm -» G7 -» C (a I -» iii -» vi -» ii -» V -» I progression) to a modified C -» E7 -» A7 -» D7 -» G7 -» C."
Of course these secondary chords will have to fit in with the chord progression and the melody line, or as Widders-Ellis and Jesse Gress conclude, with the "line's momentum" which, in a pop and rock song, they add to their statement, is always stronger than the "dictates of a key centre." In short it all depends on the power of the specific chord progression and the interweaving melody line of a song to glue things together.
A family of songs. It works the other way around as well: the chords of Minor keys can be replaced by their relative and parallel Major counterparts. The easy way in which these chords go together, has a simple reason: relative and parallel keys are closely related, as Major and relative Minor scales share the same tone material and Major and parallel Minor scales start from the same key centre. Therefore a modulation — a real change of key — can easily be avoided. Most pop songs freely exploit the possibilities of these related keys. Nowadays most people have become so used to it, that they will barely experience the feeling of a key shift. There is, however, a family of songs, where a key shift, or something resembling it, is more noticeable. One can hear the effect — a swaying, pendular movement — clearly in many hit songs, ranging from the Moody Blues' surprising "Nights In White Satin" (1967), by way of John Lennon's bitter "Working Class Hero" (1970) to the recent, melancholic songs of British pop singer Dido.
Listen to thirty seconds of Justin Hayward's composition "Nights In White Satin" as performed by the Moody Blues and the London Festival Orchestra on the album Days of Future Passed (1967)" as "The Night"
Two Dorian prototypes. Songs like "Nights In White Satin" are special because they bear the marks of the Dorian mode, a modal song type combining two scales — a Minor and a Major one — with the same notes, but with different tonics which are lying only one, whole tone step apart — like, for instance, the keys of E Dorian and D Major we hear in "Nights In White Satin".  Some of these songs are outright Dorian, many others have Dorian inflections — said otherwise, Dorian parts built into them. The song family's genealogy is quite impressive. The roots of its family tree go way back to the times of the old Celts and one can find many historical examples of Dorian tunes in the long list of well-known traditionals.  Two songs, though, tend to pop up as standard prototypes in almost any discussion of the Dorian mode. One is the shanty — or, as it sometimes used to be misspelled in the old days, the "chanty" — "Drunken Sailor", the other that all time classic of folk music "Scarborough Fair" that became widely known in the version of Paul Simon and Art Garfunkel on their album "Parsley, Sage, Rosemary, and Thyme" (1966). In his study, Van der Merwe (1989) foregoes both these traditionals. Making up for that omission, we will give them a thorough scan here. We will start with "Drunken Sailor" — overall a rather weak member of its Dorian family, though showing, as we will see, one very strong feature of its kinship.
In 1966 the Belgian skiffle, folk and blues singer Ferre Grignard (photo above) made the shanty "Drunken Sailor" into an unexpected hit. Listen to some thirty seconds of his version of the song from his album Ring-ring (1966)
What shall we do ... "Drunken Sailor" or "Early In The Morning" as some people are used to call the song, is old but again not that old. In fact, shanties — working songs at sea — as we know them today were developed during the 19th century when the sailing ships grew bigger and bigger and needed teamwork for heavy tasks like hoisting yards up a mast or heaving the anchor. Here, the shanties came in. With all sailors roaring out an appropriate song in unison, it was much more easy for them to coordinate their repetitive and heavy duties. The songs were popular with the ship mates, but because of their bawdy lyrics frowned upon and sometimes even forbidden by their superiors. In his book Capstan bars (1931; quoted by Ward, 2002) David Bone (1874-1959), who started his long seafaring career in 1890 on windjammers in Australia, dates this particular song back to the late 19th century:
"... "Early In The Morning" must be of fairly recent date, for only in a comparatively large ship could there be room on deck for "walking" a light sail aloft, the operation at which it was generally used. It was not a chanty often sung. I remember it chiefly as a showy accomplishment when all hands were employed on deck and there was an atmosphere of high good humour with us. Perhaps a day when, after long voyaging, we had taken a pilot aboard to con us into port and were wishful to impress him by our activity as real deep-watermen. We called it "man-o'-war fashion" as we stamped along the deck, handing the halyard."
A whaling voyage. Against Bone's contention, others however maintain, that "Drunken Sailor" is an older, even one of the oldest known Anglo-Saxon shanties. They are quite right, as the song — with music — is mentioned in yet an earlier book: Incidents of a whaling voyage by Francis Olmstead, dating back to 1839. Chris Robson (2001), who informs us of this fact, also mentions that the song's use was not restricted to larger ships. "Drunken Sailor", for instance, was one of the very few work songs permitted on the smaller ships of the "King's Navee" and it also could be heard on merchant ships with small crews, though here its usage was limited to the job of going "'bout ship" when the braces would be manned and stamped away with, or when hoisting light sails hand over hand. In this case only the chorus — "Way-a hay-a, up she rises" — would be sung. The song disappeared from the sailors' song repertoire with the coming of the steam ships, but was kept alive in the harbour pubs, reemerged in the days of scouting as a standard marching song and, later on, developed into a popular subject for community singing with the skiffle and folk movement of the 1950s. The tune itself even has a longer history, as it is derived from a traditional Irish air, a dancing or marching song, dating from Medieval times.
Melody line and chords. That is enough naval history for now. Let's take a look at the song itself for the marks of the Dorian song family (see figure 1 below).
Figure 1: Melody line and chords to Drunken Sailor
Dorian dispositions. The key we have chosen for our example is A Dorian, which for this song is the favourite of guitar players by far. Next to the keys of D Dorian, E Dorian and F-sharp Dorian, this still proves a very popular key for accompanying straight Dorian songs or songs with Dorian inclinations on the guitar. The tone material of Dorian songs is usually rather restricted. This also goes for the tune of "Drunken Sailor", which stays within the range of just an octave: A - A'. Next we see that the melody itself starts on the fifth step of the scale, as, by the way, many other Dorian songs do. More important, however, is the fact that, though the key centre clearly is A, the key signature itself points directly at the key of G Major, as the tone material — including an F-sharp — conforms to this key. This combination of the tone material of G Major with the key centre of A is the reason why it is commonly called the mode of A Dorian. For the same reason the sixth tone step, in this case the F-sharp, is seen as the characteristic note of the Dorian mode. Knowing this, it is now time for a deeper dive into the muddy waters of modes, scales and tonics. What is a mode, actually, and why do we call it by that name?
Why do we call a mode a mode? In music theory there are seven diatonic modes, all called by Greek names. The qualification mode as well as the Greek names are derived from the so-called "ethnic modes" of the "Glarean system", as published in 1547 by the Swiss humanist and scholar Heinrich Loris in his book Dodecachordon. In his time Henricus Glareanus, as he was called by his Latin name, propounded a new theory of twelve ecclesiastical or church modes, which in turn were named after what probably were old tuning — "harmoniai" — and teaching practices of the seven-string Greek lyre. The history of these modes is rather complicated and still under research — see for instance: Nagy, 1980 (chapter 3). Luckily, nor Glareanus nor Greek music history won't have to bother us here. The work of Glareanus itself was quickly outdated by the rise of the Major-Minor-system and the Greek names, nowadays, are just treated as convenient labels for discerning the seven diatonic scales. These are the scales we can form by starting on an arbitrary diatonic note and filling the octave with the consecutive notes on the white keys of the keyboard. The steps between these notes are different, and consequently each scale will have its own characteristic tone steps. 
The Dorian scale. Starting the diatonic scale on D, resulting in the series D -» E -» F -» G -» A -» B -» C -» D, will provide us with the standard tone steps of the Dorian scale: 1, ½, 1, 1, 1, ½, 1. Transposed to the key of A, the Dorian scale will look like the series of notes and tone steps on the third row, counted top-down, of figure 2 below:
Figure 2: Four related scales, top-down: C Major, A Minor (natural), A Dorian and G Major
Now, is A Dorian a mode or a scale? Of course, the answer to this question depends on how we define a scale. If we describe scales as restricted sets of notes with one clear tonic that locates the key, then A Dorian certainly is a scale. The label "mode" usually refers to a scale — or rather a combination of closely related scales — when next to the tonic one or more other notes play an important role — sometimes even as a cotonic — in structuring a tune. For songs with Dorian scales this is almost always the case, because the Dorian scale easily relates to some other scales that can and will provide their keys as cotonics. So the label "Dorian" also applies to a mode, be it that we have to look for the interplay of two scales.
Major-Minor relatives and Dorian twins. Figure 2 not only shows the scale of A Dorian, but yet another three related scales. They come in pairs. The two scales of C Major and A Minor — more specifically natural A Minor, which by the way coincides with the Aeolian scale — shown at the top, are paired as Major-Minor relatives. As we can readily see, both these scales, though starting from a different key, share the same tone material. At the bottom we find the pair of A Dorian and G Major, also having the same notes. As there is no official name for this relationship yet, for our convenience, we will here call these last two scales "Dorian twins".  Both pairs of scales are connected by the close relationship between A Dorian and natural A Minor. We notice just a slight, but significant difference between A Minor and A Dorian. While the usual scale of A Minor has an F for its sixth step, A Dorian instead comes up with an F-sharp. This slight mutation changes the intervals between the fifth and sixth as well as between the sixth and seventh step of the scale. The implication is that the subdominant chord of A Dorian is the chord of D Major (D-F-sharp-A) instead of D Minor (D-F-A). The other regular chords of natural A Minor, however, go together rather well with A Dorian, as both scales have exactly the same tonic (A Minor) and also share the same dominant (E Minor). So all four scales are connected and their close relationship makes it easy to step back and forth between them within the confines of one song. And, indeed, many Dorian oriented songs do show a mix of these scales and their respective chords.
Close neighbours and distant friends. Alternating between the chords of A Minor and G Major, the tune of "Drunken Sailor" shows a blend of the scales of A Dorian and G Major. There is something special about such combinations of Dorian twins. Joining the scales of C Major and natural A Minor on the one hand or mixing A Dorian and G Major on the other, are quite different things. Though melodically the key centres of G and A — separated by only one whole step — may be really close neighbours, harmonically they are only very distant friends. In reverse, Major-Minor relatives like C Major and A Minor are melodically separated, but harmonically connected. Harmony is primarily based on rather large intervals of thirds and fifths. The chord of A Minor, for instance, is built out of the flat-third interval A -» C and the fifth interval A -» E; whereas G Major in turn is formed by the third interval G -» B and the fifth interval G -» D. The harmonic relationship between these chords is based on these same intervals.
A harmonic tone grid. Figure 3 below shows the double tonics of "Drunken Sailor", with their adjacent notes, in a harmonic tone grid of fifths (horizontal) and thirds (vertical).
Figure 3: The twin tonics of Drunken Sailor and their conjoining notes
As we can see, we can relate the tonic of G Major harmonically in two ways to the tonic of A Dorian. The first one, labelled as "uncensored", is placed at the subdominant side of A, and separated from it by two intervals of a fifth: G -» D -» A. The second one, labelled as "censored" and lying at the dominant side, is even farther removed from A in the tone grid by a distance of two fifths and a minor third: A -» E -» B -» G. We will come back to these variations later on. For the moment, it is just important to notice that in both cases the tonics of G Major and A Dorian are lying far apart in the tone grid.
Jumping the tone grid. Relative Major-Minor pairs of chords, like A Minor (A-C-E) and C Major (C-E-G), share two tones. This makes it rather easy to keep to the original key if we shift from the one to the other. That is also the main reason why they can take each other's place in a chord progression (Kramarz, 1983). So, we can say, that Major-Minor relatives — though separated melodically — are closely related harmonically. With Dorian twins, however, things work out just the other way around. Whereas the chords of A Minor and C Major have two tones in common, the constituent tones of the A Minor and G Major chords are separated by an interval of at least a fifth. So, by alternating between A Minor and G Major, the sequence of "Drunken Sailor" will jump back and forth across the harmonic tone grid. In fact, it is this distance that separates both tonics of the tune, which causes the effect of a fluctuating wave, rolling back and forth, which in turn explains why the song is so fit for accompanying coordinated work. So we see, that it is not the Dorian scale itself, nor even the melodic relationship between the Dorian twin scales, which are responsible for the actual effects of "Drunken Sailor", as well as the harmonic distance between their tonics. The melodic aspects are not irrelevant, though, as the harmonic switch from A Minor to G Major finds a strong compensation in the melodic proximity of their constituent tones .
Tonal technicalities. The harmony of "Drunken Sailor" oscillates between an A Minor and a G Major chord, but does this also mean that the key really shifts between the respective tone centres of A and G? Well, that actually depends on how you interpret the song. If you lean heavily on the sounds of an equally tempered instrument like the keyboard, chances are you will be inclined to experience a firm key shift, because there is some difference in pitch between the G note sung pure, and the G note you play on the keyboard. The human voice — as well as instruments like the violin — can adjust pitches continuously and change according to the needs of the key being played. Taking the pitch of A for a starting point, the pure G note sounds considerably higher than the note of the same name on the keyboard. To measure differences like these, we commonly use the unit of a "cent". By definition, on an equally tempered scale there are 100 cents in a semitone interval.
Deviating pitches. Using this unit of measurement, the deviations of equivalent pure and equally tempered notes will deviate by 1.955 cent for the intervals of the fifth — i.e. from A to E — and the quart — i.e. from A to D. The pure fifth will come out at a pitch that is higher and the pure quart at a pitch that is lower by the same quantity. The deviation is even larger for the intervals of a major third — i.e. from A to C-sharp — and that of a minor third — i.e. from F to A. The minor third will come out at a pitch that is 13.686 cent higher and the major third at a pitch that is lower by the same amount (see figure 4).
Figure 4: Deviations in pitch of pure tones for the intervals of the minor third (F -» A), major third (A -» C-sharp), quart (A -» D), and the fifth (A -» E)
Deviations adding up. Equally tempered quarts or fifths, even major or minor thirds, do make not so much difference compared with their pure equivalents. The trained ear can perceive them, but usually the human hearing perception will adjust almost automatically. The deviations in pitch, however, all add up. So instead of the 200 cent of the equally tempered keyboard, the pure interval between A and the "censored" G will amount to (200 + 1.955 + 1.955 + 13.686 =) 217.596 cent. Again that is not yet unsurmountable for the human ear. But it will be getting bigger if you progress from this new root to other notes. For instance if you proceed from the "censored" G yet a fifth upwards to D, the difference in pitch will increase with yet another 1.955 cent. Now the deviation in pitch amounts to 19.551 cent and will threaten to sound really out of key. The problem is aggravated by the fact that there also is another D in the tone grid that harmonically is more related to the key of A Dorian, i.e. the root of its subdominant chord D Major. Being the quart of A, this D comes out at no less than 21.5 cent in pitch below its namesake. Though — as it is called — "enharmonically" equal, both these notes really do sound differently. Luckily, in "Drunken Sailor" this is not really a problem, as the D note that plays such a dominant role in the third measure of the song's melody line, is unequivocally bound to the tonic of G Major — whichever one we choose — and as such defined in pitch as its fifth.
Enharmonic notes. The important thing we learn from this, is that in modal songs the interpretation of the pitch of the notes depends on the interplay of at least two keys. That is why Van der Merwe (1989: 208-209) calls this kind of song by the name of "double-tonic tune". In fact, he argues, songs like these are built upon two "foundation notes", each defining the actual pitch of the conjoining notes — like the quart, the fifth, the major and the minor third — in the tone grid. In case of the Dorian mode, with its distanced tonics, there will almost always be some enharmonic notes in the melody line, whose pitch will differ for each of the "foundation notes". For "Drunken Sailor" — if you choose for an "uncensored" interpretation of the song — these are the C and E notes, which in the song occur in relation to both tonics. When you sing or play these notes, you will have to know to which tonic they refer to get them right. Do they belong to the harmonic structure of A Dorian or that of G Major? In case you prefer a "censored" interpretation, the problem is not too big, as the enharmonic notes here concern both D and F-sharp, which in this particular song are equivocally bound to the tonics of A and G respectively.
Hinging on a central note. Still, we will have some difficulty in balancing the tonics themselves. Here, the song lines will offer a helping hand. If we look at the "censored" interpretation — which, by the way, for this reason seems the most fitting for this song — we see that both C and E appear in the tune under the chords of A Minor and G Major. Having the same pitch in relation to both tonics, these shared tones can be used to stabilize both tonics in respect to each other. In this case the E, which starts the tune, comes ready at hand. Van der Merwe (1989) calls such a stabilizing note on which the pitches of both tonics are hinging, a "central note". Knowing all these tonal technicalities, now at last we can answer our question if the Dorian mode implies a real key shift.
Shifts of level. The style of Classical and Romantic Music stresses key contrast and modulation. Here the shift to another key is done by "grounding" each transition in a new starting point. If the scales' tonics are far apart, the new key even is accentuated and resolved to the actual pitch played on the keyboard. If you do so when playing "Drunken Sailor", you will notice a clear key shift. However, in the styles of Folk Music and Popular Music key contrast plays a far lesser role. If we accentuate the central note of a Dorian song from the very start, the song lines will almost automatically steer us towards the right pitches and we will only perceive a pleasing, slight pendular motion. We don't even need to lean on the accompanying chords in interpreting the notes of the melody line, as the central note will keep us on the right track. In fact, traditionals like "Drunken Sailor", Van der Merwe (1989) argues, are not based on the harmonies as we know them nowadays. Instead songs like "Drunken Sailor" have two or more "levels," each of which is "firmly based on a single note," the "foundation note". So, as long as we keep these foundation notes in mind for a reference, we will not experience any modulation, not even a harmonic shift, but rather a shift of level.
Melodic effects. Violating the rules of Classical Music, the Dorian mode not only let us enjoy its pendular harmonic movement, but it also provides some nice melodic effects. Again we can use "Drunken Sailor" for an example. In the "uncensored", subdominant oriented, interpretation of the song — depending on the central note of D — the notes C and E appear in relation to the grounding notes of G and A respectively. In the second measure, for instance, the C note belongs to the tonic A, whereas it has G for its grounding note in the fourth measure. The difference between these notes amounts to 21.5 cent. The preferred "censored", dominant oriented, interpretation of the song leaves no room for such enharmonic notes. Here, however, we find some interesting intervals — look at the rising line E -» F-sharp -» G -» A in the sixth measure. The pitches of these notes are all defined by the grounding note of A, which in the blues interpretation of the song in turn depends on the central note of E. Starting from E, the F-sharp belonging to A comes out 17.596 cent lower and the pure G at 15.641 cent higher. The interval F-sharp -» G amounts to 133.237 cent — in short, it is about a third larger than the corresponding semitone interval on an equally tempered instrument. This, of course, constitutes an important part of the song's charms.
A weak family member. "Drunken Sailor", no doubt, bears the marks of its Dorian family background. One feature is even very strong: the pendular movement, oscillating between the two grounding notes, represented by the tonics of the scales of A Dorian and G Major. Another mark is the tone material including the sharpened sixth step of the scale, F-sharp. The song, however, does not exploit the possibilities of the Dorian mode to the full. Because of its restricted array of chords, we can even raise some doubts about its Dorian nature. If we take the F-sharp in the melody line as a quite pleasant, but only ornamental note — a "passing note" — we can interpret the harmonic structure of the song as easily as a trade-in of Minor chords for relative Major chords — with G Major as a stand-in for E Minor, the regular dominant chord of the key of natural A Minor — or better yet, vice versa — with the A Minor chord as a stand-in for the chord of C Major. To see a more elaborated example of the Dorian mode we will have to leave ship to visit "Scarborough Fair".
At the right: the 12th century castle of Scarborough, North Yorkshire
Going to Scarborough Fair. Just like "Drunken Sailor", the traditional "Scarborough Fair" figures in many discussions as a prototypical Dorian song. It really is a Dorian song and also a real traditional as it can be dated back unambiguously to the Late Middle Ages or rather the Renaissance period, the time of growing cities and flowering mercantile ports, both in the Mediterranean area and along the North Sea shores. One of those cities was Scarborough, lying at the coast of North Yorkshire — a county in the northwest of England. In those days Scarborough became an important and prosperous harbour city, famous for its yearly fair — a big forty-five day trading event in August and September — to which the song title refers. The lyrics tell the story of a man whose lover has left him, and now asks the listener to bring her a message. It is a rather sad song, because the man seems to realize that he is asking far too much of her.
Different interpretations. Going through the verses, the lyrics add up to a long list of virtually impossible tasks, the singer wants his lover to fulfill to prove the love she feels for him. By this he recognizes how difficult it will be for her to stay faithful to him, while at the same time he clearly is doubting if she can live up to his expectations. There is no chorus, but the song's continuity is upheld by a returning phrase in the second line of each verse. Again and again, here, a whole collection of herbs is summed up: parsley, sage, rosemary and thyme. Looking for the song's origin, some historians refer to the Plague. They interpret the lyrics as the dramatic story of a sick and dying man, banned from the city because of his illness. The herbs they see as the symbols for fighting off the contamination of the Black Death. This would date the song back to the years 1348-1350. Others also point at the late Middle Ages, but provide a more romantic account as they see the lyrics as a complaint of a villager whose lover left him, looking for work in the floundering and prosperous big city. In this case the references to parsley, sage and thyme are taken as symbolically representing the powers of these herbs to counter bitterness (parsley), to offer strength (sage), to imbue a loving remembrance (rosemary), and to give courage (thyme) (De Jong, 1999).
Musical characteristics. It is difficult to decide which of these interpretations is the right one, but that question does not have to bother us here, as we are mainly interested in the musical characteristics of the song. Let us take a look at the melody line and the chord material, as shown in figure 5 below.
Listen to some thirty seconds of "Scarborough Fair" by Bert Jansch and the Pentangle as performed on their album "One More Road" (1993)
Figure 5: Melody line and chords to Scarborough Fair
Some more Dorian predilections. The bars above show a simple transcription of the song's melody and harmony. Again the tone material is rather restricted, spanning just over an octave G - A' and again A Dorian is the key in which the song is mostly played. There are some other traits one often encounters in Dorian songs. The melody starts on A in measure 1, but again gives an important role to the fifth E, appearing in measure 2. Then in measure 3 the melody proceeds to B. Series of consecutive fifths in the melody line, like these, clearly add to the feeling of a key shift. The same goes for the characteristic drone, caused by the resonating open strings in the finger settings of the A Minor chord, which is one of the reasons why the key of A is a guitarist's favourite for Dorian songs. The song is in three-quarter time, and can easily be played on a guitar with one or more of the finger-picking patterns shown in figure 6 below. Most versions of "Scarborough Fair" indeed show this sort of "arpeggio's" — harmonically related notes played separately while accentuating the melody line. It adds to the feel of "authenticity", as the song, of course, originally was accompanied by instruments like the dulcimer and the Celtic harp, fit for playing arpeggiated melody lines. It also underscores Van der Merwe's notion that songs like these are built upon shifting "foundation notes" instead of modern full harmonies. Both the consecutive fifths and arpeggiated chords are things we often find in Dorian songs.
Figure 6: Finger-picking patterns
The British folk revival. When you listen to the version of Simon and Garfunkel, you will notice some differences with the arrangement we have shown. For one thing, Simon plays the song in the key of E Dorian. However, a capo is set on the seventh fret of his guitar, so the finger settings still conform to those of A Dorian. They also use another arrangement, originally made by British folk singer Martin Carthy. There is a gossip story attached to this version, worth telling. Paul Simon learned the song from Carthy in the early 1960s during his stay in London, but pirated the arrangement, giving no credits to Carthy (Harper, 2000; Torvund, 2002). Only recently things were set right, clearing the atmosphere for a musical reconciliation, which took place at Simon's concert at London's Hammersmith Apollo in October 2000. Carthy's arrangement became very popular in the wake of the British folk revival and clearly shows some leanings toward the old Celtic tunes by its use of a more elaborated finger-picking style based on an interplay of A7sus4, Asus2 and G chords — see for instance the discussion of the song by David Hodge (2002). The song can easily be played fully harmonized with standard triadic chords, though, as we have shown in figure 5 above.
At the right: folk singer Martin Carthy
Harmonic ambiguities. Just like "Drunken Sailor", "Scarborough Fair" shows an interplay of the chords of A Minor and G Major. The harmonic structure of the song, though, is far more complex as it adds two more chords to the chord material: the chord of C Major, we find in measures 6, 12 and 13; and that of D Major in measure 8. The introduction of these chords adds to the tonal ambiguity, as both these chords can be related to the tonic of A Dorian as well as the tonic of G Major. The song's flexible melody, moreover, lends itself to many harmonic variations.  A few other chords, for instance, can be replaced by their relative Minors or Majors, implying that also the keys of A natural Minor and C Major are imported in the song. In his arrangement folk singer Russ Shipton (1975), by example, trades the A Minor chord in measure 2 for a C Major chord, and the G Major chord in measure 18 for an E Minor chord. One has to admit, that it sounds rather good. Not all chords, though, are equally fit for this treatment. If we trade the G chord in measure 17 for an E Minor, something clearly goes wrong with the lines of the song. One way or another, the G chords accompanying the bars of measure 17 and 18 have a different feel and seem to be two different chords — as in fact they are.
Censored and unce nsored variants. As we already saw in "Drunken Sailor", we can take our pick out of two varieties of the G chord — the "uncensored" and the "censored" one. Why not combine these variants in one and the same song? There is nothing against it and, as we shall see, that is exactly what happens in "Scarborough Fair". Only one of those G chords can be easily substituted by its Minor relative and that is the "censored" variant. Substitution of the "uncensored" variant by its relative Minor is rather difficult, because this E Minor chord really lies far away in the harmonic tone grid, and therefore its use will seriously endanger the key. So, the fact that the G chord in measure 18 can be traded for its Minor relative, offers us a clue that this one regards the "censored" variant. Vice versa the G chord in measure 17 rather concerns the "uncensored" variant. Both chords not only are different in pitch, but there are also different semantic meanings attached to them. This will need some further explanation.
Chord semantics. Chords are not just musicological devices, but also add some meaning to a song. On a semantic level, there is a relation between chords and lyrics, at least in popular music. This even applies to the basic chords, which all refer to the stance taken by the song's protagonist — voiced by the singer of the song — as an acting subject. The tonic symbolizes the place, where the singer stands voicing the subject's individual point of view. With the subdominant the song takes a step back as if the singer retreats in an inner world to think things over. The dominant on the other hand can be interpreted as a step forward. Combined with this chord we often find lyrics in which the singer addresses someone else explicitly or shouts his or her point of view out at the outside world. The "inside" here is the world of inner thoughts, the "outside" the world confronted by the song's protagonist with his or her actions. The switch between those positions is mediated by the tonic. In this way the basic chords of respectively subdominant, tonic and dominant represent the three core elements of human agency: reflecting, deciding and declaring (Tillekens, 1999).
At the left: in the Renaissance period, playing the dulcimer was an accepted pastime for noble women
Semantic shifts. Chord substitution, such as trading Minor chords for their relative Majors, almost always implies a shift in semantic meaning too, and therefore must fall into accord with the song's lyrics and their vocal interpretation. Paired, the combination of relative Majors and Minor chords, refer to different contexts of conversation. As with many nowadays pop and rock songs, relative Minor chords usually support the voicing of lyrics with a more confidential tone, as used in conversations between friends, an thus more open to doubts and questions. The Major relative, in turn, gives access to a more open voice for a wider public and therefore carries a more formal, polite, and thus censored, but also more socially binding ring.
Semantic combinations. The semantic meanings of the basic chords and those of chord substitution can also be combined. Again "Drunken Sailor" is a good example, if we take its chords for a Minor tonic and a Major dominant. Hearing the first line of the verses, accompanied by the tonic chord of A Minor, one can easily imagine a bunch of sailors sitting around a table, confidentially making plans and trying to decide what to do with their absent, drunken comrade. Next, shifting to a Major dominant, borrowed from the relative key of C Major, we hear them repeating their propositions in the second line, now shouting their ideas out aloud for all to hear as an open, formal declaration intended for a wider audience. Something of the sort will also happen in the first line of "Scarborough Fair" if we, as Shipton does, substitute the A Minor chord in measure 2 with a C Major chord:
1 2 3 4
Are - you go- ing to Scar- bo- rough Fair? -
|Am |C |G |Am |
Here we stay to the tonic, but the last part of the phrase "Are you going to ..." also will carry a more polite ring. It will sound as if the singer knows that this part of this question has already been formally decided and that the real question — "remember me to one ..." — will come later on.
Trading Minor chords for their parallel Majors. If we compare the basic chords of A Dorian with those of natural A Minor, we can say that the Minor subdominant of the latter is replaced by its parallel Major in the former — as instead of a D Minor we find a D Major chord. Parallel Majors, like these, also have their own excess of meaning to add to the lyrics. The D Major in the second line of "Scarborough Fair" shows how:
6 7 8 9
Pars - ley, sage, - rose- ma - ry and thyme - -
|C |Am |D |Am |
We don't have to question the place of this chord in the tone grid. Stuck between two A Minor chords, it clearly is the regular subdominant of A Dorian. Indeed, as each subdominant, it focusses attention on thoughts circling in the back of the mind of the song's protagonist. Maybe it is not a coincidence, that the herb "rosemary" metaphorically refers to a loving remembrance, and because of the emphasis on the syllables, we may even surmise that the woman in question is called Mary. Compared with the D Minor chord, the regular subdominant of A Minor, this Major chord, seems to add some emotional value to the lyrics. Just like the relative Majors, it contributes to a more open and public voicing of the lyrics. There is a difference, though. It sounds as if there are also some private emotions involved, strong enough to shine through. In this case, there is some plaintiveness to the words. That is why we qualify this chord here "uncensored". Of course, the same goes for the G Major chord at the subdominant side of the tonic, which for that reason we did label as "uncensored" earlier on. But, how does this particular chord relate to the basic chords?
The far flat-seventh. For an answer to that question, we will have to return to the first phrase and continue to measure 3 were Scarborough Fair is mentioned as the destination of the listener and the location of the singer's lover, and the accompaniment shifts to a G Major chord. In this context this chord is a good example of the "far flat-seventh," which is often used in folk and popular music to denote a distance in time and space between the singer and his beloved one. Under this chord we can find such lyrics as "Far away from home," "She's gone and left me," and so on. To this end in a song in a Major key, usually the Major flat-seventh chord is applied (Kramarz, 1983: 53; Tillekens, 2002: 125-129). In a song in C Major this is the chord of B-flat, which also can be traded for its relative, G Minor. In the Dorian mode of "Scarborough Fair" it again can be a parallel Major chord, which will add an "uncensored," emotional feeling to the feeling of distance, giving the lyrics an plaintive ring. Or, as the key of A Dorian also provides an enharmonic "censored" G chord, it can be voiced in a more polite way. Both these interpretations can even be combined, as for instance in measures 16 to 19 of the fourth line:
16 17 18 19
She - once was - a true love of mine - -
|C |G (uncensored) |G (censored) |Am |
Of course, it all depends on how the singer will catch the pitches and voice the lyrics. Indeed, Dorian songs open a window to many interpretations.
Shifts in subject identification. By going back to the third line of the song, we now can take things yet one step further, as this line illustrates perfectly that a shift of level sometimes also can imply a change in functional harmony. If the singer, by accentuating the G note in measure 14, allows the melody to really shift to the key of G Major, we will also experience a shift in the logics of functional harmony. Whereas we still can take the C chord in measure 12 for the relative Major of A Minor, in short as a secondary tonic, we will have to view the C chord in measure 13 in retrospect as the subdominant of G Major.
11 12 13 14
Re- mem - ber me - to one who lives there - -
|Am |C |C |G |
A Dorian: |i |bIII |- |- |
C Major: |- |I |I |- |
G Major: |- |- |IV |I |
Now, if we take a closer look at the lyrics, we will notice that with the key also the song's perspective is shifting. Underscoring the phrase "me to one", the singer tries to identify with the woman living in Scarborough Fair, turning her into the subject of this part of the song. Semantically this change in perspective and subject identification, here, perfectly fits to the content of the song's lyrics. As we have argued, a subdominant (IV) usually refers to the thoughts of the acting subject, and by turning the C chord from a tonic in measure 12 into a subdominant in measure 13, the song's protagonist is bringing himself virtually into the mind of his lover. Such shifts in perspective and subject identification are made possible by the double tonics of the Dorian mode, and we find them in many other Dorian songs.
The Dorian twin tone system. With all these shifts of level, shifts of semantics and shifts in functional harmony, the Dorian mode in "Scarborough Fair" comes to the fore as a fully developed twin tone system: a tonal system equipped with two tonics, each surrounded by its own set of basic chords, as shown in figure 8 below:
Figure 8: The Dorian twin tone system
Within this system we find the basic chords of A Dorian, with A Minor for its tonic, D major for its subdominant, and E Minor for its dominant. Added at the subdominant side, as the subtonic, there is the "uncensored" G Major. Next to that there is another set of chords having G Major as its tonic and C Major as its subdominant. Both sets are hooked into each other by the chords of C Major and E Minor. C Major serving both as a substitute and secondary chord for the tonic (I) of A Dorian as well as the subdominant (IV) of G Major. Likewise the E Minor can be used to refer to A Dorian's dominant (V) or serve as a relative Minor substitute (vi) of G Major. G Major itself can shift between "censored"and "uncensored". The system can be extended at will with other chords, like the F chord serving as a secondary subdominant in the key of A Dorian or a subtonic in the key of G Major. With all these ambiguities, the twin tone system opens many possibilities for composers and performers alike. Recalling the words of Widders-Ellis and Gress (1994: 99), it will all depend on "line's momentum" and, we can add, the interpretation of the singer.
Never reaching the end. That is not yet all we can learn about Dorian songs from our examples. Remarkably too, the musical marks of Dorian seem to favour certain lyrics. With their harmonies swinging back and forth endlessly, the songs' lyrics often take the form of unending stories. Old sources of "Drunken Sailor" mention only a few verses, but in the song's history many additional verses were made, advising to put him in a long boat, in his cabin or in the brig, to trice him up in a running bowline, to stick him in the crow's nest, to hang him from the yard arm, and so on. Taken together, they form an endless list of places fit to sober up a drunken sailor. The lyrics of "Scarborough Fair" show a similar pattern. Originally the song had at least ten verses, of which the Simon and Garfunkel version skips six, meanwhile skilfully weaving their own counterpoint melody — "Canticle" — through the remaining ones. In the original version the verses, however, go on to list a whole series of impossible tasks: sewing a seamless shirt and washing it in a dry well, finding an acre of land between the sea and the shore and plowing it with a lamb's horn, reaping the harvest with a leather sickle and so on. There's really no end to both songs and most recorded versions, indeed, just fade-out at the end. The same effect pops up in today's rock songs with Dorian inclinations — think again of those "nights in white satin", which as the Moody Blues sing, are "... never reaching the end."
Analytical tools. Luckily, this article is not a Dorian song and nearing its end, we can summarize our findings. One thing has become clear. What at first sight seemed just simple songs, prove to be rather complicated musical constructions. Now some last remarks can be made about the labelling of songs as Dorian. In folk music and popular music, tunes always have priority above scales. Obviously, songs like "Drunken Sailor" and "Scarborough Fair" were not made by picking a scale like the Dorian one and building a tune around it. Popular songs and folk tunes can only be analysed afterwards by musicologists using the concepts of scales, or modes as combinations of scales. In respect to folk music and popular music, scales and modes are just analytical tools. So, there may be Dorian scales and Dorian modes, but we still have to be careful in stamping songs with the Dorian label. For one thing, as we have seen, the harmonic qualities of the Dorian mode are far more important than the melodic ones. From that point of view, the Dorian song family rather is a collection of overlapping sets of songs, than a strict category.
At the right: the Moody Blues
Modal typecasting. Harmonically, the Dorian scale relates easily to a whole series of other scales and as a result Dorian songs will differ accordingly. Some songs will be dominated by the substitution of the Minor chords of the Dorian scale by their relative Majors. As we have seen, neglecting the F-sharp in its melody line, "Drunken Sailor" can be analysed that way. In other songs, as in "Scarborough Fair", the Dorian scale will more clearly interact with its Dorian twin scale. Moreover, some Dorian songs will lean predominantly towards the subdominant side, while others will stress the dominant side of the Dorian key. There are still more variations, as some Dorian songs also import secondary chords borrowed from their parallel Major key. That other well-known traditional, "Greensleeves", for instance, not only combines the scales of A Dorian and C Major, but also includes a Major dominant chord, borrowed from the parallel scale of A Major. History makes things still more fuzzy, as different songs can be retraced to the same roots. During their development and geographical spread many tunes, indeed, did change considerably. If we strictly keep to categorizing songs to their scales, this would imply that different versions of the same tune will have to be ascribed to different modes. Sometimes it even proves difficult to pinpoint their exact scales, as many Dorian tunes show weak thirds, wavering between Major and Minor. So, labelling a double-tonic tune as Dorian, will not shed any light on its harmonic characteristics. For that reason Van der Merwe (1989) rightly refuses to typecast the many examples in his study according to their scales.
The Dorian heritage. Taken as just a scale, the qualification Dorian actually is an anachronistic misnomer. Dorian songs, however, have some things in common, worth looking into, as they are built on two harmonically distanced keys. This really is what makes them special, compared with other songs. Some of the marks of these songs are less important, such as their preference for series of consecutive fifths in the melody line, and their predilection for arpeggiated harmonies and drones, all strengthening the feeling of a wave rolling back and forth. Marks like these can be derived from what lies at the core of the Dorian song family, as all members show a mix of a Minor and a Major key, lying one whole step apart, sharing the same tone material, and hinging harmonically around a "central note". Melodically this facilitates the use of enharmonic notes with different pitches, with which pleasing effects can be effectuated. Harmonically this combination of scales can easily be enriched with chords, borrowed from other related scales, and used for shifting between different semantic contexts and for addressing different subject positions. In respect to the lyrics, there is the effect of an unending story. As "Drunken Sailor" shows just a few, well-chosen chords will suffice to sort these effects. No wonder, the Dorian heritage was kept preserved and exploited in the idiom of pop and rock music — for instance, it is the main musical component of Pink Floyd's album The Dark Side of the Moon. There is more to tell about that subject, but we will keep this in store for an another article.
1. For those unknown with the musicological vernacular, a "tonic" in relation to a scale refers to the one note defining the actual pitch of the other ones. The concept "mode" has different meanings in the field of music studies. Some scholars see modes as a specific subset of scales, comprising the seven permutations of the diatonic scales, while others use the concept to denote the whole, arbitrary collection of folk scales. Most of the times, however, the label "mode" or the adjective "modal" is used to refer to songs when next to the tonic one or more other notes play an important role — sometimes even as a cotonic — in structuring the tune. Here Van der Merwe (1987: 101-106) uses the concept "modal frames". He also takes another stance in regard to scales and modes. He defines a scale as just a "bare set of notes" and specifies a mode as a scale with one note playing the role of the tonic. Consequently he speaks of an "atonic mode", when a scale is missing a clear tonic. More in line with the standard conventions, we here use the concept of scales for what Van der Merwe defines as "modes" and "modes" for what Van der Merwe defines as "modal frames". So when we speak of a scale we mean a restricted set of notes with a clear tonic that defines the scale's key, which in turn gives it its name.
2. Many people still think that most British, Irish and American folk songs are modal, where in fact modal songs constituted only a small part of the body of original Anglo-Saxon folk music (Kramarz, 1983: 40). The reason for this identification of folk music with "modality" probably lies in the fact, that during the 19th century modality became the main criterion to distinguish Classical music from other musical styles. Van der Merwe (1989: 18-20; 241-242) calls the divergence of styles the Great Musical Schism, where folk and popular musicians were promoting and Classical composers were increasingly avoiding the lavish use of double tonics and the resulting enharmonic notes. During this process modal songs got prominence in the domain of folk music above the usual Major ones, as they were kept alive by performing artists. Besides, in the same period folklorists became interested in preserving folk music and, as Dave Harker (1981; 1985) notes, many songs were "modalised" when they started paying for what sounded to them as real, authentic — i.e. modal — material, thus involuntarily furthering its production.
3. To name them all, the seven diatonic "modes" comprise the Ionian (starting on C), Dorian (starting on D), Phrygian (starting on E), Lydian (starting on F), Mixolydian (starting on G), Aeolian (starting on A) and Locrian (starting on B) scale. The ordinary Major scale coincides with the Ionian scale, just like the Aeolian scale does with the natural Minor scale. Though in a lesser degree than the Aeolian scale, the Dorian, Phrygian and Locrian scales all have a Minor flavour. Not all these scales are as relevant for the musical idiom of pop and rock music. Next to the Mixolydian scale the Dorian one is the most welcome guest in this domain of popular music. The other scales are predominantly found only in folk music. The Phrygian scale is characteristic for much Spanish — especially from Castile and Andalusia — and some Italian and Hungarian folk music. The Locrian scale is only occasionally found, for instance in Icelandic and Greek music.
4. As the scale of natural Minor coincides with the Aeolian scale, we could call pairs of Major-Minor relatives by the name of Aeolian twins. Indeed, songs in a Major key but incorporating a massive dose of relative Minor chords — like many folk-rock and Brit-pop songs — often are called Aeolian by the music critics. By the way, the list of related scales, presented in figure 2, goes full circle, as the scales of C and G are also modally connected by the intermediary scale of G Mixolydian that shares its tone material with C Major. The role of these scales in popular music, however, is a story that I will reserve for another article.
5. Imitating the way humans speak, melody usually progresses by small steps, whereas harmony is built on larger intervals. Or said otherwise, melody is based on the principle of tone proximity as harmony is on the principle of harmonic distance. In chord progressions, usually, both principles are at work. In a simple progression such as I -» IV -» V -» I, for instance, the parts I -» IV and V -» I may be said to be harmonically motivated, whereas the transition IV -» V is melodically motivated. The sociologist Max Weber (1904/1905), by the way, erects his theory of western music largely on the fundamental conflict between both these principles in that style of music.
6. Among the recorded versions of "Scarborough Fair", there are some variations in the melody line. Many transcriptions of the song, for instance, present an A Minor chord in the last measures of the third line. Consequently, we will then find a closing A in the melody line, which will restrict the tone material of "Scarborough Fair" to just one octave. Often the melody here also will slightly waver between G and A.
Becker, Howard S. (1994), "American popular song. Sharing the standards of the American sound." In: Soundscapes, July 2002.
Bone, David (1931), Capstan bars. Edinburgh: The Porpoise Press, 1931; New York: Harcourt, Brace and Company, 1932.
Harper, Colin (2000), Dazzling stranger. Bert Jansch and the British folk and blues revival. London: Bloomsbury, 2000.
Harker, Dave (1981), The making of the Tyneside concert hall. In: Richard Middleton and David Horn (eds.), Popular music I. Folk or popular? Distinctions, influences, continuities, Cambridge, Cambridge University Press, 1981, 27-56.
Harker, Dave (1985), Fakesong. The manufacture of British "folksong", 1700 to the present day. Milton Keynes: Open University Press, 1985.
Hodge, David (2002), "Scarborough Fair. Songs for intermediates no. 5." In: GuitarNoise.com, retrieved from the Internet on July 8, 2002.
Tillekens, Ger (1999), "The amazing grace of Never Ever. The power of a four-girl band and a three-chord song." In: Soundscapes, November 1999.
Tillekens, Ger (2002), "A flood of flat-sevenths." In: Russell Reising (ed.), Every sound there is. London: Ashgate, 2002, 121-136.
Ward, John (2002), "Shanties and sea songs." In: John Ward's Web Pages, retrieved from the Internet on July 20, 2002.
Weber, Max (1904/1905), The rational and social foundations of music. Translated and edited by Don Martindale, Johannes Riedel and Gertrude Neuwirth. Carbondale: Southern Illinois University Press, 1958.
Widders-Ellis, Andy, and Jesse Gress (1994), "Take a rad song and make it better. Exploded views of three Beatles classics." In: Guitar Player, 1994, 28, 9, 95-104, 106, 108.