Canadian Journal for Traditional Music (1986)
Guidelines for Harmonizing English-Language Folk Songs: General Outlooks
Over the last few years, I have been engaged in transcribing several hundred English-Canadian folk songs. These were originally collected in the field by Edith Fowke and have been preserved on tape and made available to the scholarly community in the archives of the Listening Room of York University's Scott Library. The items I have transcribed so far include both children's and adult songs that Professor Fowke has collected from singers in Ontario. In musical style, these are similar to other English-language folk songs found throughout North America and Britain.
In preparing the songs for possible publication, I have tried to provide not only a faithful rendering of the texts and tunes but also indications of appropriate chords (by way of letter-name chord symbols such as "D," "E mi" and "G7") for the modern "folk-song revival" singer who might wish to perform the material with the accompaniment of a chord instrument such as the guitar. In the process of adding chord symbols, I have had to solve a number of problems that are not only musical but also aesthetic and practical in nature. I feel that I have arrived at some fairly reasonable solutions to these problems and, accordingly, I would like to share with a wider public both the solutions and the reasoning that led to them.
General Outlook
In principle, there is no single "best" harmonization of a folk song. In some instances, a singer might wish to perform the songs as they were originally performed in the oral tradition, that is, without any instrumental accompaniment at all (i.e., monophonically). In other cases, a simple drone accompaniment might be very effective. Such an accompaniment might merely consist of the tonic of the key or mode (i.e, doh), or both the tonic (doh) and the dominant a fifth above (i.e., sol). Beyond such non-existent or exceedingly simple accompaniments, there seems to be no limit to the degree of complexity that might be found in an effective folk-song arrangement.
Intermediate degrees of complexity might involve a few simple strummed diatonic triads, similar chords in figurated "picking" patterns, vocal "harmony" parts (e.g., in thirds and/or sixths), chromatic and modal harmonizations, ostinato melodic figures, and even distinctive counter-melodies that might serve as a foil for the main tune. In addition, at the upper end of the spectrum of complexity, one can cite various effective "art-music" arrangements made by classically-trained composers, arrangements that have entailed choral, orchestral and even electronic resources.1
Although the gamut of complexity through which effective folk-song arrangements might range is quite great, the sorts of arrangement that seem to be most useful to a modern folk-song revival singer who resorts to printed versions of folk songs fall within a fairly narrow range. By and large, such performers appear to be best served by an indication of a few chords that will provide a relatively simple harmonic underpinning for the tune.2
Although the traditional tunes of English-language folk songs were generally sung without any chordal accompaniment at all when they first flourished, chordal accompaniment has become the norm in contemporary performances both by professionals and by amateurs. And for the most part, simple chordal accompaniments do not seem to distort the musical effect of traditional songs, because the melodies of these pieces seem clearly to embody strong harmonic implications. Although the tunes might not have been performed with chordal accompaniment when they were first sung, they appear to project rather strongly certain implicit harmonizations.3
One of the best head-starts for trying to find a reliable way of choosing chords for these tunes is provided by the numerous harmonizations that have been made in the past. Recordings of English-language folk songs by professional folk-song revival singers and printed collections of folk songs by professional arrangers constitute a readily available source of ideas that can help one "get off on the right foot."
Much less useful is any training one might have received in so-called "classical" or "conservatory" harmony. (Indeed, this article might have been subtitled "Cautionary Notes to a Conservatory-Trained Musician.") Although the harmonic style of folk songs overlaps, to a certain extent, the harmonic style of classical composers (at least, insofar as this style has been portrayed in the standard conservatory textbooks on harmony), there are important divergences between the two idioms (see below). Therefore, in seeking guidance for how one might go about providing chord symbols for the folk songs that one has transcribed, I have looked for clues and general "rules-of-thumb" in recordings and printed collections of English-language folk songs rather than in the prescriptions offered by conservatory texts on harmony.
Whereas a student of "classical" harmony might look ultimately to the theoretical writings of a Rameau or to the pieces of a Beethoven for guidance, the arranger of folk songs might find that those who have successfully recorded or published their own harmonizations of folk songs provide his best models, for the efforts of these "folk-song masters" have been tried in the fire of highly knowledgeable and sensitive audiences and readers.
If one critically examines the harmonizations of such masters, one finds that their arrangements are remarkably consistent. Indeed, one can summarize their procedures in "rules," which, though more flexible than the conservatory's "rules of harmony," nonetheless provide very reliable guidance to the neophyte arranger. These "rules," or better, "guidelines," are not to be considered edicts or regulations against which one must not offend. Rather, one should view them as summaries of ways in which successful arrangers seem to have arrived at their solutions (whether consciously or unconsciously). And if one's initial choice of chords seems to be "off" at times, the guidelines can often help one analyze the situation in order to find a more satisfactory solution. What I offer below, then, is a set of guidelines that I have developed through an analysis of the harmonizations that professionals have made in the past, and I hope that these might provide valuable guidance for those who may never have harmonized an English-language folk tune before.
Although these guidelines are based on the practice of successful professionals, they do not cover all the situations one might encounter in their arrangements. In some cases, deviations from the guidelines on the part of professionals seem merely to have been the result of a slip of the hand or a misprint. In other instances, it seems to me that professionals have been unnecessarily "fancy" (i.e., complex) in their harmonizations. Indeed, in professional arrangements, the most consistent deviations from the guidelines presented below seem to be in the direction of unnecessary complication rather than in the direction of oversimplification (relative to the overall style).
My feeling is that one can always further complicate a harmonization, but it is often rather difficult to simplify it. One of the frequently formidable challenges in choosing chords for folk songs is to keep one's arrangement as simple as is feasible (again, relative to the style). In the event that one might wish to complicate a simple arrangement, I will offer some suggestions about this in an article that is to appear later.
Although one will find some exceptions to the guidelines in the arrangements of professionals, these exceptions are rare and tend to be in the direction of overcomplication rather than oversimplification. From the opposite vantage-point, if one applies the guidelines consistently to the melodies of English-language folk songs, one will find, by and large, that one's harmonizations sound quite "professional." I use the phrase "by and large," because the guidelines do not solve all the problems that one might encounter.
From time to time while working on this study, I ran into a tune that, at certain points in its course, resisted harmonic treatment according to the guidelines I had already developed. In some cases, it seemed to me that the tune itself was not harmonic in conception, at least not in the sense of the triadic conventions that I describe below. In other cases, seemingly recalcitrant tunes encouraged me to re-examine the guidelines I had earlier formulated and led to a slight revision. The reader who applies the guidelines described below might well discover instances such as these, and, I hope, refine the guidelines as I have attempted to do. In this way, the guidelines might evolve by way of a collective effort into an ever more versatile system that would be equipped to handle all but the most "difficult" melodies that one might wish to harmonize with a basically triadic vocabulary.
Soon after I began examining the harmonizations of professionals, it occurred to me that, from a harmonic point of view, there are two major stylistic categories in the repertoire of English-language folk song. One category consists of songs that are in a major key and that can be harmonized quite effectively with the chords I, II(7), IV, and V(7).4 (In D major, these chords would be D, E or E7, G and A or A7, respectively). The other category consists of everything else: major tunes that call for minor chords (e.g., Em, F#m and Bm in D major), minor melodies, modal songs (e.g., songs in Dorian, Aeolian, and Mixolydian), pieces with so-called "gapped" scales (i.e., tunes lacking one or more of the usual seven-scale degrees), and tunes that use chromaticism (i.e., notes that do not belong to the scale of the piece and call for accidentals in their notation).
Among the major tunes that can be effectively harmonized with just the I, JJ(7), IV, and V(7) chords, there are some subgroups. For example, many children's songs can be harmonized effectively with only I and V(7) chords.1 In addition, many adult songs call for I, IV, and V(7) chords and often feature the IV chord much more prominently than in so-called "classical" music. Additionally, there are several songs that employ the full resources of the category: I, II(7), IV, and V(7) chords.
Because this category as a whole involves the most straightforward problems of harmonization and because a solution of these relatively straightforward problems provides a rather firm foundation for a consideration of the other, harmonically more complex, category of songs, the remainder of this article is devoted to what might be termed the "I -II(7) - IV - V(7) type." In this way, I deal with general considerations that arise in a very large important group of songs. And these considerations will be found of value when one attempts to harmonize more complex melodies that are minor, modal, chromatic, etc.6
The I - Ii(&Amp;) -Iv - V(7) Type
In order to determine whether a song belongs to the I - II(7) - IV - V(7) type or not, one can begin by simply singing (or scanning) the melody. If the tune appears to be major, it might well belong to this type. If one finds on applying the guidelines set out below that the song can be effectively harmonized using only the I, II(7), IV, and V(7) chords (or a subset of these), then one has confirmed that the song belongs to the I - II(7) - IV - V(7) type.
The reasons why I recognize this group of songs as a special type are twofold. First, from the point of view of the analyst, this group merits consideration as a special type on the purely empirical grounds that there are so many songs that belong to the group. Secondly, from the point of view of the performer, the songs in this group are special because, by and large, they make the smallest demands on the technique of someone who plays a chord instrument. Indeed, if they were transposed to a single key, one could accompany the songs in this group — and there are hundreds, if not thousands of them — even if one only knew how to play from two to six chords.
Choice OF Key
Unlike the situation in so-called "classical" music, the key of a folk song typically varies from performance to performance. If one is pedantic, one might insist on notating and/or performing a given folk song in precisely the same key in which one originally heard a traditional performer sing it. One's scholarly motivation here might be to preserve information on the absolute pitch of each of a performer's songs in order to study the ranges that the singer used.7 If one's goal were ease of analysis, one might choose to notate all folk melodies as far as possible with no key signature (e.g., in C major) and with no accidentals, and one could then make a note of the interval by which one transposed the tune upward or downward from the traditional version originally heard (e.g., by means of a short phrase such as "original is 5 semitones higher").8 However, if one wishes one's arrangement of a tune to be taken up by the widest spectrum of performers, it might be wise to consider certain special aspects of the performers' situation.
First, as far as the melody is concerned, one can take into account the normal vocal ranges of untrained singers, and indeed the majority of trained singers. There seem to be relatively few "true sopranos," "true altos," "true tenors," or "true basses" in the overall population today. To
judge from my experience with trained and untrained singers, it seems that the largest segment of the singing population consists of women with a fairly modest mezzo-soprano range and men with a correspondingly modest baritone range that lies about an octave lower than the mezzo range.9 By and large, these singers seem most comfortable vocally when singing between middle d and the a immediately above if they are women (see Example la), and between bass-clef d and the a immediately above if they are men (see Example ib). In addition, children also seem to be most comfortable in the upper version of this pair of ranges (i.e., between middle d and the a immediately above — see Example 1c).
By and large, children's songs tend to be narrower than adult folk songs. Indeed, they often do not range much beyond a fifth. 10 Thus, it is generally possible to locate a children's song so that most of its notes fall within range from middle d to the a immediately above. Adult folk songs tend to be wider than children's songs: about an octave or a ninth.
Some adult songs have an authentic range and lie entirely, or almost entirely, in a range that extends about an octave above the tonic (i.e., from low doh to high doh). Other adult songs have a piagal range and lie for the most part from the
fourth below the tonic (i.e., from low sot) to the fifth above the tonic (i.e., to high sol).11 By and large, the mezzo sopranos and baritones that one encounters begin to have difficulties above their respective high d's, especially if they have to sing for a long time in this part of their range. Similarly, mezzos and baritones have a certain amount of difficulty below their respective low a's, especially, again, if they have to sing for a long time in this part of their range. Much the same holds for children, in fact, even more so, because their ranges are generally not as wide as those of adults. Accordingly, one's arrangement of a melody will be directly and effectively performable by the greatest number of singers if one observes the ranges described in Example idandle.
For children and women, the central fifth (d-a) should be the location of most of the notes of the melody and the overall range of an eleventh (rarely called for in folk songs) from aI to d' should form the outer boundaries within which one's transposed version of the melody is fitted (Example id). For men, the same applies, except down an octave (Example le). Accordingly, by a judicious choice of key, one can arrange a melody in such a way that the largest number of children, women, and men can sing it with the least difficulty. Moreover, since most people (including men) seem to be able to read music in the treble clef more fluently than music in the bass clef, it seems best to locate one's melody in the treble-clef range described in Example id. Thus, to reach the greatest number of readers, one might have to transpose a traditional melody into a key where the guidelines of Example id can be applied. In many cases, by virtue of the prevalence of authentic and plagal ranges, this key might be D major, E major, F major, or G major. And if one's arrangement is also aimed at a scholarly readership, one can specify the interval of transposition by the means of a short phrase (see above).
Choice of key can also be made with the player of a chord instrument in mind. To judge from my experience with novice guitarists, it seems that D major, G major, A major, and E major are the keys in which they can perform I, II(7), IV, and V(7) chords (or a subset thereof).13 As a general guideline, then, I would suggest C major or, better, D major for authentic melodies and F major, or better, G major for plagal tunes.
All the same, one cannot overlook the fact that even novice guitarists can transpose a group of chords upward very easily by means of a so-called "capo," so that when one is in doubt, it might be wise to notate a tune a little on the low side. Indeed, by and large, D major seems to be the safest key in which to place a melody and its attendant chords, and one would be well advised to try a song in D major before rejecting this key in favour of another (except where one's arrangements might be aimed directly at novice performers on the ukulele that is tuned in C). The more advanced player's taste for a wider variety of chords can be satisfied by tunes in the other, more complex harmonic styles to be dealt with in the sequel to this article. Notwithstanding the more elaborate style of songs in the other, more complex category, the bulk of songs in the simpler I - II(7) - IV - V(7) style can be set effectively in such a way that only the chords D, E (or E7), G, and A (or A7) are used.
Harmonic Rhythm
If one examines professional arrangements of folk songs, one finds that, by and large, as few chords as possible are used and these are changed as infrequently as possible. For the novice instrumentalist, this practice is a blessing, because, other factors being equal, the more rarely and the less rapidly one has to change chords, the easier it is to play an arrangement. For the student of conservatory harmony who attempts to harmonize folk songs, however, this practice often seems to go "against the grain" at first, for the initial style that such students master features relatively frequent and rapid chord changes in the idiom of hymn and chorale settings. In learning to harmonize folk songs, a person who has studied conservatory harmony often has to "unlearn" the somewhat frenetic style of harmonic rhythm that is found in hymns and chorales. By and large, in a folk style, one is wise not to change chords unless the melody "forces one's hand," as it were, whereas a rather dense style of harmonic rhythm, as exemplified by the wonderfully complex arrangements that Bach made of German chorale melodies, is often aspired to in formal conservatory studies.14
For the most part, it seems to be easier for an instrumentalist to add chords to an arrangement than it is to leave chords out of an arrangement. Musical notation seems to have a coercive, imperative voice in which every symbol seems to say "Play me (or else!)." Therefore, one's arrangements will tend to be less frustrating to the novice if one errs in what might seem to be the direction of specifying too few chords rather than in the direction of what will probably be too many. However, if one wants one's arrangements not only to be accessible to novices but also to encourage more advanced players to use and develop further their technique, one might adopt a convention whereby "extra" chords are placed in parentheses.15 These "extra" chords might be overlooked by the novice and played by the more experienced.
There is, of course, a danger if one introduces an unfamiliar notational format and the danger here would seem to consist of the possibility of bafflement among one's readers. As well, the adoption of a parenthetical notation might constitute an irresistible appeal to those novices who, through pride or compulsion, might attempt to play precisely the chords that are not aimed at them, "rising to the challenge," as it were. And when relatively inexperienced players try to negotiate several rapid chord changes, their "brute-force" efforts often lead to a disruption of the music's rhythmic flow which can only be unsatisfying to them and their listeners as they stumble through an arrangement. Nevertheless, I sometimes employ the parenthetical format below for a number of reasons.
First, I hope to convey the idea that differences between arrangements often amount to a mere addition or subtraction of a few chords here and there without a change in the overall "harmonic skeleton" of the arrangement. Second, I hope to show that more than one harmonization of a given melody can be quite effective even within the limitations of the relatively simple I - II(7) - IV - V(7) style. And third, I hope to encourage those who arrange folk songs to try out different solutions to the problems that arise, rather than stubbornly sticking to one.
Metre And Chords
If one examines professional harmonizations of folk songs, one finds that, for the most part, chords are changed on strong beats of the measure. For example, in melodies that are accurately notated in 4-4, generally the chords change very frequently on the first beat, less often on the third beat, quite rarely on the second and fourth beats, and very seldom "off the beat" (e.g., on the second eighth note of a beat). This procedure is consistent with the overall melodic style of English-language folk songs, a style in which syncopation is rare and the chord changes that are implied by the melodies are quite infrequent. Moreover, from the performers' point of view, changing chords on relatively strong beats serves to provide a strong rhythmic underpinning for the tune. In addition, if a chord is changed on a relatively weak part of a measure (e.g., in 4-4, on the third beat vis-à-vis the first beat, or on the second beat vis-àvis the third beat), then there is generally a change of chord on the next stronger part of the measure that arrives (e.g., if the chord has changed on the second beat of 4-4, there will be a change of chord on the immediately following third beat of the same measure, and if the chord has changed on the third beat of 4-4, there will be a change of chord on the immediately following first beat of the next measure). In this way, one avoids "syncopating the chord changes," and further provides a strong metrical underpinning for the melody. However, there are some exceptions.
In the first place, the guideline for unsyncopated chord changes need only apply within a phrase, corresponding to a full line of the text, for phrases constitute a main unit of chord progression in the style. An
instance of this sort of seeming syncopation is found between the third beat of the fourth (full) measure and the first beat of the fifth (full) measure in my setting of "The Battle of the Windmill" (Example 2). In the song as a whole, the appearance of the D-chord on the third beat of the fourth measure is at the end of the first phrase, whereas the next appearance of the D-chord (on the first beat of the fifth measure) belongs to the second phrase.
In the second place, one frequently finds that in order to provide a perfect cadence (i.e., a progression from V or V7 to I — e.g., from A or A7 to D, in D major — at the end of a phrase), one might insert the V or V7 chord on a highly unaccented part of the measure (e.g., the very last quarter note or the very last eighth note of a 4-4 measure), thus seeming to offend against the guideline according to which one would prefer not to change chords on extremely unaccented parts of the measure. However, this formula is encountered so frequently in folk-song settings that one can count on performers — even novices — generally having mastered it. And, one should note, the I-chord of this formula (e.g., the D-chord in D major) virtually invariably will arrive on the following downbeat (i.e., the first beat of the immediately following measure) in such a way that, though a quick chord change is called for, the unsyncopated nature of the chord changes is preserved.
In my setting of "The Doctor-Man" (Example 3), I have inserted such quick changes of chord in parentheses at the ends of the first and fourth phrases (i.e., in the row of chords, between the G- and D-chords that appear at the beginning of the third measure and the beginning of the fourth measure, and at the beginning of the fifteenth measure and the beginning of the last, sixteenth measure). An insertion of the A-chord at these spots is somewhat optional in this instance because of the overall tendencies of the style to which it belongs (see below).
Choice OF Chords: Prominent Notes
What chord should one choose at a given point in a melody? A good thing to keep in mind here is that the chords form an accompaniment for the melody. One of the implications of this statement is that the chords should, as far as possible, support the most prominent notes of the tune. A technique by which a chord can support the most prominent notes that a singer produces consists of having the notes include among themselves the most prominent notes that the singer is producing at any given time. For example, jff# is prominent in a piece that is set in D major, then one might accompany this note with a D-chord, which includes d, f# and a among its members. If c# is prominent in a similar piece, then one might accompany this note with an A-chord (or an A7-chord), for this chord includes a, c# and e (or a, c#, e and g). In this way, the chord that one chooses will "harmonize with" (i.e., be "consonant with") the most prominent notes that are being sung at any given time.
What makes a note prominent? In the first place, if a note belongs to a melodic leap, it is thereby more prominent than if it merely belonged to a stepwise melodic progression. In the second place, if a note appears on a relatively strong beat, it is more prominent than if it appeared at a less accented part of a measure. In the third place, if a note is relatively long, then it is more prominent than if it were relatively short. And in the fourth place, if a note is repeated immediately, then it is more prominent than if it is followed by a note of another pitch.
Most of these points can be seen in my setting of "The Mulberry Bush"
(Example 4, below). Here the d's in the first measure are prominent by virtue of being repeated immediately, appearing on the relatively accented first and fourth eighth notes of 6-8 (in the case of the first and fourth d's), being relatively long (in the case of the fourth d), and belonging to a leap from d tof# (in the case of the fourth d). Moreover, thef# in the first measure is prominent by virtue of a) belonging to the leap from d tof#, and b) belonging to the leap fromf# to the first a of the following measure. The a at the beginning of the second measure is prominent by virtue of its appearance on the first beat of the measure, its immediate repetition (in the form of the second a of measure two), and its membership in the leap fromf# to a. And so forth. Clearly, a chord containing, d, f# and a (e.g., a D-chord) will harmonize most simply with the notes of the first two measures of "The Mulberry Bush." Moreover, although the a's at the beginning of the second measure might have been harmonized by an A-chord (or an A7-chord), the guideline (above) that suggests that one keep chord changes to a minimum would mitigate against such a change of chord at this point.
Another factor to consider is the tempo of the piece. A note that might seem to be quite short and unobtrusive in a fast tempo could take on a fair amount of prominence if one performed the piece in a much slower tempo. Whether a dissonance would arise between the vocal and instrumental parts in such cases might also depend on the rate at which, for example, a guitarist might strum the chords. For instance, in "My Seventy-Six Geared Wheel" (Example 5, below), one might wish to insert an A-chord into the second half of the third measure (indicated as an alternate harmonization by means of parentheses) if the tempo were slow and if one were strumming on every third eighth note in 6-8 (i.e., on the first and fourth eighth notes of each measure). Such an A-chord would harmonize well with the fairly prominent e of the melody at this point, whereas if one continued to strum a B-chord (or a B7-chord) at this point (cf. the row of chord symbols above the third measure), and if one were performing in a slow tempo, there might be a rather unruly dissonance between the e of the melody and both the d# and the f# of the accompanying chord (which would include b, d#, f# and, maybe, a). However, if one were performing in a fast tempo and strumming only once per measure (i.e., on the first beat of the measure), any possible clash between the e of the melody and the B-chord (or the B7-chord) might be overlooked as the e would recede in prominence.
Before leaving the topic of prominent notes for the time being, I should point out that, since there are several criteria according to which a note might be judged as prominent or not, there is some room for ambiguity in determining which are the most prominent notes and hence which chord should be chosen at a given point. For instance, in "The Battle of the Windmill" (Example 2, above), one might feel that the first note of the first measure (i.e., b) is quite prominent by virtue of its appearance on the relatively strong, first beat of the measure. However, if one set this note to a chord that contains it (e.g., a G-chord), one would find, on setting the remaining notes of the measure to a D-chord (which seems to be unambiguously called for at this point, by virtue of the clearly prominent a, f# and d), that there were more chords than necessary and even more importantly, that one had introduced a harmonic syncopation (in that the D-chord would be introduced on the second beat of the measure and there would not be a chord change on the following, third beat — indeed, there would not be a chord change until the first beat of the following, second measure). This example shows that one cannot apply the guidelines described thus far in a wholly mechanical way and yield effective results. In part, one's ear can provide a reliable guide in cases such as these. But in addition, yet another guideline can account for what is happening here.
Often one finds that a note that seems prominent according to one or more criteria functions as a mere adjunct to a chord. Typically, such a note is preceded and/or followed by a note in the melody that unambiguously forms part of a chord, and that the other note (or the other notes) is (or are) just a step away. In the parlance of conservatory theory, the original note is a "non-harmonic" or "non-chord" tone that might be called a "passing note," an "accented passing note," a "neighbour note," an "auxiliary note," an "appoggiatura," an "anticipation," or an "échappé." Generally, it takes several lessons in conservatory harmony to gain a facility in the use of such notes and terms, but the arranger of folk songs need not bother with the theoretical distinctions that they embody. For such an arranger's purposes, it will suffice to observe that an otherwise prominent note might not correspond to a note in the chord that accompanies it if, in the melody, it is preceded and/or followed by a melody note that a) is a step away, and b) belongs to the chord that accompanies
it. A converse to this general guideline is that one should take great care in ensuring that notes that are both the second and first members of leaps (e.g., the first f# in "The Battle of the Windmill," Example 2, above) belong to the chord that is being played at the time.
My setting of "The Battle of the Windmill" (Example 2, above) points up yet another aspect of folk-song harmonization that might surprise someone who is versed in conservatory harmony. As one can note at the very outset of the song, no chord is provided for the first two notes. Whereas, in conservatory harmony, one is taught to harmonize every note of the melody, one finds that in folk songs, one ignores any notes that appear before the first downbeat. 16
Related to this practice is another aspect of folk-song harmonization that might be surprising to the conservatory-trained musician as well. Often the melody anticipates the chord that is about to come on the next relatively strong beat. For instance, in "The Jolly Old Miller" (Example 6, below), a rather prominent a appears on the second beat of the fourth measure just before the chord changes to D on the following third beat. Although the a would seem to clash strongly with the notes of the Gchord that accompany it, especially if the song were sung slowly, it is better accounted for as a prominent anticipation of the D-chord that follows it.
In another children's song, "Old Roger is Dead" (Example 7, below), the second-to-last note (f#) anticipates the arrival Of the D-chord in the following, more accented part of the measure (i.e., the second beat of the last measure). Although the a of Example 6 and thef# of Example 7 seem quite prominent, one would delay the arrival of the D-chord until the following beat in both cases in order to avoid a harmonic syncopation, and one can view the apparently discrepant notes as anticipations that are "assimilated" by the arrival of the following D-chord in a metrically strong position.
Summary
In sum, one can note that the following general considerations might arise if one attempts to harmonize English-language folk songs: the singing ranges of those who will use one's setting, the performance capabilities of those who might perform one's arrangement, one's taste for relatively simple or complex harmonizations, and the underlying harmonic structure of the songs. These factors can influence one's choice of a key, the frequency and rhythmic arrangement of the chords one selects, and one's assessment of prominent notes in the melodies. I have dealt with these factors, because they might emerge in the process of harmonizing any English-language folk song. And I have illustrated my approach to the problems that arise from a consideration of these factors by means of songs that can be harmonized effectively using only I, JJ(7), IV, and V(7) chords because such songs are by far the most frequent in the repertoire and the easiest to perform (from the point of view of the novice instrumentalist). Other factors tend to be specific to given types of tunes (e.g., modal, gapped and chromatic) and, accordingly, are to be dealt with in a sequel.
York University,
Downsview, Ontario
1. A selective listing of classical arrangements for orchestra, band, chamber ensemble, piano, choir, and voice of English-Canadian folk songs appear in Edith Fowke, "A Reference List on Canadian Folk Music," Canadian Folk Music Journal, 1 (1973): 52 - 54; recordings of English-Canadian folk songs by folk-song revival performers are listed in ibid. :55 - 56, and in the same author's supplements to this list in the Canadian Folk Music Journal, 6 (1978): 55 - 56, and 11(1983): 58-60.
2. A sampling of lead sheet arrangements of English-Canadian folk songs with melody, text, and chord symbols appears in Edith Fowke's "A Reference List," Canadian Folk Music Journal, 1 (1973): 49 -51, and the 1978 and 1983 supplements to this list (see note 1, above): 46-49 and 51 -54, respectively.
3. Certain regularities in English-Canadian children's songs can only be accounted for in terms of the notion that chords are implicit in the tunes. On this point, see Jay Rahn, "Stereotyped Forms in English-Canadian Children's Songs," Canadian Folk Music Journal, 9(1981): 43-53.
4. In this study, I employ the notation "V7" as an abbreviation for "V or V7" and "II(7)" as an abbreviation for "II or II7." One can understand II as what is referred to in functional harmony as "V of V" and II7 as "V7 of V." For what follows, it might be useful to have a working knowledge of elementary theory. However, I try to provide concrete examples for all the technical points.
5. For example, all of the children's songs in the "London Bridge" type can be harmonized effectively with just I and V7 chords. On this point, see Jay Rahn, ibid.
6. These more complex songs are to be dealt with in a later article.
7. This seems to have been the practice followed in a number of the arrangements referred to in note 2, above.
8. Transcribing oral tradition material using as few accidentals as possible is practised by some ethnomusicologists. See, for example, Mieczyslaw Kolinski, " 'Barbara Allen': Tonal Versus Melodic Structure," Ethnomusicology, 12(1968): 209 - 12.
9. My remarks on trained and untrained singers, guitarists, and ukulele players are based on my experience in accompanying and conducting community, church, elementary-school and college choirs, teaching musicianship to music-majors and non-specialists at the university level, and directing the music in a theatre school for children and adolescents.
10. For representative ranges of English-Canadian children's songs, see Edith Fowke, Sally Go Round the Sun (Toronto, McLelland & Stewart, 1969) and Ring Around the Moon (Toronto, McLelland & Stewart, 1977). Note that often a children's song only has a range of an octave by virtue of a few isolated and unaccented notes; otherwise, its range might be merely a fifth or sixth.
11. In this way, the characterization of melodies as authentic and plagal can take on a practical significance for the arranger of folk songs.
12. This observation is corroborated by a survey of of first chords that are taught in guitar and ukulele instruction manuals.
13. On the use of ukulele, especially in elementary music education, see Lois Birkenshaw, Music for Fun, Music for Learning, 2nd ed. (Toronto, Holt, Rinehart and Winston, ©1977): 210.
14. I am thinking here of those who have only studied introductory harmony, which is normally taught in the idiom of the hymn- or chorale-setting. Those who have studied more advanced topics such as homophonic forms will find that the harmonic rhythm of folk songs is often quite close to what is found in the freer textures of classical composers.
15. This practice is encountered in some publications ofjazz and popular music lead sheets.
16. Similarly, chords often do not enter until the first downbeat in a classical phrase that has an upbeat beginning (or "anacrusis").
Resumé: Jay Rahn est spécialiste en l'étude de la chanson folkiorique canadienne. Son article souligne les elements qui gouvernent l'harmonisation des chansons folkioriques anglaises. Il discute les capacités at l'étendue vocales de ceux a qui sont destinés les arrangements musicaux, compare les mérites respectives des harmonies simples et complexes, et examine la structure harmonique de ces chansons. Ensuite, en se servant des examples canadiens, ii développe une analyse detaillée qui sert a démontrer comment l'harmonisation des chansons les plus communes peuvent s'accomplir par l'emploi d'un seul accord.
© Canadian Journal for Traditional Music