RATTLIN ROARIN WILLIE — FIDDLE — I
GENERAL INFORMATION ON PIPE & FIDDLE
SETTINGS
• Time signatures and Tempi
Scores are written in either 9/4 or 9/8
according to source, but the meaning is the same, reflecting conventions at the
time of writing, so the 9/4 scores are not to be played at half the tempo of
the 9/8 ones. The tempo of the “listening” versions is at the performer’s
discretion (Oswald’s instruction is Brisk, Riddell’s is Allegro),
while the tempo of the dance versions is naturally determined by the dance in
question. Atkinson’s and
• Repeats
It is difficult to be certain whether
repeat marks in the earlier texts indicate an actual repeat or simply a
strain-end marking. Later sources are not consistent, some indicating repeats,
some not. This tune may be an exception to the general observation that 8-bar
strains in 9/8 tunes are not repeated. The choice is again at the performer’s
discretion.
• Key signatures
The prevailing mode of all versions
except Wright’s D minor set is Mixolydian, the major scale with a minor 7th
leading note. This is spelt differently in the different texts, some opting for
a major key signature with natural signs for the minor 7ths, some for a modal
key signature with sharp signs where major 7ths are required. So, a setting in
D may have a D major signature (2 sharps) with C naturals indicated by
accidentals, and a setting in A may have the same apparent key signature, with
any G sharps indicated by accidentals.
• Midis
A basic midi file is given with each
score. The limitations of the midis are obvious, both in the lack of
ornamentation and in the flatly literal reading of rhythm values, with none of
the rhythmic weighting within note groupings which gives the lilt or lift
essential to performance. Their only advantage is that those who do not read
staff notation can hear at least something which relates to the discussion.
THE FIDDLE TUNE
The many fiddle settings of the tune fall
into two categories, those in D, which use all four strings and most of the
range of the fiddle’s first position, and those in G and A, which use a smaller
range more like that of the song version, with most of the tune keeping to the
highest pair of strings.
The D sets are closely related to each
other. Covering a time span of one century, they have the appearance of
episodes in the development of a single vigorous strand of the tune which grew
as it travelled. With the G and A versions there is a broad consensus among the
many short sets, but no indication of an overall pattern shared by the two
longer ones. This also applies to the longer pipe sets.
THE D FIDDLE SETS
Five D sets have been found, three in
manuscript sources:
Henry Atkinson, Northumberland, 1694
James Gillespie,
William Vickers,
(There is in addition a two-bar fragment
of the tune, probably an aide-memoire, in the Gairdyn manuscript, NLS Glen 37,
which David Johnson dates 1700-c.1740.)
and two published:
Daniel Wright,
Robert Riddell, Edinburgh, 1794
The three English sets are comparatively
plain versions while the two Scottish sets are much more ornate.
It is clear from the Atkinson manuscript
that the writer (Atkinson’s book is the work of several hands) was struggling
with his notation in this tune. It is metrically erratic and has some bars
missing, so a reconstruction in the light of other versions is offered. The
order of strains is unchanged but later versions suggest the sequence 1, 3, 4,
2.
ATKINSON SCORE AND MIDI FILE (opens in new window)
Daniel Wright’s Rantin Bille is
next in date. It is the odd one out of all versions in that it appears to be in
a mixture of D minor and D major. It is given as published, but is highly
suspect; as a
WRIGHT
SCORE AND MIDI FILE (opens in new window)
Vickers’ three strains are remarkably
close to the corresponding ones in Atkinson, but the musical idiom is now
recognisably modern in sound and appearance and this setting would be at home
at a dance or a fiddlers’ session today.
VICKERS SCORE AND MIDI FILE (opens in new window)
In both Atkinson and Vickers the
accidentals have been edited here to conform to the more literate published
Scottish versions. Both originals are reproduced on the FARNE website.
Although close in date to Vickers, and
close in geography to Niel Gow, Gillespie’s set is very different from both. It
appears to be be a direct descendant of something like the Atkinson version,
but cast in what we might call the Native Baroque style. Blending distinctly
Scottish (and elsewhere, Irish, English and Northumbrian) melodies and harmonic
structures with European methods of variation and decoration, this is a musical
current which flowed strongly in
We should not imagine any conflict
between variation sets and dance sets, even of the same tune. Gillespie himself
certainly saw none, and his manuscript makes this clear, with sections devoted
to each. Interestingly, he classes his variation sets as “Scots Tunes” and they
form the largest section of his substantial collection. (For an excellent
overview of the variation style and its relationship with 17th century art
music see Dick Hensold’s article on the subject in Out of the Flames,
published by and available from LBPS.)
Gillespie barred his Roving Willie
incorrectly in 6/8, the same lapse in musical grammar which makes many of
GILLESPIE SCORE AND MIDI FILE (opens in new window)
The last of the D sets was published by
Robert Riddell of Glenriddell in his Scotch, Galwegian and Border Tunes
(
Riddell’s set is very close to
Gillespie’s. While there are many differences of detail between them the thrust
of their musical narrative is the same: they are not sets of different
variations, but different sets of the same variations, and much can be learnt
about the flexibility of the idiom by comparing them closely. Riddell has two
more strains than Gillespie, his last being based on familiar filler phrases.
His note on the tune is illuminating:
“This Air (of which a very fine set is
here given,) is said to have been the composition of JOHN COWAN, a very noted
performer on the Fiddle, at Newton Stewart in
Riddell does not claim that the set he
publishes is as John Cowan or Peter MacNaughton played it, and this is doubtful
considering the consistency of the earlier known versions; it seems rather to
be “traditional” in that is the work of more than one hand, especially given
its close relationship with Gillespie’s earlier version.
Riddell’s bass line, given for strain 1
only, is not reproduced in the score but is included in the midi file.
RIDDELL SCORE AND MIDI FILE (opens in new window)
Taking the D sets together, some
interesting musical characteristics emerge. These are inextricably interwoven
in the tune but they can be considered as a series of four layers of
patterning, and while these do not tell us everything about the tune they are
useful guidelines as far as they go:
• Rhythmic structure (foundation)
• Harmonic pattern (framework)
• Melody (fabric)
• Overall design (form)
• Rhythmic structure: whether scored as
9/4 or 9/8, strains are 8 bars long rather than the more common 4 bars repeated
or 8 bars unrepeated (some versions call for repeats, some do not). Versions
tend to be more or less rhythmically busy; expressed in 9/8, some are mainly
crotchets and quavers while others are mainly quavers and semiquavers. The
simplest explanation for this difference is that it reflects their purpose,
whether they are for dancing or listening to.
• Harmonic pattern: this is implicit in
all versions and explicit in those where a bass line is given. The bass lines
from three disparate sources in dfferent keys all agree, showing that all the
18th century arrangers who wrote a bass line heard the same pattern, and in
Riddell’s case, where the bass line is given for strain 1 only of a variation
set, assumed that it would underpin all following strains. Expressed as chord
symbols in the key of D, it is:
||: D / / | C / / | D / / | D / / |
D / / | C / / | D / / | D A D :||
and for the sets which follow
in G:
||: G / / | F / / | G / / | G / / |
G / / | F / / | G / / | G D G :||
and in A:
||: A / / | G / / | A / / | A / / |
A / / | G / / | A / / | A E A :||
(The bass E on beat 1 of bar 8 of the Gow
version below makes a 2nd inversion A chord with the melody.)
Line 1 has the very common 3:1 ratio of
“home” and “away” chords in one of its two simple forms, XYXX (the other simple
form is XXXY). Line 2 is a close echo apart from the last bar. In the 2nd beat
of the last bar the dominant or V (rather than the subtonic or bVII) chord is
important in relation to the pipe versions which follow: the literate Scottish
sources are nearly all in agreement in this respect, and anyone who might think
that they were using the dominant to make the tune more harmonically
“respectable” will find plenty of examples where the same sources are happy to
stay in the mixolydian mode throughout. (Disblair is the partial exception in
that his melody at this point implies the subtonic chord in some strains and
the dominant in others.)
Bar 4 of strain 3 of the Gillespie and
Riddell sets can be harmonised:
| D G D |
or, transposed to A:
| A D A |
This is a common minor deviation from a
strong basic pattern, and is also hinted at by the harmony in the lyra-viol
version at this point. Disblair’s and Oswald’s sets (see: Fiddle - II) are less
tied to the sequence than any of the others, but while their melodic invention
frequently carries them away from it, mainly at strain-endings, it is still
felt as the background or default pattern.
• Melody: this is both more sophisticated
and more consistent in the D sets than in the others. Bar 1 consists of a
melodic shape or phrase (M1) on the tonic chord, and bar 2 of a contrasting
melodic phrase (M2) built on the subtonic chord. Bar 3 is as bar 1, while bar 4
repeats the melody of bar 2, but a step higher on the tonic chord. Bar 5 is as
bar 4, bar 6 as bar 2, and a concluding or tag phrase (T1-2) forms bars 7 and
8. This sounds complicated, but it is very easy to hear and sense. It can be
expressed diagrammatically:
||: M1 | M2 | M1 | M2 |
| M2 | M2 | T1 | T2 :||
It is noteworthy that in many of the
lyrics given earlier line 5 is an exact repeat of line 4, just where the tune
repeats a bar, with the same melody on the same chord. The recurring M1s and
M2s are not of course exact repeats of each other, and they are more obvious in
some strains and in some versions than in others. In strain 1 of all the D
sets, for example, M1 begins on the 5th of the relevant chord, and M2 on its
high tonic, and the end of the bar changes according to what follows it. In
Vickers’ strain 2 (Atkinson’s strain 3) M1 and M2 are the same. There is
variety in the tag phrases between versions, but beat 2 of T1 is nearly always
note 6 of the home key (B in D, F# in A).
• Overall design: this is as much a
practical as an aesthetic consideration. A listener may be mystified, hearing a
variation set, as to how the player remembers his or her place, but a formal
template held before the performer’s mind keeps the content and its sequence in
place. Of course, whether the listener is bored or interested at any one time
depends not only on the listener’s quality and span of attention but also on
the musical detail; once a workable design or form is established, the detail
can be varied or improved just as in telling a story, where the narrative
outline is known by the teller but the detail changes from one telling to the
next. Form can of course degenerate into formula, so some variety of form
between different tunes is desirable.
With this tune we need to look beyond the
individual versions and try to read between the texts to discern a formal
design. It is possible to discern a pattern of alternating odd- and
even-numbered strains where the first complete bars of the odds begin on the
5th degree of the scale and the evens on the 8th or 1st, the high or low tonic.
We might then expect all these sets to be comprised of multiples of two
strains.
Wright and Vickers have three strains,
but as these three begin, in both cases, on 5, 8 and 5, the alternation is
still present as far as it goes. Gillespie and Riddell are interesting in that
their strains 1-4 begin on 5, 8, 5, 1, but thereafter they do not keep to an
alternating pattern. I would suggest that their strains 1-4, whose ancestors
are in Atkinson’s version in a different sequence, and in Wright’s and Vickers’s
versions in the same sequence as far as they go, represent the core of this
strand of the tune, and that it has a satisfying and easily memorable formal
pattern; the composers who added extra strains followed the harmonic framework,
but placed them in simple pairs of alternating lower and higher registers.
We may make a guess at the date when this
strand originated. If we suppose that the song tune with its relatively
restricted range is the earliest form of the melody, dating from or soon after
the time when Rattlin Roarin Willie was alive, then as well as being piped and
sung, it could have been played on a variety of instruments including the older
“fiddle”, the instrument that was displaced in Scotland by the modern Italian
violin around 1670-80. The strain which uses the bass string seems to be made
for the new stronger-toned violin rather than the old fiddle, but this is of
course conjecture. If John Cowan of Newton Stewart was an old man by 1725, he
might well be the fiddler who, some time around 1680-90, composed (“put together”) not the original tune
itself but a version of it which moves the main melody to the middle two
strings and gives scope for contrasting strains, higher and lower, on the outer
pairs. This version could easily have found its way to Atkinson in
Northumberland. Cowan’s claim to authorship, supported by the tradition that
Riddell recorded, is plausible but not provable. Whoever designed the setting,
the written record shows it to have been durable, mobile and fertile.