paul roberts lecture

Anyone looking through the wealth of material in these old English fiddle manuscripts must be struck how different it is to the music of recorded 20th century English fiddlers, and must inevitably wonder about how it was played. The following two essays represents an initial attempt to answer that question. The first was given as a talk at Sidmouth Folk Festival in August 2000, the second as a talk at the Aberdeen "North Atlantic Fiddle Convention" in August 2001. They are presented here in their 'raw' state, for which no apologies. It is the author's intention to eventually produce a more thorough, better organized, academically conventional and reader-orientated piece, but it is felt that the issues and ideas raised need an immediate airing.

The Village Music Project Lecture - Sidmouth Festival, 10th August 2000given by Paul Roberts

English fiddle styles 1650-1850: reconstructing pre-Victorian technique

The period from 1750 to 1850 was the age of the Industrial Revolution and it saw dramatic, fundamental and unprecedented change in every area of English life. Much of this change was later and more abrupt than is often realized, especially in the area of popular culture, where something of a sea-change seems to have taken place around the 1840s and 50s. This took many forms. We find the culture of the chapel replacing that of the pub; trades unions and political parties taking over from riot and terrorism; the works of Samuel Smiles supplanting the chapbook lives of criminal heroes; and in popular music we see brass bands and light orchestras replacing village string bands; polkas, waltzes and imported ballroom dances ousting jigs, reels and hornpipes; accordians and concertinas replacing fiddles and bagpipes. Thus English traditional music as documented during the 20th century was dominated by Victorian ballroom dances and commercial pop music typically played on some form of free-reed instrument.

But if we go back to the pre-Victorian era we come across a repertoire of jigs, reels and hornpipes similar to the one we now associate with Irish and Scots music and we find the dominant instrument is the fiddle. Elements of this older music were preserved by 20th century musicians as an accompaniment to certain specialized dance forms that survived from the pre-industrial era, notably step-dancing and ceremonial dances like the morris, but It was often valued more highly by collectors than by the musicians themselves.

Ultimately 20th century musicians are a poor source for pre-Victorian English vernacular music, and we definately have to be cautious about extrapolating too much about pre-Victorian playing styles from the playing of post-Victorian English fiddlers. There are two main sources for the old music, printed collections of Country Dance tunes, which come in a steady stream from the 1650s through to the early 19th century, and the Manuscript tune books kept by fiddlers themselves. There are serious limitations to both these sources.

Printed collections were aimed at professional musicians and dancing masters eager to exploit the fad for Country Dancing among the post-Restoration gentry. They adapted whatever music they could make fit the bill, including large numbers of jigs, hornpipes, and reels, essentially solo step-dance or small-group dance tunes from a lower-class oral culture. Tunes may well have been subtly altered and amended to give a more polite tone - tune titles certainly were - and doubtless distinctly regional or non-standard music was ignored or standardized. And whole genres of music intended for listening or complex stepdancing are inevitably absent - for example long variation sets, improvized descriptive pieces, and instrumental airs. Still, these books preserved a vast body of music which would simply have been lost otherwise. .

As for the manuscript tune books, those recovered so far show a definate social, chronological and even regional bias. They were typically the work of a small and distinctive section of the population, a working-class elite of independant craftsmen, and most of them date from the 40 or so years at the very end of our era - mostly from around 1800 to 1840, and especially the 20 years from around 1820 to 1840, which presumably reflects an advance in literacy among this group at this time. On current findings they also show a definate regional bias towards the north. The repertoire in these books is very similar throughout the country with a great many tunes occuring repeatedly - some so regularly as to form a distinct "hit parade". It consists of an eclectic mix of older indigenous forms such as jigs, reels, and hornpipes and more recent imports like cottillions, quadrilles and waltzes, and there is a strong strain of both religious music and light classical music running thoughout. The oldest native forms - 3/2 and 6/4 hornpipes, variation sets etc. - are rare, as is purely regional music. All this presumably reflects both their time and their social group. Overall the content is considerably less "folky" than that of the much smaller number of 18th century manuscripts, and only partly parallels literary references to the repertoire of country fiddlers of the period. For example, Henry Hobson's poem The Northern Minstrel's Budget lists a staggering 232 tune titles allegedly played by a Northumbrian fiddler of about 1800 - the point of the poem is to cleverly rhyme the titles.

Whilst there is a considerable overlap with the material in earlier Northumbrian collections like Dixon (1733) Vickers (1772) and Peacock (1805) there is only a slight match with the material in the John Moore MS, a typical artisan fiddler's collection from the same area dating to around 1840. John Moore of Tyneside appears closer to John Moore of Shropshire than to Hobson's "minstrel" or the earlier Northumbrian collections. So we cannot presume these books accurately reflect the music of every social and regional group, and they certainly don't reflect the music of the entire pre-industrial era. One thing we can be sure of is that the archetypal fiddler of 18th century art and literature, the blind itinerant playing in street, ale-house, or farm kitchen, didn't write his music down. After all, the typical 20th century traditional fiddler was able to read and write but in general remained musically illiterate. Indeed, my experience is that most non-classical musicians are still musically illiterate or at best semi-literate, despite often being highly educated, and despite having endured school music lessons.

So it seems safe to presume that in pre-Victorian England with its low levels of popular education the typical lower-class musician was musically illiterate. Now, imagine if all that was known about 20th century popular music was what was written down! There are certainly enough tantalizing hints in the the mass of printed collections and manuscript books and from literary sources to suggest that there was a lot more to the music of the era than we will ever know. Nethertheless it is my belief that by careful use of a wide range of sources we can still find out a lot about this music - above all about how fiddlers actually played it. It is my intention to use a variety of sources to reconstruct some key elements in the technique and playing styles of English fiddlers in the period roughly from 1650 to 1850. These sources include the playing of 20th century English fiddlers, the observations of contempories in art and literature, and above all the internal evidence of the fiddle manuscripts and printed collections themselves.

There is also much value in a comparative approach involving both neighbouring traditions and Baroque Art music. English, Irish, Scots and American music were a lot closer in the 18th century than now, as were popular and art music. Basically I am applying what historians call the "triangulation" method whereby different kinds of sources are used to confirm each other rather than relying on one type of source alone.

Lets begin at the beginning. For most of this period the violin itself was a different instrument from the one we know - its neck shorter and angled differently, the bass bar lighter, the sound-post thinner, the bridge flatter, the chin-rest not yet invented, the strings made of gut, the bow shorter and without the modern curve. The modern violin and bow have evolved over the years to meet the demands of changes in classical music, and most of this development took place towards the end of our period and was not instantly accepted by everybody. There is nothing intrinsically superior about the modern violin and bow anymore than an areoplane is superior to a car, or a car to a horse - the modern instrument simply reflects the demands of modern music, and ideally one would use an 18th century fiddle and bow for 18th century music. I presume few of you will want to go to this extreme, yet there's no doubt the baroque instument sounded different - it was apparently a lot less legato for example - and aspects of your technique would probably change if you habitually played one. You may, however, want to adapt aspects of baroque technique to the modern instrument, something traditional fiddlers have been happily doing for the past 200 years - holding the fiddle against the chest for example, or under the right not the left side of the chin, gripping with the left hand not the chin, sloping the neck downwards, holding the bow further up the stick or in the "French grip" (thumb under the frog), playing with a flatter bridge etc

Though we now associate these things with old country fiddlers they were shared with classical players as well in the 18th century. One technique which seems unknown to classical violin at any stage but was formerly a common feature of folk fiddling in both Britain and Europe is the practice of moving the fiddle instead of the bow. Some east European fiddlers still play like this. This was probably something carried over from the medieval fiddle. I would certainly recommend those of you wedded to modern classical technique to experiment with a more relaxed stance and different bow positions. For example the tight under the chin grip is bad for the neck and shoulders and only essential if you regularly move out of first position. Or again, traditional fiddlers often hold the bow nearer the middle. It seems to be easier to do fast string changes and to play rhythmically like this, possibly because it's closer to the point of balance.

However, there is probably nothing in the modern classical stance which is totally incompatible with this music. The real lesson of baroque technique is "do what you will" - the attitude was simply less rigid. If there is one area where modern violin and traditional fiddle technique do appear seriously incompatible it is what we might call the primary aesthetic.Basically, the classical player gives priority to creating a particular tone - in the past somewhat hard and full, based on the human voice, nowadays rather sickly sweet, based apparently on a jar of syrup - and wants to play cleanly and precisely. The traditional dance fiddler is more concerned with rhythm, energy, and volume. These two approaches can be quite difficult to reconcile. Loud, rhythmic, dynamic fiddling tends to work against sweet tone and clean playing, and vice versa. Most recorded fiddlers - not just from English tradition - play with a hard, edgy tone, and some are positively harsh and scratchy. And whilst some are noticeably clean players, others seem quite unconcerned by squeaks and grunts, and even consciously exploit them as decorations.

All of this tends to shock the classically trained vioinist, but given that traditional fiddling is generally closer to baroque technique than is modern classical technique, one suspects that Geminiani and Corelli sounded more like Jinky Wells or Jake Hutton than Vanessa Mae. Nethertheless, there is great variety in tone, clarity and precision in recorded English fiddlers, and this was probably so 200 years ago as well. What we can be sure of is that the tonal model in baroque violin was not a tin of syrup but the human voice. OK, we've got our fiddle, baroque or modern.

Time to get it in tune. I raise this issue because pre-Victorian fiddlers did not always play in standard tuning. The medieval fiddle and related instruments like the crowd used a variety of tunings and some of these were carried over into folk violin - indeed, in the 18th century even classical music occassionally used them. In classical tradition they are referred to as "scordatura" tunings, but Anglo-American fiddlers call them "crosstunings" . Three of these crosstunings occur in the old English collections, these are high D in which the G string is raised to A, used mainly in D tunes; AEAE, the commonest of all, used mainly for A tunes; and Csharp - the same as AEAE except the top E is dropped to C-sharp to give what the Ozark fiddler Lonnie Robertson describes as a "weary blend". These three cross-tunings are also documented in 18th century Scots music and were carried over to America by English and Scots settlers. In some districts of the American south fiddlers only ever played in cross tunings and US tradition includes a great many more tunings than these three. It is hard to say how common or widespread crosstuning was. Standard tuning evolved with the violin itself and would have come over with the first violins.18th century classical violinists and 20th century fiddlers both in England and in related traditions for the most part used standard tuning. Cross tunings are only occasionally mentioned in the old books. So it would seem a pretty good guess that pre-Victorian fiddlers for the most part played in standard tuning. But the American example suggests that some individuals and certain social and regional groups would have seen it it as no more than one tuning amongst many, some possibly using crosstunings most of the time. This would not really be documented in the tune books because cross tuning presents serious problems of notation and in any case the books reflect the playing of the most musically literate and classically influenced fiddlers. The American example would suggest that crosstunings would have been commoner lower down the social scale and at the geographical margins, precisely where one might expect pre-violin techniques to survive best and precisely where they would be least documented.

There is also some evidence that these tunings become more common the further back you go. They are mentioned in early collections like Playford and Marsden's 1705 collection of Lancashire Hornpipes but the latest I've found in an English printed collection is 1769. But again, this may reflect nothing more than people giving up on trying to notate them. Johnson notes that in Scotland most written crosstuning is relatively late, dating to the second half of the 18th century, and that this seems to bear no relationship to its real incidence before or after this period. The advantages of these tunings are threefold - firstly the fiddle becomes louder, an important factor for a dance player in the days before amplification, secondly the tunes become easier - in A-tuning (AEAE) for example you don't have to stretch for the sharp G on the D string. Above all a cross tuned fiddle sounds better than standard because of the reverberation of the sympathetically tuned strings. Here's the Ozark fiddler George Helton recorded in 1956 playing a typical 18th century single reel he calls Jenny Nettles in AEAE tuning - not the Jenny Nettles which occurs in many of the old English and Scots collections but a melody definately out of the same stable. Note the use of pizzicato, not unusual in cross-tuned pieces. (George Helton, JENNY NETTLES. From The Old Time Fiddler's Repertory, Missouri Old Time Fiddler's Association 107/8) (Tune file not yet available)

The novice fiddler's first question is usually "what do I do with the bow?" When I was learning, received wisdom in the English music revival was that English traditional fiddlers always played one bow stroke per note. I suspect this idea was invented by melodeon players because it gives a choppy sound that blends well with the melodeon. In fact it was instantly apparent to me from listening to recordings of English fiddlers that very few of them actually played this way - although working out exactly what they were doing wasn't easy, because single note bowing is the only technique which is easily discernible on tape, the moment people start slurring notes it gets hard to follow what's going on. In the case of pre-Victorian English fiddling, however, we are fortunate in that many of the players actually wrote bowing patterns down. Some wrote them in detail, others now and again, some not at all. Some appear to be writing the bowings down in precisely the places where you might not expect to use a particular pattern - which if you think about it makes sense. All in all the notation of bowing was probably like the notation of decoration - fairly random. For some people bowing patterns would be something so ordinary and second-nature, or so subject to spontaneous variation, that notation would be pointless, whilst others would only put bowings down where it seemed essential to the character of the tune or otherwise important. However, there are enough bowings notated in enough collections to reach some fairly definate conclusions - pre-Victorian fiddlers habitually used a number of distinct bowing patterns, of which one stands out pre-eminently. This is the pattern American fiddlers call the "Nashville Shuffle" and it seems to have been as basic to old-time English fiddling as it is to old-time Anglo-American fiddling. Imagine a common-time tune divided throughout into groups of four quavers. The first two notes in each group are played on one bow stroke, the next two on separate strokes (demonstrate). In some of the manuscripts this is simply written as a slur on the first two notes, but others add staccatto dots above the second two, perhaps to suggest these should be accented - indeed, American players often heavily accent these two notes, especially the first one (i.e. the third of our four) for example by hitting the adjacent string on the third note, and it can be heard occasionally in recordings of English players.

The Victorian Scots violinist William Honeyman in his various publications describes the Nashville Shuffle as the basic pattern for bowing reels, and in Anglo-American fiddling it is regarded as the basic bow pattern for all common-time tunes, though in the real world many players never use it and most players don't use it absolutely literally throughout every tune. It doesn't fit into every passage of real tunes for a start. A good player would regard it as a basis from which to depart and extend, much as one might regard the basic vocabulary of a language. But I would recommend anyone who seriously wants to play the old English music to learn it, even if you ignore everything else I have to say! Of all the bowing patterns I've come across it is undoubtedly the most effective and the most useful, and gives a real lift to most common-time tunes - and to some triple-time hornpipes as well. It gives a flowing but driving feel with an accent on the offbeat than can be as slight or as heavy as you want.

Once you start using the shuffle it becomes instantly apparent that a lot of old English tunes are actually structured around it - all manner of tunes suddenly start to make more sense. The only English or Scots fiddler I know of whose bowing has been systematically analysed is the Scots border fiddler Tom Hughes. Here he is recorded in the late 1970s applying the Nashville Shuffle to the Flowers of Edinburgh. This is followed by the Northumbrian fiddler Adam Grey recorded in 1954 playing Tom Hepple's Polka. He applies the shuffle to a crotchet/two quaver pattern rather than the literal four quavers, but note the way he accents the first of the quavers and hence the offbeat. (Tom Hughes, THE FLOWERS OF EDINBURGH. From Tom Hughes and his Border Fiddle, Springthyme SPR 1005. Adam Grey TOM HEPPLE'S POLKA, from Holey Ha'penny, Topic LP) (Tune file not yet available)

Some American fiddlers occasionally use the Nashville Shuffle backwards - slurring the last two notes of the four instead of the first two - and I've seen passages like this in the old books. A simpler pattern that occurs in the old books is repeated paired slurring - in our group of four quavers one and two are slurred together and then three and four. This pattern was used for 3/2 as well as common-time - Marsden's 1705 collection of triple-time Lancashire hornpipes has almost entire tunes played this way. Honeyman describes a variant of this pattern in his discussion of the "Newcastle" or "clog" style hornpipe - that is, the "dotted" hornpipe. With two sets of four quavers the first note is played separately, then 2 and 3 slurred together, then 4 and 5 and so on. Several English fiddlers have been recorded using this pattern, on undotted as well as on Newcastle style hornpipes. On undotted tunes it creates a lightly syncopated rant sound which can give the illusion of dotting due to the short pause between each of the slurs. Unsurprisingly the pattern crops up in recordings of rant playing as an alternative to single bowing. Honeyman also claims that single stroke bowing with the normal up and down pattern reversed is an important feature of the Newcastle hornpipe style. In general the tune books rarely give any directions as to up and down bows, but some Anglo-American fiddlers habitually reverse their bow strokes, creating a sound accurately described as "digging it out".

Here's the Herefordshire fiddler Stephen Baldwin playing The Gloucester Hornpipe making heavy use of the Newcastle double slur in a performance that seems to straddle the fence between dotted and undotted, followed by the Northumbrian fiddler Jim Rutherford applying the doubled slur to a rant version of Corn Rigs. (Stephen Baldwin, THE GLOUCESTER HORNPIPE, Leader LP/Jim Rutherford, CORN RIGS, North Country Rants and Reels, Folktracks FSC-60-121

)(Tune file not yet available).

Less common but still to be found are patterns based on the 1-3 division of our group of four quavers - first note on one bow stroke, then the last three on one stroke. In the tune books this seems to have been a form of passing decoration, though as an extended pattern it can be quite efective on highly dotted music like strathspeys. American fiddlers sometimes use a variant of this which they call the "Georgia Shuffle". Here the downbow is diplaced to the offbeat so the slur joins notes four, five, and six, breaking up the natural division into fours. Somehow they seem to make it work on undotted tunes though I think it works best with dotted rhythms. (demonstrate) Bowings are less commonly notated with jigs, which is probably significant.The commonest figure we find is a jigtime variant of the Nashville shuffle. In the basic group of three quavers the first two - or occassionally the last two - are slurred. Sometimes we find passages where all three quavers of a group (or a crotchet and a quaver) are slurred together, which helps give a sleeker and more sensuous feel - it's very good at toning down the over-bounciness of some jigs.

Occasionally whole tunes make heavy use of the two-quaver slur but in general slurs seem to occur in short passages and my impression is that it was more a form of variation or decoration and that jigs were largely played with single bowing . In America single stroke bowing - whatever the time signature - is actually called "jig bow" which is probably significant. Recordings of 20th century English fiddlers seem to confirm this picture. Typically they use single bowing on jigs with occassional double and triple slurred sections. There are a lot more things to be said about bowing than I can really go into here. My feeling is I've hardly started following up the clues in the old books and in recordings of English fiddlers.

One thing that does seem clear from all these sources is that the best players varied their bowings throughout tunes - alternating passages of single bowing, shuffle bowing, and paired slurs for example - so the whole thing becomes an important part of their decorative bag of tricks.

A simpler way of using the bow to decorate a performance is the playing of drones and chords or "double stopping". This was a fairly prominent feature of 20th century English fiddling. Drones are particularly useful for the solo dance player because they double the volume as well as adding colour to the performance. And if you use a repetative bowing pattern the droned strings become almost a separate rhythm instrument. English fiddlers recorded in the 20th century ran the whole gamut, from those like Fred Pigeon who played a continuous bagpipe-style drone on almost everything, to those like Ned Pearson who played almost entirely on single strings. Most come somewhere in between, with southern English fiddlers generally leaning towards the Fred Pigeon extreme and northern fiddlers towards the Ned Pearson extreme, though there are too many exceptions to see it as a purely regional thing. Indeed, some of the border fiddlers varied their use of double-stopping dramatically from tune to tune. Jake Hutton for example was recorded playing with both a constant drone and playing totally single string. Tom Hughes played his working dance material with lots of what he called "double string work" but he played his competition hornpipes in a sparse single-string style - because judges in contests didn't like double-stopping which was seen as old-fashioned and incompatible with "good" (i.e. classical) technique.

In fact throughout the British Isles, Europe and America droning and heavy double stopping does tend to be regarded as a particularly old-fashioned thing, and there's a lot of evidence to support the idea. It's generally commoner among older players, those from remoter areas, and those less influenced by classical or commercial popular music, while single string playing is commonest among younger players and the more classically influenced. This is so much so that it is tempting to imagine that all 18th century fiddlers played with continuous drones or at least heavy double-stopping The reality was probably more complex.

I've no doubt that droning and heavy double-stopping was even more common than in the 20th century because the violin took over from the medieval fiddle on which drones and double stops were apparently the norm - some medieval fiddles even had reverberating drone strongs like the hardjanger fiddle. Moreover, the chief rival to the fiddle was the bagpipe, and effective competition must have involved imitating the bagpipe's much loved drone. Indeed, one sometimes comes across specific references in the writings of the time to the bagpipe-like droning of fiddlers, which seems to have been seen as something rustic and uncouth. However, when the violin came over here the Italian single-string sonata style must have come with it and been adopted by some fiddlers. Looking at the music itself all we can say is some tunes cry out for drones or are highly compatible with double-stopping whilst others - for example some of the complex variation sets and the later competition hornpipes - would be much easier to play cleanly on single strings.

It is perhaps significant than in recorded 20th century English fiddling heavy droners rarely played the more complex hornpipes while single-string fiddlers showed an equally strong prediliction for them. Interestingly, exactly this distinction is recognized in the USA, between drone inclined "breakdown" fiddlers and single string inclined "hornpipe" fiddlers - indeed, it is enough to favour single-string playing and complex, notey tunes to be classed as a "hornpipe fiddler", actually playing hornpipes is optional. A perfomance can be decorated with fingering as well as the bow. In this respect 20th century English fiddlers were generally quite unadventurous, making sparing use of a few fairly simple gracenotings. However, the old books make clear that some pre-Victorian players habitually used a wide variety of often quite complex gracings, including single gracenotes above and below the melody note, long semi-quaver runs between melody notes, the rapid movement of the bow the Scots call the "birl" - the same note rapidly bowed 3 or 4 times - plus all the standard baroque decorations like the Mordant, the Shake, and the Turn. The mordant is when you tap the note above or below after landing on the melody note, the Shake or Trill is the repeated beating of the note above or occassionally below the melody note, perhaps the archetypal baroque decoration (demonstrate). Vibrato in this period was regarded as a variant of the trill and was thus only used as an occasional decoration. The Turn appears to have been particularly common - you'll be familiar with this because its the movement Irish players call the Roll. It's played by hitting first the note above then the note below the melody note

The mordant, shake and turn were shared with baroque art violin and were represented by standardized symbols above the stave. Contemporary accounts make plain they could be given a variety of speeds - modern Irish players tend to do these gracings very fast but in their heyday this was only one of several approaches including dividing the time equally between each of the grace notes and the melody note. Played like this, or played between notes, they start to become indistinguishable from the long semi-quaver runs that also feature in the old books. Different gracings could also be run together to extremely elaborate effect. For example the trill was often followed by a mordant, an effect like an extended roll - in art music this was called the trilled turn or the turned shake. Another baroque decoration that occasionally surfaces in modern Irish playing and is common in Anglo-American playing is sliding up or down onto a note. It occurs very occasionally in 20th century English fiddling - Ned Pearson was recorded doing it occassionally and I have a recording of old Peter Beresford doing it just once. All this would seem to indicate it was probably a feature of old English fiddling, but it's not the sort of thing people would be likely to write down. It would seem in fact that there's nothing in modern Irish or Scots fiddle decoration that doesn't occur in old English playing, and there's a good deal more besides. What's not immediately clear is how extensively these decorations were applied - just how elaborate was the norm. On the one hand we have the fact that 20th century English players were very sparing in their use of gracenotes, and so are the descendants of 17th and 18th century English immigrants in the American south - people who have maintained a strong and unbroken fiddle tradition dating back to our period. And most of the old books contain no gracing indications or only moderate amounts. Yet a number of books do show quite elaborate combinations of gracings and runs. Moreover, as with bowing, these are the sort of tricks people use spontaneously and inconsistently and one wouldn't expect the true levels of decoration to appear in books. The reality was probably a range of decorative complexity from very plain to very elaborate, encompassing a variety of regional styles themselves subject to family and individual preferences - basically the same situation as in recent Irish and Scots playing. However, that elaborate decoration was fairly common does seem clear.

To demonstrate the point I would like to play you something I consider little short of miraculous - an actual recording of an English fiddler made in the year 1810! No I don't have a time-machine - but in 1810 Admiral Parry organized an expedition to the north pole and decided to take a mechanical fiddler to replace the live example that was usual on board ship - doubtless he thought it would be less troublesome. To this end he commissioned a pipe organ programmed to imitate the classic trio of fiddle, cello, and tambourine playing what reads like the tune listing from one of our artisan fiddlers' manuscripts. Here it is playing The March in Bluebeard, one of the smash hits of the era, followed by two popular reels, Lady Montgomery and Lord Howis. (Parrys barrel organ Saydisc LP)(Tune file not yet available).

As well as rolls and other baroque gracings Parry's mechanical fiddler uses semiquaver runs of varying complexity to link notes. This is quite common in the old books. It occurs in all types of music but was particularly a feature of the variation set which we will discuss in a minute.

By the 20th century English fiddlers seem to have dropped the idea but some Irish fiddlers and a few English wind players like the piccollo player Billy Ballantyne continued to use them. At times these could develop an almost anti--rhythmic character - this seems particulary the case with jigs and possibly dates from around the mid 18th century as 6/4 tunes were gradually converted to 6/8 - it seems people often kept the same semi-quaver runs without amendment for the different timing. In the hands of a master the resultant clutter is quite beautiful. And here are three masters, first of all the great small piper Tom Clough with his brilliant 1930s 78rpm recording of the jig Holey Hapenny .........(Tom Clough, HOLEY HA'PENNY. Reproduced on Holey Hapenny Topic LP and recently on The Northumbrian Small Pipes, Topic TSCD487).(Tune file not yet available).

To give an idea of how it must have sounded in the hands of a fiddler, here's short snatches of two Irish players recorded in the 1950s, Patrick Kelly from Clare and Patrick Ahearne from west Limerick. You'll hear a lot of extra notes in the A part of the first track and the B part of the second........... You won't hear playing like this down your local Irish session.........(Patrick Ahearne, JIG/Patrick Kelly, BANISH MISFORTUNE. Probably recorded by Ciaran MacMathuna , Taken from RTE Radio in the 1970s) (Tune file not yet available).

Runs forms a bridge between gracenoting and another form of decoration - variation of the melody. There are numerous approaches to this, from the relatively simple alteration of a few notes that most fiddlers introduce occassionally, to deliberate interference with basic conventional structure - cutting out notes, bars or passages or introducing extra ones to subtly destroy the "correct" phrasing and length of a strain. American fiddlers call this "crooked" playing and It is very common amongst recorded English players. It is less common in the manuscript books, which were inevitably compiled by people with a greater degree of formal musical education and subsequent rigidity, but it certainly does occur and its a shame that in many of the published versions of the old collections both crookedness and syncopation have been edited out. Modern listeners tend to interpret "crooked" playing as accidental mistakes by the uneducated, but a sense of structure and timing is nothing to do with formal education. In America a good crooked player - whose rule-bending is so subtle you don't notice it till you go mad trying to learn the tune or fall apart trying to play backup - is highly respected. "Crooked" playing is a close relative of syncopation, which interferes with conventional timing and rhythm rather than overall structure.

Syncopation is more amenable to spontaneous introduction and even less likely to have been notated - in any case it is notoriously difficult to notate clearly and accurately. Written versions of US fiddle tunes generally don't even try though American players frequently use syncopation. For example, American fiddlers habitually syncopate the tum-tum-tum ending of hornpipes to tuuuum-ta-tum but I've never seen this written down. It's my belief that many of the apparent mistakes in the old manuscripts are actually attempts to notate syncopation - and my ABC computer programme seems to agrees with me! I'm not suggesting the old players used jazz levels of syncopation but the light syncopation typical of Anglo-American fiddlers is probably a guide. Here is the wonderful Harry Lee recorded by Ken Stubbs in Borough Green, Kent in 1962 playing the Flowers of Edinburgh. This is not only seriously and tastefully crooked it also features a beautifully executed syncopation using a couple of pizzicato flicks on the repeat of the A-part - if it still counts as the A-part, we seem to be past the regulation 16 bars by then! (Harry Lee, FLOWERS OF EDINBURGH. From private tapes made by Ken Stubbs. Also reproduced on Boscastle Breakdown, Topic LP) (Tune file not yet available).

Beyond such tinkering with the basic structure and rhythm lies the playing of complex variations on the melody or around the chord structure that amount to the introduction of new strains. In the 17th and 18th century these were often referred to as "divisions" because one technique was to divide the notes of the melody. Up to around the middle of the 18th century "division" playing was a widespread practice and many written sets survive, usually containing around 6 to 12 strains, though 20 or more strains are not unknown. A few variation sets crop up in the early-19th cent fiddle manuscripts, though it is thanks to small-pipers that the form has lived on into this century.

Those of you who have seen John Offord's "John of the Greeney Cheshire Way" or who are familiar with the Northumbrian pipe tradition will understand what I am talking about, for the rest of you here is Colin Ross and Carol Robb playing Cut and Dry Dolly from Peacock's collection of 1805. Sorry about the two glitches where I pressed the wrong button! (CUT AND DRY DOLLY. From Cut and Dry Dolly, Topic LP) (Tune file not yet available).

In case you're thinking I've strayed from the issue of style and technique into the area of repertoire, I should stress that while numerous variation sets were standardized, written down, and memorized, the form involved varying degress of improvisation. Many written variation sets are very similar over time and place. For example it is remarkable how close some of Tom Clough's settings are to those in Peacock's collection of 1805 and the William Dixon manuscript of 1733. Nevertheless, no two sets are ever identical as the form was particularly susceptible to the folk process including conscious improvisation. David Johnson comments "....Some fiddlers would have habitually left certain strains out...others would have added new strains....Strains would have been shuffled in order...Long variation sets were generally the work of several hands...arrived at by a process of trial and error, composed partly on paper, partly on the instrument, and partly through improvisations that were later remembered and transmitted". In fact the biggest crack of all was not repetition of someone else's set or even of your own but the but the spontaneous improvisation of variations.

In the 17th century several "idiot's guides" were published to help the less talented to fake this. Indeed, in Simpson's the Division Violist of 1659 he talks about improvising the basic melody as well as the variations! You would get the bass instrument to lay down a bass line or "ground", play it through a few times, and then improvise your tune and ever more complex variations over the top. It seems two or three fiddlers would sometimes improvise together Dixieland style, or take breaks in swing style, and Simpson refers to the practice of calling out "breve" or "semibreve", very like the jazz practice of calling out "fours" where each player takes four bars of the tune in turn. Improvised variation playing probably went back back a long way. The 17th century composer Mathew Locke refers to "the tearing of a consort into pieces with divisions, an old custom of our country fiddlers". Variation playing was maintained by Locke's "country fiddlers" into the early 19th century, though it seems to have become much rarer around the mid to late 18th century, probably because the money now lay in simple two part Country-Dance settings and because as more fiddlers became musically literate they began to ape the rigidity of classical tradition and abandon improvisation. Variation sets in the early 19th century fiddle manuscripts aren't usually as complex, they sometimes introduce changes in rhythm and tempo inspired by the classical sonata, and the principle seems to have been more often associated with airs, particularly the popular "Folk Baroque" form, than with dance music. It's hard to say how far they still retained an element of improvisation, but there are a few accounts from early 19th century Scotland that specifically describe extempore variation playing and it seem reasonable to presume it wasn't totally dead south of the border. Here is a 1930s 78rpm recording of Tom Clough playing his variations on the Keel Row, which is a shortened and simplified version of the setting in the manuscript of the Tyneside fiddler John Moore, circa 1840 .........(Tom Clough, THE KEEL ROW. Reproduced on Holey Ha'penny, Topic LP. Recently reproduced on The Northumbrian Smallpipes, TopicTSCD487) (Tune file not yet available).

Related to the variation set and like it involving varying degrees of improvization was the playing of descriptive pieces like The Fox Chase. These were rarely notated, not least because they tended to rely on trick effects. I know of no recordings by 20th century English fiddlers of this form, though, as so often the case, it can still be found in America.

In general the whole idea of improvisation survived better among among American fiddlers and in the 20th century blossomed into the Country Swing and Bluegrass idioms - and indeed into Jazz itself, which was originally string music. Here is the very first Country Music record, the stunning 1922 recording of Sally Goodin by the Texas fiddler Eck Robertson. It's hard to imagine this led to Garth Brooks......This is undoubtedly a variation set, though Eck's variations don't have the complexity of the classic division form. In particular it's missing the long, flowing runs, but it does feature pedal variation - repeated jumping off a lower note - a technique that was very popular among 18th century fiddlers and pipers. Eck always maintained he improvized this in the studio - which probably means he drew on stock strains and phrases without a preconceived plan, as with most jazz improvization. (Eck Robertson, SALLY GOODIN. From Old Time Southern Dance Music Vol.2, Old Timey OT)(Tune file not yet available).

Finally we come to the soul of any performance - the whole area of keys, modes, rhythm, timing and tempo. This might seem the area where it is hardest to come to firm conclusions without a time machine, yet conventional musical notation is designed to give definate instructions in this area. The indications are that pre-Victorian fiddlers had a more flexible attitude and a much wider range of options than is general in revivalist circles today.Perhaps this is best examined by looking at the main instructions as they appear on the the musical stave:- Keys and Key signatures: The fiddle is easiest to play in the keys of A, D, and G major, and A, B, D and E minor. All these keys occur in the old English collections, as well as G minor, C minor and the harder major keys of C, F, B-flat and E-flat. Yet in general English revivalist fiddlers rarely play outside the keys of G and D, a quite incredible kow-towing to the tyranny of the melodeon. It has been general practice for some time to transpose tunes in other keys into the melodeon keys - the leading fiddle key of A has been particularly cruelly treated. But there is a reason why the old players put particular tunes in particular keys and you will almost certainly lose something by transposing them. 'A' for example has a particular high screaming quality and allows extended droning on the same string in a way that is not possible in other keys except on the low octave - the Eck Robertson tune you've just heard is a classic A piece and would have been a totally different piece of music in another key.

I recently had a quick scan through the hornpipe section in the Dragonfly edition of the Lawrence Leadley collection from Helperby in north Yorkshire. Of 45 tunes approximately 1/4 are in G, 1/4 in D, 1/4 in A, and 1/4 in the so-called "unnatural" keys of C, F, and B-flat. I certainly wouldn't claim that those exact proportions stand throughout all the collections, but there is no doubt that the correct proportions are not reflected in todays playing of this music.

Given the general flexibility of pre-Victorian fiddling, it is all the more striking that the same tunes tend to occur in the same keys in different collections, as if the ideal for each tune was soon found and agreed. Flexibility is more apparent as regards modality. Different settings of the same tune appear with quite different modality and Jeremy Barlow's work on Playford found many cases of the same tune being awarded quite different modality in different editions. This may be why some collections happily leave out key signatures. Time signature: variety and flexibility also comes across in the area of meter or 'measure' as it was called. Pre-Victorian fiddlers used a greater variety of dance rhythms and time signatures and there was much more overlap between them. At the start of our period jigs were commonly in 6/4 or 9/4, by the end they were commonly in 6/8 or 9/8 or 12/8, but for a long time all these different rhythms co-existed and some books contain the same tune with both 6/4 and 6/8 (or 9/4 and 9/8) time signatures. Though some 6/4 tunes play very like modern 6/8, most 6/4 and 9/4 tunes are closer to the modern French bouree. In the 20th century some Scottish and Irish dance bands recorded 9/8 jigs with a 1-2-3- waltz backing, a curious echo of the days of 9/4. 3/2 time, which I have heard described as "a reel and a half", was particularly popular in the early years and survived into the 19th century - these tunes were sometimes written as 3/4 or 6/8 which may have implied a slightly different rhythmic approach as both 6/8 jigs and 3/4 tunes similar to waltzes - minuets for example - were perfectly familiar in the 18th century. We sometimes find 3/2 tunes totally converted to 6/8 and even to 4/4. With common-time we often find the same tune written in 2/2, 2/4 and 4/4 though it's not apparent that this really implied a different timing. More definately, tunes we now know as reels appear in the earliest collections with the same time signature as today but doubled note values and lengthy quaver passages that clearly imply a different timing and speed (demonstrate). By the end of our period the modern form was dominant, yet both approaches to these tunes can be found at the same time and in the same collections from an early date.

All this is really the tip of the iceberg - the careful observer will often find the same tune in quite unrelated metric disguises - for example, the popular air "Sir John Fenwick" or "Mary Scott" also appeared as a minuet, a jig, a reel, a hornpipe and a waltz. . Tempo: One of several myths which long held a deadening grip on revivalist English music was the idea that it should be played at a funereal pace with a plodding rhythm. This seems to have originated in the 1970s as a deliberate attempt to place clear blue water between English music and the revivalist Irish music that was often played at headbanging speeds back then. In reality, most recorded English traditional musicians - exactly like their Irish counterparts - used a variety of tempos ranging from mega-relaxed to totally manic. Some of the greats of English music - William Kimber for example - actually played very fast. The evidence from the old collections is that a similar range of approaches existed back then. Our sense of tempo relates to the workings of the human body and probably hasn't changed over 200 years. But even if we presume our idea of "normal" tempo applied to the 18th century it could only be a median, beyond which lay a range of options depending on the tune and the occasion.

Many of the old English collections contain all the standard classical tempo codes - "adagio", "moderato", "allegro", "presto", and so on. The use of these make clear that exactly the same type of tune could be approached many different ways. If we cross the border the Gow family's publications are particularly helpful as they speak in plain English. They cite three alternative speeds to the norm, especially for the more complex or highly decorated jigs and strathspeys - "slowish", "slow" and "very slow". Overall, different forms may have tended to different speeds For example, some of the complex variation sets may have been played rather slower than two-part version of the same tune. My experience is that quite complex variations in 6/4 and 9/4 can actually be played at a fairly energetic dance speed but when these settings are changed to 6/8 and 9/8 - as became common later in the 18th century - they need to be played much slower to get all the notes in clearly. In Ireland 6/8 variation sets were known as "pieces" and Irish pipe manuscripts distinguish between a fast "jig" tempo and slow "piece" tempo. This distinction is well known to English pipers too, who sometimes play 6/8 variation sets almost like slow airs. Again, looking at the various tempo indicators in the old books I get the impression that hornpipes tended to be played faster than reels, the reverse of modern practice, though I wouldn't want to lay money on it at this stage. But in general the rule seems to have been variety and flexibility. We are talking a wide range of possible tempos for any tune type adopted according to taste, mood, or the specific melody or dance. Remember that a typical 18th century fiddler would have played for a greater variety of audiences and situations than today. Different dances done to the same tune may have demanded different speeds - think of the difference between the rather stately tempo of modern competition clogging and the wildness of some English pub stepping. Or think of the difference between 20th century rapper speed (very fast) country dance speed (normal) and Cotswold morris speed (slow). Our old world fiddler played for all age groups too. Nowadays its rare for different generations to socialize together but any DJ who does weddings knows that drunken teenagers demand faster and wilder music than old ladies and he has to cater for both. Our fiddler didn't just play for the people who go to Sidmouth ceilidhs, he also played for the people who go clubbing in Ibiza and the people who go to old-time sequence dancing at the Conservative club. And in the ale-house with his cronies tempo would have related to the mood and the amount of White Lightning taken, just as today. Tempo seems to be the area which causes most bad feeling within bands today perhaps because revivalist musicians approach the subject in a rigid and unimaginative way.

Whether we look at 20th century tradition or the pre-Victorian collections it's clear a much more relaxed attitude prevailed in the past, or at least a wider range of acceptable options. Incidentally, the subjective experience of tempo varies enormously within any individual. It has been demonstrated that the same individual will class exactly the same speed as slow, normal or fast on totally different occassions. In your heads you and Fred are probably playing the same speed, so maybe its you that's got it wrong. Accenting and internal timing: Again, all the indications are that much variety existed in this area too, though it has to be one of the greyest areas in our study. The indication of accenting by dots and tails only shows major alterations to note values. Try leaving out the largely unconscious slight accenting all players from all traditions use and play the note values literally as written and you will find the music feels quite strange. Subtle and variable accenting and phrasing that can't be adaquately notated is at the very heart of individual and regional style. This is what makes Scots reels sound so different to Cape Breton versions for example. Try playing the literal note values exactly as written with a jig then slow right down. A lot of tunes start to turn into 3/4 or 6/4 - this suggests that the gradual conversion of 6/4 to 6/8 may have resulted from subtle long-term changes to tempo and accenting. It was slowing reels down and exagerating the accenting that produced the strathspey - originally seen as no more than a regional way of playing reels.This exactly parallels what happpened with the common-time hornpipe in England. In the old books the vast majority are written undotted and play like rather bouncy reels - this is basically how most 20th century English and Anglo-American traditional musicians played them. The "dotted" hornpipe like the strathspey-reel seems to have arisen through slowing tunes down and exagerating the accenting, and to have been originally seen as no more than a regional style, variously Northumbrian or Lancastrian. And like the strathspey it ended up being classed as a separate genre, for a while at least - Victorian collections like Ryan's/Coles and Kerrs Merry Melodies distinguish between dotted "clog dances" and undotted "hornpipes". Whatever - in the old collections the occasional dotted hornpipe often appears alongside the usual undotted examples showing the style was an accepted alternative from an early date and beyond the north-east or Lancashire, which puts in perspective pointless modern arguments about the "right" way to play hornpipes. Recorded playing of all tune types shows a wide range of accenting and internal rhythm can be applied. In particular each bowing pattern has its own special feel, though the main bowing patterns are all very adaptable.

Here's three hornpipes played by the same fiddler, and all making heavy use of the same 'Newcastle' double slur bowing, but each has its own quite different rhythm and accenting because each was played for a different type of step dance. This is Harry Lee again, recorded in 1962 playing The Monkey Hornpipe, The Breakdown and The Clog Dance. There is a gradual progression from undotted to dotted, and in case its too gradual there's a quick reprise of Monkey Hornpipe at the end to remind you where we started. (Harry Lee, Borough Green, Kent, 1962. From private tapes made by Ken Stubbs) (Tune file not yet available).

It cannot be stressed enough that pre-Victorian English fiddling must have seen a huge variety of regional and local styles, not to mention the inevitable differences between individuals and the differing fashions of different periods. There is no such entity as a uniform, homogenous English (or Irish or Scottish or American......) fiddle style in this period - nor was there in the 20th century until nationally minded revivalists set about creating them. I am talking about general, widespread features from which those who wish can pick and choose to develop their own style.

It's time to summarize these general features.

  • The pre-Victorian English fiddler played with his fiddle against the chest or shoulder or under either side of the chin. He tended to hold it sloping downwards and he gripped with his left hand rather than the chin.
  • He used a flatter bridge and a shorter bow, holding the bow in a variety of ways and often further up the stick than today. He probably used standard tuning for the most part but would have been familiar with crosstunings, and some players may have habitually used them.
  • He would have played mainly in G D and A - the latter considerably more often than today - and the easy minors, but from around the mid-18th century players seem to have begun experimenting with harder keys like C, B-flat and F.
  • He played mostly in first position though shifting was not unknown and some of the more complex hornpipes required it.
  • He used a variety of bowing patterns and techniques, especially the so-called "Nashville shuffle", the "paired slur", and the "Newcastle Hornpipe" style, and he probably varied his bowing within pieces as a form of decoration.
  • He made plentiful use of drones and double stops, even more so than today, though some players will have preferred single-string bowing, especially on the more complex hornpipes.
  • He decorated with an impressive mix of gracings including single gracenotes, triplets, mordants, trills and the baroque turn, the same movement as the modern Irish roll, also the staccatto repeated semi-quavers that Scots call the birl.
  • He also used fast semi-quaver runs, some leading smoothly from one note to another, others almost breaking up the rhythm - this was particularly the case with 6/8 jigs where the fast runs typical of the 6/4 tunes were often carried over without amendment. Some at least would have delighted in playing "crooked", that is with extra or missing notes and phrases, and light syncopation was probably not unusual.

He played basic two-part jigs, reels and hornpipes but with a much greater variety of rhythms and time signatures than today including 6/4, 9/4, 12/8 and 3/2 time as well as the more familiar 6/8, 9/8, 3/4, 4/4/ and 2/4, and he had a more flexible approach to tempo, rhythm, accenting and timing, or at least a greater range of acceptable options.

For much of the period he also delighted in the playing of divisions or long variation sets, and whilst some settings became standardized the best players liked to spontaneously improvise their variations, sometimes in a jazz-like small group context. This was music for dancing, but also for public listening and personal amusement. If we remove our nationalist blinkers it is possible to use 20th century sources to imagine the sound of the archetypal English fiddler of the pre-industrial era. With his shuffle bowing, cross-tunings and preference for drones, he must have sounded very like an old-style fiddler from Kentucky or Virginia - with his use of baroque gracing and decoration he must also have sounded like the more elaborate of the western Irish players.

If you imagine a cross between Tommy Jarrell and Michael Coleman or John Salyer and Packie Dolan improvising Tom Clough's intricate variation settings of Holey Ha'penny or The Keel Row you wouldn't be far wrong. Quite mind boggling really! Of course, variety was the order of the period. I doubt if any of the performances used to illustrate this talk would have sounded seriously strange to the man on the Clapham stage coach. But none really comes close to matching our archetype. The problem is that 20th century Irish players used the elaborate rolls and gracings but not the Nashville Shuffle bow and the drones, English and American players used the drones and the shuffle bow but not the gracings, and onlypipers now play the long variation sets. But the following two performances surely come very close to sounding like an archetypal Georgian English fiddler. First we have the great Longford fiddler Packie Dolan recorded in New York during the 1920s playing The Liverpool Hornpipe. It's one of the most popular tunes in the old English collections, he plays it old-style and undotted, and he not only uses the baroque gracings he uses English or American levels of double-stopping, something quite rare in recorded Irish fiddling............(Packie Dolan. GROVE HORNPIPE (Liverpool Hornpipe). Reproduced on Farewell to old Ireland, Proper CD) (Tune file not yet available).

....................Finally, a remarkable performance of The Foxhunters Reel by the Clare fiddler Patrick Kelly, recorded in the 1950s. He use birls, rolls, and other baroque gracings plus English or American levels of drones and double stops, he plays in the classic pre-melodeon fiddle key of A and - incredibly - in crosstuning, the only case I know of its use in Irish tradition. The tune itself is a typical 18th century single reel of the type so popular in the old English collections and though its not a classic long variation set, with its five distinct strains it comes closer to it than most modern fiddle tunes. This is surely very close indeed to the sound of the fiddler Admiral Parry sought to replace with a machine, and a fitting conclusion to this talk. Sorry about the abysmal quality but I recorded this off the radio over 20years ago..(play Paddy Kelly THE FOXHUNTERS REEL. Probably recorded by Ciaran MacMathuna in the 1950s. Taken from RTE radio in the 1970s)(Tune file not yet available).



"Crossing Boundaries" - North Atlantic Fiddle Convention, Aberdeen, July 2001
English fiddling 1650-1850: reconstructing a lost idiom and beyond

The popular dance music of pre-Victorian England was dominated by the fiddle and a repertoire of jigs, reels, and hornpipes similar to the one we now associate with Scottish and Irish tradition. This rich musical culture was largely swept away in the middle decades of the 19th century by a wave of new music - by brass bands, accordeons, and imported ballroom dances.

Sources for this older music are fragmentary and limited. Period art and literature contains scattered information. 20th century English fiddlers, though rarely recorded, displayed many archaic traits within a basically Victorian repertoire. The hundreds of 'Country Dance' collections published between 1650 and 1850 are an invaluable source, though they were largely aimed at professionals working the gentry market and only document a limited area of vernacular music-making. Above all, the manuscript tune collections compiled by some of the old fiddlers themselves open a very direct window into the world of pre-Victorian fiddling, though even this source has serious limitations, in particular a strong social, regional, and chronological bias.These books were typically the work of a distinctive minority, a working-class elite of independant craftsmen, they mostly come from the north, and they date overwhelmingly to the very end of our era, in particular the 20 years from around 1820-40. The content - when compared to the handful of 18th century manuscripts, or to various literary references to the repertoire of country fiddlers - suggests a music heavily defined by time and social group, and we should not automatically equate the music of these respectable artisans with the music of the archetypal fiddler of period art and literature, an altogether much more disreputable character.

Nethertheless, by careful use of all available sources we can still find out a lot about pre-Victorian English fiddling. In particular, I believe it is possible to reconstruct the key elements from which period fiddle styles would have drawn - to stage by stage reconstruct the main components of an archetype. Part 1: the reconstruction (Stage 1: instrument and stance) For most of this period the violin itself was a different instrument to the one we know, it's neck shorter and angled differently, the bass bar lighter, the soundpost thinner, the bridge flatter. It used gut strings and lacked the chin rest. The bow was shorter and without the modern curve. Nor was it held in the modern stance. It was held against the chest or shoulder or under either side of the chin, typically sloping downwards, and gripping with the left hand not the chin. The bow was held with a variety of different grips, and it seems some fiddlers moved the violin as well as the bow, a technique probably inherited from the medieval fiddle.

(Stage 2: tuning)

Also inherited from the medieval fiddle were several alternative tunings, in particular ADAE, AEAE, and AEAC#. Though these so-called 'cross-tunings' seriously restrict the choice of key they have definite advantages for the dance player: increasing volume, making fingering easier, and adding harmonic colour. It is hard to estimate how common and widespread the practice of cross-tuning was. The old collections are probably not a good guide, because cross-tuning presents problems of notation, fiddlers might see no need to specifically refer to it, and the books reflect the most progressive fiddling of the time.

Such archaic pre-violin techniques would probably be most common where they would be least documented - lower down the social scale, in remoter districts, and further back in time (they are certainly commoner in the older collections). I would tentatively suggest a similar situation to 20th century Appalachia, with the most old-fashioned players habitually using cross-tunings, many fiddlers using them occassionally, and the most progressive fiddlers hardly using them at all.

(Stage 3:bowing - tone)

Modern classical bowing is heavily concerned with tone and precision, seeking a rather syrupy tone and a clean overall sound. The dance fiddler has different priorities - rhythm, energy, and volume - and 20th century fiddlers in England and elsewhere tended to use a fairly heavy and dynamic attack, producing a hard, thin tone and often a rather 'dirty' sound. This was presumably as true of the 18th century as the 20th. Where the old fiddlers gave any conscious attention to tone they would presumably have followed the tonal model of baroque Art Violin, the human voice.

(Stage 4: bowing patterns)

Bowings are often marked in the old tune books, and it seems that in their pursuit of rhythm our fiddlers used a number of distinct bowing patterns, in particular the one American fiddlers call the 'Nashville Shuffle', which seems to have been as basic to old-time English fiddling as it still is to old-time Anglo-American fiddling. Imagine a common-time tune divided throughout into groups of four quavers: the first two notes in each group are played on one bow stroke, the next two on separate strokes, giving a flowing but driving feel with an accent on the offbeat. Also very common was the repeated two-note slur - in a group of four quavers one and two are slurred together, then three and four, and so on. This pattern was used for 3/2 as well as common-time tunes. An important variant of this was used in playing dotted or 'Newcastle style' hornpipes. In our basic group of four notes the first is played on a separate bow, then 2 and 3 slurred, then 4 and 5, and so on. Several 20th century English fiddlers were recorded using this pattern on undotted tunes as well, where the short pause between the slurs gives a choppy, lightly syncopated feel. Although the books rarely give instructions as to bow direction, when playing dotted hornpipes it seems the natural down-up pattern would sometimes be reversed. Bowings are less commonly notated with jigs, which is probably significant. The commonest figure we find is a 6/8 variant of the Nashville shuffle. In the basic group of three quavers, the first two or last two are slurrred. Sometimes we find passages where all three quavers are slurrred, giving a rather sensuous feel, and toning down the characteristic bounciness of 6/8. Ocassionally whole tunes make heavy use of these devices but in general they seem to occur in short passages and to have been more a form of passing decoration. My impression is that jigs were largely played with one bow stroke per note, and it may be significant that in the US such bowing is actually called 'jig bow', whatever the time signature.

(Stage 5: bowing - chordal decoration)

The playing of drones and double-stops was fundamental to the medieval fiddle - some even included sympathetic drone strings. Many 20th century English fiddlers also played with a continual drone or used heavy double-stopping. So this was almost certainly an important feature of the centuries in between, and Georgian writers do sometimes refer to the bagpipe-like droning of country fiddlers. We cannot, however, presume all fiddlers always played this way. The Italian single-string sonata style must have come over with the violin and been adopted by some players. Some of the more complex music in the old books - variation sets, competition hornpipes, tunes in flat keys - would not only be difficult to play with heavy double-stopping, they would lose all clarity. In 20th century England heavy droners rarely played the more complex hornpipes, while single-string players showed an equally strong prediliction for them, paralleling the American distinction between drone-inclined 'breakdown' fiddlers and single-string inclined 'hornpipe' fiddlers. I would suggest that the use of drones or heavy double-stopping was very common, but that the more progressive or technically advanced players probably tended towards single-string playing.

(Stage 6: fingering - melodic decoration)

20th century English fiddlers were very sparing in their use of gracenotes, but it's clear that some pre-Victorian players made extensive use of a wide variety of gracings, including long semi-quaver runs between melody notes, the movement the Scots call the birl - the same note bowed rapidly several times - and a series of decorations that were shared with period art music, in particular the mordant (made by rapidly playing the main note and an adjacent note before the melody note), the turn (the same figure as the modern Irish roll, played by hitting first the note above then the note below the melody note), and the shake or trill (the repeated beating of the note above, or sometimes below, the melody note). Vibrato in this period was regarded as a variant of the trill and only used as an occasional decoration. In general these gracings seem to have been performed fairly fast, but contemporary accounts make plain they could be given a variety of speeds. Played slowly or between notes they start to become indistinguishable from the long semi-quaver runs that also figure prominently in the old books. Different gracings were also spliced together to extremely elaborate effect - the shake was often resolved in a turn for example. Traditional musicians tend to use gracings spontaneously and inconsistently and we can't expect the old tune books to show the true levels of decoration. The reality was probably a range from very plain to very elaborate, encompassing a variety of regional styles themselves subject to family and individual preferences. But there is no doubt that some old English fiddlers used extremely elaborate decoration because there still exist some early mechanical organs that were programmed to imitate them - as close to a time machine and a tape recorder as we can get. Some of these use complex gracings and long semi-quaver runs almost to the point of a-rhythmic clutter, a style of playing which has survived into the 21st century in the hands of the Clough school of small-piping.

(Stage 7: melodic variation)

Runs could be seen as a form of melodic variation rather than gracenoting, and were very much a feature of the long variation set - elaborate multi-part variations on a melody or its chord structure, typically containing around 6 to 12 strains, though 20 or more were not unknown. Such variations were often called 'divisions' because one basic technique was to divide up the notes of the melody. 'Division' playing was widespread up to around the mid-18th century, but sets still occur in 19th century fiddle manuscripts and the form has survived amongst small-pipers into the present day. Our concern is with style not repertoire. What brings the variation set within our remit is the importance of improvisation. Sets were often standardized, written down, and memorized, but at the heart of the form lay spontaneous improvization. In the 17th century several 'idiot's guides' were published to help the less talented fake this.These describe a phenomenon very like jazz. We learn that several fiddlers might improvise together Dixieland style, or take breaks in Swing style, and one book describes the practice of calling out 'breve' very like the jazz practice of calling 'fours' where each player takes four bars in turn. Sometimes variations were improvised over an extempore bass line without reference to a specific melody - add to this the note-dividing technique and it must have sounded remarkably like Bebop. Even the language used has uncanny echoes - when the 17th century composer Mathew Locke refers to " the tearing of a consort into pieces with divisions, an old custom of our country fiddlers" I can almost hear Bob Wills shouting "tear it up boys!"

(stage 8: some extremes of variation)

Beyond the melodic variation set lies the playing of descriptive variation sets like the Fox Chase. These are often referred to but rarely notated, presumably because of their dependance on improvisation, trick effects, and a cavalier attitude to conventional structure and rhythm. The latter could be brought to bear on simpler pieces too. 20th century English fiddlers were very given to adding colour to ordinary dance tunes with both light syncopation and what the Americans call 'crooked' playing - deliberately interfering with conventional structure by cutting or adding notes, bars or longer passages. Both syncopation and crookedness are reasonably common in the old manuscripts, and given that they were compiled by the most formally educated and hence probably the most rigid players, these techniques may have been even more common than the books suggest - particularly syncopation, which is both very amenable to spontaneous introduction and notoriously difficult to notate.

(stage 9: the stave)

Staff notation is designed to give explicit instructions in the areas of key, mode, meter, tempo, and accenting. Many of the stylistic subtleties that distinguish the playing of one individual or region from another lie in these areas and here the stave is rather a crude tool. It tells us that English fiddlers played mostly in G, D, A and the easy minors and in first position, but that during the 18th century the use of harder keys like C, F, Bb, Eb, Cm and Gm and experimentation with second and third position became increasingly common, paralleling developments in Art music. They played in a variety of meters including 6/4, 9/4, 3/2, 3/8, 12/8 and 2/2 as well as the more familiar 6/8, 9/8, 4/4, 2/4 and 3/4, and were not as averse to changing the meter of a specific melody as seems the case today. A wide range of tempos were available for any tune, depending on context, dance, audience and mood, and if the stave doesn't reveal the subtle differences in accenting and phrasing that are an important feature of personal and regional style it does reveal some not so subtle ones like the occasional use of the 'Scotch snap' (which in written music appears in England almost 100 years before Scotland) and the evolution in parts of the north of the dotted or 'clog' hornpipe in a process analogous to the evolution of the strathspey-reel in Scotland.

(stage 10: the archetype)

Having outlined some common and widespread features of pre-Victorian English fiddle style, perhaps we can put them together to describe an archetype - a kind of composite English fiddler of around 200 years ago. He held his fiddle against the chest or shoulder or under either side of the chin, sloping downwards and gripping with the left hand, and using various bow grips. He used both standard tuning and crosstunings. He played mainly in G,D, and A and in first position, but sometimes utilized both the harder keys and shifting. He used a variety of bowing patterns, especially the 'Nashville Shuffle' and both the repeated two-note slur and its 'Newcastle' variant. He made plentiful use of drones and double-stops but was familiar with the single-string sonata style, especially for the more complex pieces. He decorated with an impressive mix of gracings including the birl, the mordant, the turn, the shake, and long semi-quaver runs that at times almost broke up the rhythm. Sometimes he liked to play a bit crooked or to throw in a little syncopation. He played with a greater variety of time signatures than today and with probably a more flexible attitude to meter, tempo, rhythm and accenting. He also delighted in the playing of divisions or long variation sets: if good enough he would spontaneously improvise his variations, sometimes in a jazz-like small group context. And if he was of a progressive bent he maintained a certain interest in Art music and its techniques. In modern terms this sounds remarkably like a hybrid of older-style Appalachian, western Irish, and Scots fiddle styles. Which, it seems to me, has important implications which transcend the parochial concerns of English musical antiquarianism.

Part 2....and beyond

Not only are many of the stylistic features outlined in this talk still to be found in related traditions, to some extent they now demarcate the boundaries between them - turns, mordants and trills are now seen as peculiarly Irish; the birl, the snap, and respect for classical aesthetic and technique as distinctively Scots; the Nashville shuffle bow, cross-tuning, the heavy use of drones and double-stopping, and a fondness for crookedness, syncopation, and improvisation, as archetypally American. Given that all these features can be found in pre-Victorian English fiddling it begins to appear almost as a 'missing link'. In the 18th century England was one of the most densely populated countries in the world, around 80% of Britons were English, white Americans were probably around 70% English in origin, and an Irish population less than half that of England included a substantial minority of fairly recent English origin.

Thus the English probably did have a greater variety of fiddle techniques and styles than their neighbours, and must have played a central and influential role in this music that is hard to imagine nowadays. But demographics are not the whole story. On closer examination it seems that a certain standardization and simplification has been taking place in Irish, Scots and American fiddling over the last 200 years and that many of the stylistic features outlined here in an old English context were once fairly widespread and general - often well beyond the Anglo-Celtic world. The stance was common to Europe during the baroque era, among Art violinists as well as fiddlers, and aspects of it still survive in many areas of Europe and America. The main crosstunings were also known to Scots and European fiddlers, and were even used in 18th century European art music. The 'Nashville' bowing pattern also occurs in old Scots collections. Droning and double-stopping are widespread features of vernacular bowed instrument technique throughout the world, and were formerly a lot more common in both Irish and Scots music than today. Turns, mordants, trills and other gracings were shared with both baroque Art violin and several European vernacular traditions. The birl was common to the entire British Isles and was formerly common in America. Decorative runs are also common in old Scots collections and are still used by some Irish players. The 'Scotch snap' was found throughout Europe, sometimes with similar regional associations (in France and Italy it was the 'Lombard snap') which are probably just metaphors for archaism and rusticity. Respect for classical aesthetic and technique has always existed among the more progressive fiddlers everywhere. Long variation sets are found in old Irish and Scots collections too, and syncopation, crookedness and improvisation are found throughout the world. It would seem, in fact, that in the past there was a greater degree of commonality than is the case today, particularly within the British Isles and their American colonies, to some extent within Europe generally, and even between Art violin and vernacular fiddling.

We can draw lines between human beings anywhere we choose. Whether National frontiers are always the most meaningful places to do so in the study of popular culture is questionable. Indeed, the widespread equation of traditional music and national identity seems to me positively misleading. It has not only tended to obscure the kind of supra-national commonality discussed above, it has tended to play down the crucial importance of the sub-national - of regional, local, family, individual, class, and generational differences - and to ignore the reality of distinctive cultural regions that straddle the frontiers, like the Anglo-Scots border country and the 'Bristol channel zone' of south-west England, south Wales, and south-east Ireland. This is not to deny the reality of a national dimension, but 'National' fiddle style should perhaps be seen as the sum total of all the varied styles found within a given political border rather than as something monolithic, homogenous, static, self-contained and totally unique. This paper has examined in detail the main stylistic features of English fiddling in the pre-Victorian era, and has given, I hope, some idea of its richness and variety. It may seem perverse to turn round at the end and emphasise the areas of wider commonality, but it is time scholars started biting this particular bullet. Political and other boundaries have never stopped fiddlers from extending and developing their music, and unless we follow their example our understanding of their music will always be partial and stunted.

(Paul Roberts, Hebden Bridge, July 2001