Tutti Reedi

<< Double Reeds
Bagpipes >>

The Forked Shawm – An Ingenious Invention

This was originally a paper, given at a conference on Near-Eastern music at the Maison Français in Oxford in 1996 and subsequently published in the Journal of the International Council for Traditional Music the following year.

The forked shawm is a type of instrument which is in common use around the southern and eastern shores of the Mediterranean, from Morocco (Cherki 1981:38),1 across to Egypt (Collaer & Elsner 1983:48), in Turkey (Picken 1975:485ff), in parts of the Balkans, and certainly as far east as Azerbaijan (Vertkov et al. 1975:fig.415), probably Georgia (ibid.: fig.456), and perhaps Uzbekistan (ibid.:fig.582), What is important is that this particular variety of shawm, with the simple and ingenious device of the fork, is the most important development in the history of the shawm in two and half thousand years.

It is necessary, to begin with, to define some terms and to explain some characteristics of the shawm and other reed instruments. A shawm is normally understood to be an instrument of fairly wide conical bore. Its internal profile expands fairly rapidly from the player’s end, the proximal end, where it is narrow, to an open bell at the distal end. It is played with a double reed, made either of two pieces of cane or from a flattened plant stem. The difference seems to be a matter of cultural preference, rather than depending upon botanical distribution. North of the Mediterranean, from the Atlantic to the Adriatic, shawm reeds are made from two blades of a cane-like reed, most commonly Arundo donax. South and east of the Mediterranean and the Adriatic, the reed is normally a flattened plant stem. Picken (1975:358) gives full details of the Turkish variety; other varieties, from Morocco to China, have not, so far as I know, received the same scientific attention. Irrespective of its material, the reed, instead of being gripped between the lips, is normally held within the mouth cavity, where it can vibrate very freely, which gives rise to the very loud sound immediately recognisable as characteristic of the shawm. Thus the shawm is quite unlike our orchestral oboe both in bore and in playing technique, for the oboe’s bore is fairly narrow, and its double reed, again of two blades of Arundo donax, is longer and narrower than that of the shawm, and is gripped firmly by the player’s lips. Examples of the shawm on the northern shores of the Mediterranean are the Basque and Navarrese gaita, the Valencia dulzaina, the Catalan tiple and tenora and the gralla, the Italian ciaremella, and. further north, the Breton bombarde and the chanters of many bagpipes, from the Spanish gaita gallega to the Scots piob mhór.

We also speak of the cylindrical-bore shawm, an instrument whose bore does not expand and which is also played with a double reed. Some examples are the Japanese hichiriki, the Chinese guan, the Persian balaban, and the Turkish mey. All of these, and their cognates in other countries, can be shown to be connected with the ancient Greek aulos, and to be a type of instrument which travelled down the Silk Route between Europe and China. In which direction they did so, and which are ancestral to others, and indeed the connexions themselves, are not our concern at this moment and in this context. Unlike the type of shawm we have already described, their bore is a straight cylinder rather than a cone, but like the reeds of the conical shawms also east of the Mediterranean, the reed is again a double reed, again of a flattened plant stem, but one very much larger and very much wider in diameter.2

The shape of the instrument’s bore, whether it be a cylinder or a cone, is important and controls the acoustical behaviour of the instrument. A reed-driven conical bore, an instrument which is played with a reed and which has an expanding bore, such as our oboe and our saxophone, as well as the conical shawm, overblows octaves. A reed-driven cylindrical bore, on the other hand, an instrument played with a reed and whose bore does not expand, such as our orchestral clarinet, and the cylindrical shawms, not only overblows twelfths instead of octaves, but produces a lowest note which is considerably lower in pitch than that of a conical instrument of the same length. This is why, of the soprano saxophone and the clarinet, both of which are built in the key of B flat, and both of which are much the same length, the conical soprano saxophone only goes down to the A below middle C as its lowest note (its sounding pitch, not its written pitch), whereas the cylindrical clarinet goes down to the D a fifth lower (again its sounding pitch), even though both are about the same length. Because the bore of the saxophone is conical, it therefore overblows to the upper octave; because the bore of the clarinet is cylindrical, it therefore overblows to a twelfth. By ‘overblows’ one means opening a vent, usually a thumb-covered hole, in the upper, more proximal part of the bore, covering all other fingerholes, and blowing rather harder so that the air column jumps to a higher mode of vibration.3

The same applies to shawms, so that those of conical bore such as the gaita and dulzaina (to cite only two examples) overblow to the octave, whereas those of cylindrical bore, such as the hichiriki, mey, and so on, overblow to the twelfth, if at all.4

It is easy to make a cylindrical bore, whereas it is difficult to make a conical one. For a cylindrical bore, one can use a length of bamboo, which has a natural cylindrical bore, as with the hichiriki, or one can simply burn a hole with a hot iron in a piece of wood, or use a drill to make the hole. To make a conical bore, one needs a tool called a reamer. This is a sharp-edged, conical, metal tool which is shaped, with files, by forging, or more easily on a metal-working lathe, to the precise profile that it is intended to produce inside the instrument. To use it, one must first drill a cylindrical pilot bore through the piece of wood, and then, with this tool, ream that out to a cone. If one has a reamer, this is fairly easily done on a lathe, though it takes time and considerable care. Without a lathe the task is considerably more difficult because of the problems of keeping everything straight and properly aligned. Without a reamer, or without the skill to make one, then one must use a tool with a straight, non-tapering blade. This must be held against the side of the bore, gradually pushing it further and further in, gently enough that the wood will not be cracked but firmly enough that the blade does not chatter in the bore, keeping the angle correct all the way, in order to turn a cylinder into a cone. The whole process is difficult, especially if one wants a fairly wide cone, and really is only easily done in shaping the flare of the bell. We can see widely flared bells on some early Turkish instruments such as those portrayed by Carpaccio in Venice around 1500,5 and also on one from Turkey in the Laurence Picken Collection which is now in the Museum of Archaeology and Ethnography at Cambridge University.6 The operation is particularly tricky in its final stages when one has to reach right in to the top of the tube with the blade, for it is very easy to distort the lower part of the bore as one does so. Some conical shawms, including that just referred to in the Picken Collection, and several Nigerian and Mexican shawms in my own collection, appear to have been carved by hand, rather than on a lathe, which is even more difficult to do. However, once all this trouble has been taken, and all this work has been done, one has an instrument which is far more efficient than one of cylindrical bore. With the conical bore overblowing an octave, we have an instrument which can, with only six fingerholes, cover a full range. Say, for simplicity, that we are starting on C, with all the fingerholes covered. If, for simplicity of explanation again, we take the instrument to be tuned to our diatonic scale, opening six holes in sequence would take us up D, E, F, G, A, to B; opening the thumbhole and covering all the other holes again and overblowing would produce the upper C, and then one can go on up, through the next octave. With a cylindrical bore, the six holes would take us to B again (if we started on C), but covering them all and overblowing would produce the high G a twelfth above the C, not the C an octave above. Thus there would be a gap in the compass, between the B and the high G, which could only be filled by adding more fingerholes. Another four fingerholes would be needed, for the upper C, D, E, and F. Using ten fingerholes is difficult when we only have ten fingers and thumbs to do it with and we have to hold the instrument up at the same time!

This is why, in antiquity, the development of the conical-bore shawm, which seems to have happened among the Faliscans,7 somewhere around 500 BC (Becker 1966:Abb.3), was such an advance. This, too, is why we find the conical-bore shawm over so wide a geographical area today. As we have already suggested, it is used from Northern Europe, where it is often blown through a bag as the chanter of a bagpipe, through all of southern Europe, into north and west Africa, across the Middle and Near East, through India, Pakistan, and Tibet, and much of the southern parts of what was the USSR, to China, Korea, and Indonesia, and, presumably as an ex-colonial import from Spain, into Central America.

Forked Shawm figure 1

Figure 1

Now let us turn to the subject of this article, the forked shawm (fig. 1). This instrument appears to have a cylindrical bore, if one discounts the terminal flare of the bell (the clarinet has this flared bell, as well, without affecting its cylindricity) and thus it seems to be more primitive than the conical shawms. And yet what historical evidence we have strongly suggests that it is a fairly recent development. Laurence Picken’s evidence (1975:499) is quite clear: that the forked shawm must be later in use in Turkey than the widely-flared, conical-bore, hand-carved instrument already mentioned, and which must originally date to 1800 or so, perhaps as early as 1700. Any of us who have heard shawms played from Morocco to Turkey and into the Balkans, know that it overblows octaves, not twelfths. Thus we know that it works as though its bore were conical, despite its cylindrical appearance.

Forked Shawm figure 1

Figure 2

How is this achieved? The answer is the fork in the head. This is a delightfully simple and yet a highly ingenious way of converting an easily made cylinder into a cone. For a cone to be effectively conical, to be acoustically conical, it does not need to be what a geometrician would call a cone. A simple stepped cone made by putting three bits of different diameter bamboo together, stepping one into the next, works perfectly well, as Nazir Jairazbhoy’s recordings and report from the Dumbu or Paidi have shown us (Jairazbhoy 1988:32). Experiment with sections of aluminium tubing of different bore diameters will confirm it and will, perhaps, demonstrate the acoustical principle more securely in that being factory made, they are quite certainly cylindrical. And this what we have with the fork. The reed goes on to a metal staple, a short tube which itself is usually, but not always a short cone; it is cylindrical in some examples. That staple is fixed into the top of the wooden fork (fig. 2) which forms the cap of the instrument and which has a short cylindrical bore drilled through it, slightly wider in bore than the staple. At the bottom of this bore through the cap, the wood is carved into a fork. The gap between the tines of the fork is slightly wider than the width of the previous bore, but the effective bore is wider still because the sides of the fork, at 90° to the tines, are open to the full bore of the instrument into which the fork is inserted, so the bore is both wider and in shape something like an oblong with rounded ends. If, when looking down the bore, one were to take the fingerholes as the north side and the thumbhole as the south, the east and west sides would be formed by the two tines of the fork. The north side opens to the full width of the bore first, to allow access to the uppermost fingerhole, thus forming the third step (staple, bore through the head, access to the north side of the bore). The fourth step is to the south side of the bore, at the thumbhole. Thereafter patterns differ: some forks have tapering tines on the east and west, leading fairly smoothly into the bore, as in fig. 2. Others remain the same thickness all the way, and step only to the full width of the bore at the end of the tines, this forming the fifth and last step, where the bore becomes circular again. Even this can be complex in geometry because the distal ends of the tines are usually cut to a point on their north and south sides. Thus a mathematical program for this device would be of very high complexity but the practical result is that we have a short, stepped cone set into the top of our cylindrical tube.

Now where did this simple but ingenious device originate? We do not know. There is, as yet, very little work done in historical ethno-organology. All that we do know is that the instrument is used from Turkey northwards into Macedonia, but not much further, south and west from Egypt throughout North Africa to Morocco, and east into Central Asia at least as far as the Caucasus.8 The easy assumption is that it was a Turkish invention which spread throughout the Ottoman Empire, but it could just as easily have been a Moroccan invention which trickled along the southern littoral of the Mediterranean until it met the Ottomans in Egypt, whence it travelled to Turkey and so into what used to be Yugoslavia. Or devised in Central Asia, where so many of our instruments seem to have begun, the lute, the gong, the long trumpet, the fiddle bow, and thence travelled again to the centre of the Ottoman Empire. Or, of course, anywhere in between. It has been suggested9 that Turkey, as the heart of the Empire, was the technological centre and thus the obvious point of origin. But this is not necessarily so; the best swords, for example, were made in Syria, and other areas also had their specialities. It is, besides, an invention, and the spark of invention can strike anywhere.

As for when it was invented, all that we can say is that it must surely be later than Carpaccio’s time (the early sixteenth century) and equally it must be earlier than a hundred years ago. My own guess would be fairly early in the nineteenth century, but it is only a guess.

So here we have an ingenious device which was quite easy to make. In England gypsies used to make clothes pegs in large numbers with a bodger’s lathe, almost exactly like these forks, which they sold from door to door to housewives to hang up their washing. Thus it is clear that anybody with quite simple tools could make such a device. Once this device was created, it became far easier to make the body of the shawm, especially to drill its bore, for this could now be a straight cylinder with a conical bell. The bell is easily turned just with a knife, for it is easy to reach. So there is no doubt at all that the forked shawm was a progression and not a retrogression, far easier to make and equally efficient.

It is not usual, in ethnomusicology, to discuss technological developments in the ways in which we do in European10 systematic organology, but the development of the fork was a progression which, in its way and within its culture, was just as important to the history of instruments, and to musical life and use, as such developments within our pop and orchestral culture as the electric guitar, the Boehm system flute or the invention of the valve for brass instruments.

© Jeremy Montagu, 1997


Baines, Anthony, 1992, The Oxford Companion to Musical Instruments, Oxford: Oxford University Press.

Becker, Heinz, 1966, Zur Entwicklungsgeschichte der antiken und mittelalterlichen Rohrblattinstrumente, Hamburg: Hans Sikorski.

Cherki, Salah, 1981, Musique Marocaine, Mohammedia: Imprimerie de Fedala.

Collaer, Paul and Jürgen Elsner, 1983, Nordafrika: Musikgeschichte in Bildern, Bd.1, Lief.8, Leipzig: Deutsche Verlag für Musik.

Jairazbhoy, Nazir Ali, 1988, A Musical Journey Through India 1963-1964, Los Angeles, Department of Ethnomusicology, University of California, Los Angeles.

Picken, Laurence , 1975, Folk Musical Instruments of Turkey, Oxford: Oxford University Press.

Sadie, Stanley, 1984, The New Grove Dictionary of Musical Instruments, London: Macmillan.

Vertkov, K., G. Blagodatov, E. Yazovitskaya, 1975, Atlas of Musical Instruments of the Peoples Inhabiting the USSR, Moscow: State Music Publishers.


  1. In addition to the sources cited here and elsewhere, most of the instruments mentioned are described, and some are illustrated, in Sadie 1984 and Baines 1992. Most are represented in my own collection.

  2. Picken (1975:476) suggests from the same plant, despite the size difference, as the very much smaller reed for the zurna.

  3. Somewhat a simplification, but probably sufficient in this connexion; it can also help to jog the end of the reed with the tongue.

  4. It is often much more difficult to persuade cylindrical-bore reed instruments to overblow than those of conical bore.

  5. E.g. The Turkish Ambassadors in Venice.

  6. CUMAE 77/55. Illustrated in Picken 1975:plate 41n.

  7. The Faliscans were a sub-tribe of the Etruscans,

  8. See note 1 and the references in the earlier part of this article.

  9. By Dr Eckhard Neubauer when a shorter version of this article was presented as a paper at the ICTM Conference on Arabic and Beduin Music at the Maison Française in Oxford in 1996. I would here express my thanks to Professor Dieter Christensen for his encouragement to convert that paper into this article.

  10. By which I mean the pop- or symphony-orchestral-music culture in which we, in Europe, America, Japan, and much of the world, also live.

<< Double Reeds
Bagpipes >>