Like language, music itself leaves no traces and can only be studied indirectly through the material remains of musical behaviours, but archaeological finds of musical instruments prove beyond doubt that Homo sapiens was making music towards the end of the Paleolithic.
We shouldn’t conclude though that music is no older than these surviving instruments. Modern hunter-gatherers often make instruments out of natural materials like skins, wood, hollowed plant stems and gourds, which are perishable, and we have ample evidence that Paleolithic humans had the tool-making skills necessary to make comparable objects. We know from wooden spears found at Schöningen in 1995 that early humans were skilfully working wood, for example, at least 400,000 years ago. Also, we don’t need accessories to make music: the most versatile musical tool of all (and probably the oldest) is the human voice, which can create an immense range of sounds. This could have been accompanied by handclaps or other forms of percussion, such as hitting objects together. Music in this form could in theory predate tool-making. So if the only evidence we consider is the surviving instruments, we will have only the most limited conception of Paleolithic music-making.
Even when instruments still exist, reconstructing the development of music in the Paleolithic is very difficult because only the invention of time travel could reveal to us what music was created with them. Ways of writing music down appeared relatively late — our earliest example of notation is a tablet from Nippur in Mesopotamia dating to around 2000 BCE. Even today, a huge amount of the music created is never written down.
Musical instruments in the Paleolithic
We don’t know when in the Paleolithic humans began making instruments. The simplest percussion instruments — such as rattles, shakers and drums — must predate sophisticated objects like flutes, possibly by many millennia, and they predominate over melodic instruments amongst modern hunter-gatherers. One unusual discovery was a set of mammoth bones found in Mezin in the Ukraine. These were ornamented with red paint and located with a hammer and other items, which invite interpretation as percussion instruments — even as a kind of Ice Age ‘orchestra’. These date to around 20,000 years ago, and it remains to be seen whether we will unearth anything more ancient.
Archaeologists have found many Paleolithic objects which are or may be musical instruments. All the human-made items are made of bone or ivory, because these preserve much better than materials such as wood and skins. There is still considerable debate about whether some of these objects should be identified as instruments.
The archaeologist Iain Morley has concluded that the objects fall into five main types: flutes and pipes, pierced phalanges, bullroarers, rasps, and caves (as exploited for their acoustic effects).
The object regarded as possibly the oldest surviving musical instrument is a perforated bear femur found at the Divje Babe site in modern Slovenia, thought to date to 43,000 years ago. This object was found in a Mousterian level, associated with Neanderthals. If it was made by Neanderthals, it would predate the Human Revolution often associated with the flowering of symbolic behaviours and art. Archaeologists are still disputing whether this object is part of an instrument or just a bit of bone with two holes in it. A detailed study by Francesco d’Errico and colleagues in 2003 was sceptical, concluding that the perforations were made by a carnivore . Morley considers some rival interpretations and also concludes, albeit with reservations, that this object probably wasn’t a musical instrument. There is little other evidence that Neanderthals made music, so for the moment it is better to err on the side of caution. One of the suggested reasons why our species survived while Neanderthals became extinct is the possession by Homo sapiens of unique cognitive and social powers, which could have made a huge qualitative difference to our ability to create music.
Unambiguous flutes and pipes made by our own species do exist, and are the most numerous sort of instrument, with over 120 claimed examples having been found. (A pipe is blown into directly, whereas a flute is played by blowing across a hole, but not every writer makes this distinction.) More than thirty unequivocal examples survive, all from the European Upper Paleolithic, spanning the Aurignacian, Gravettian and Magdalenian periods. These are often made from bird bones — which are naturally hollow — and have a variable number of holes. The oldest examples were found in caves in Germany and date to around 35,000 years ago. The most complete was from Hohle Fels in Swabia: a flute with five finger holes made from a vulture bone. Two other examples from nearby caves at Geissenklösterle were made of swan bone and mammoth ivory.
Paleolithic bone flute, ca. 35,000 years old, from Hohle Fels.
The latter in particular would have required considerable skill, as mammoth ivory is a dense material and difficult to carve. This implies what the flutes’ discoverers call a “well-established musical tradition”  that must date further back than these particular specimens. These finds represent the first known unambiguous evidence of music-making, as all other extant Paleolithic instruments post-date 30,000 years ago. The prehistorian Friedrich Seeburger has reconstructed what these prehistoric flutes sounded like using replicas, which you can listen to at the nature.com website. These pipes were not mere crude ways of making a noise, but in fact are not far removed from modern flutes and capable of expressive melody.
Another important set of pipes was found at Isturitz in south-west France, where more than twenty items have been found spanning a 15,000-year period from the Aurignacian to the Magdalenian. The status of some of these is questionable, but the finest examples have drilled holes and incised decorations and were clearly intended for making music.
The next set of instruments are pierced phalanges or phalangeal whistles. These are made of the phalange bones — in humans, these are the bones of the fingers and toes — of animals such as reindeer and ibex, pierced with a single hole. Blowing across this hole produces a high-pitched sound.
Phalangeal whistles from France. The two left-hand items are from the Aurignacian site of Tarté; the right-hand item is from Le Moustier. Photo: Duncan Caldwell.
The ease with which such perforations can appear sometimes makes it hard to tell exactly if an object was manufactured. But there is no doubt that some were deliberately made by Paleolithic humans. Whether the whistles were in fact used for music, rather than as noise-makers for other purposes, such as signalling, remains hard to say.
Less common is the third category, the bullroarers — these are flat pieces of bone or wood that make a buzzing or humming sound when spun in the air. (You can hear a sound sample here.) These have been used by many cultures, including modern hunter-gatherers such as the Australian Aborigines. A bullroarer can be made very easily by drilling a hole in a piece of bone, flint or wood and spinning it on a cord, so many perforated Paleolithic objects are open to interpretation as bullroarers. Many have been described as pendants or fishing weights, or as not of human provenance at all. But some of the evidence is stronger, such as the oval specimen carved from reindeer antler found in the cave of La Roche de Birol in the Dordogne, which is incised with a geometric pattern and pierced at one end. It is not certain that this and similar objects from the Paleolithic are bullroarers, but in tests they perform very well as such.
The fourth type of instrument is the rasp, a scraped form of idiophone or vibrating instrument. This is a piece of wood, stone or bone carved with grooves, against which the musician scrapes another object. Likely candidates include a Mousterian mammoth bone discovered in Belgium and dated to 50–40,000 years ago, which has regular striations down one side; a Magdalenian item made of reindeer antler from Pekárna cave in the Czech Republic; and further Magdalenian examples from the French sites of Mas d’Azil and Bruniquel. These objects perform as rasps, but, as with the bullroarers, we still cannot be sure that their purpose was musical.
We have referred before to the bas-relief image from Laussel known as the ‘Woman with a Horn’. Morley points out that the crescent-shaped object in the woman’s right hand resembles an incised bison horn similar to rasps from Mexico.
The ‘Woman with a Horn’ from Laussel, ca. 25,000–20,000 years ago.
The final type of instrument is a telling example of how Paleolithic peoples could have used their environment to make sounds in a variety of creative ways. Several prehistorians have observed correlations between cave acoustics and parietal art. Iegor Reznikoff and Michel Dauvois studied three caves in the French Pyrenees, making sound tests to draw up a ‘resonance’ map of the caves, and found that most cave paintings are situated one metre from points where the caves resonate to certain notes. They concluded that the paintings’ locations were chosen according to the acoustics of the caves, even when the point of resonance was poorly accessible and did not leave the artists enough room for a full image.
It follows that there may have been rituals or other social events that involved both the paintings and music — above all the human voice, to which the caves’ resonances responded more than to whistles and other instruments. As the archaeologist Chris Scarre noted, “the image of the cave artists chanting incantations in front of their paintings may not be too fanciful.”
Researchers have also noted that stalactites, stalagmites, and other points can produce a tone when struck, thus acting as lithophones.
The caves at Nerja in Andalucia in Spain. Image: Luzzyacentillo
One example of this is the cave of Nerja, a natural amphitheatre in which concerts are performed today. The Belgian archaeologist Lya Dams observed that the so-called ‘organ’, a section of the cave compromising a series of fluted calcareous folds, can produce clear notes when struck. These had been painted with both abstract images and animals, and bear signs of having been hit with hard objects. It is hard to believe that the cave’s Paleolithic occupants would not have noticed this property, so Dams concludes that the ‘organ’ was being used as a lithophone, and that the parietal art was closely connected with this musical purpose.
“Whereas we now visit painted caves in a hushed reverence,” Steven Mithen commented, “they probably once reverberated with the sounds of pipes, stalagmite xylophones, singing and dancing.”
The instruments we’ve considered are all from the Eurasian Upper Paleolithic. This is not down to Eurocentric bias but to the reality of the archaeological record. Instruments from other parts of the world, e.g. the bone flutes found at the Jiahu site near Henan in China, currently date only from the Neolithic. This does not necessarily mean that musical instruments were unique to Europe during the Paleolithic, only that we haven’t found any elsewhere. This absence could be for a variety of reasons — e.g. the preservative quality of the materials used. There is a disparity, unexplained as yet, that anatomically modern humans seem to have emigrated from Africa around 70,000 years ago, but the earliest unequivocal musical instruments appear not in Africa but in Europe nearly 40,000 years later. This gap in the archaeological record may simply be down to the different intensity of research in Europe as opposed to Africa, and future excavation may yet resolve it.
The purpose of Paleolithic music
It is impossible to say what form music took in the Ice Age. If we look to modern hunter-gatherer societies for parallels, we can see that music is likely to have been highly diverse. Bullroarers, for example, are used in quite different ways across cultures; amongst Australian Aborigines alone, their uses vary according to context and to tribal group: in initiation ceremonies, in burial rituals, to ward off spirits, to imitate the voices of spirits, to warn women away from exclusively male rituals, and so on.
It would be a mistake to make bold assertions about the music-making of Paleolithic humans based on our observations of peoples who are separated from the Stone Age by many thousands of years. Music developed during the Pleistocene, in part when humans were still undergoing speciation. No such environment exists today for us to study. It would also be a mistake to view hunter-gatherer societies as somehow ‘frozen’ in a Paleolithic state, as thousands of years have passed since the Pleistocene and no society undergoes zero change. So all such parallels can do is open our minds to some of the possibilities.
Nonetheless there are grounds for some educated speculation. Iain Morley studied four diverse hunter-gatherer groups — Native American, African pygmy, Australian aborigine and Eskimo — and found strong parallels between their musical-making.
All four groups meet up with fellow communities during their most difficult subsistence season, during which time there is increased performance of ceremonial and communal music and dance. In all four cultures, the ceremonial and social use of music is very important, is communal and is almost always accompanied by rhythmic dancing. In all four cultures music is also performed purely as a communal activity for pleasure.
All four have music which is predominantly vocal, and accompanied mainly by percussion instruments. The use of melodic instruments is minimal and, when used, invariably consists of end-blown pipes. These are usually single-toned instruments. All the instruments, whether percussive or melodic, are made from naturally occurring organic materials.
All four peoples believe themselves to have come from the land, to be akin with the other fauna of their environment, and use music to try to influence the world around them. Music and dance can have important uses in engendering group cohesion, altering mood, as an aid to the teaching of dance, and can facilitate group interactions and communality, within and between groups. In the majority of these instances the music itself has no inherent symbolism, but it can be used to accompany symbolic activities.
A great deal of these societies’ music didn’t require any instruments. Instead, the group vocalised, with accompanying foot-stamping, hand-clapping and so on. This may be significant given the absence of instruments older than 35,000 years. Morley also found that these societies sometimes used music as a mnemonic, helping to preserve tribal lore. It is much easier to remember texts that are highly structured with pattern and repetition.
The findings are important because the groups studied occupy different continents and developed their musical practices independently of each other. Morley concludes that there are “fundamental similarities in the nature and roles of music between these diverse groups” which may indicate an extremely ancient tradition, and proposes that “there are important evolutionary driving forces towards those common behaviours, either as a consequence of subsistence method or of human biology.”
Bone pipes and most other instruments can only be made with an advanced level of tool-making. As technology improved so did the sophistication and range of our instruments, which were often artistic objects in themselves. By the rise of early civilisation around 3000 BCE, we see very finely-made instruments such as the lyres found in the Royal Cemetery of Ur, decorated with beaten gold. This technological process has of course continued into the present, adding electric guitars and samplers and many other things to our musical inventory.
On the current evidence, there appears to have been a relatively sudden proliferation of instruments from about 35,000 years ago. The creation of musical instruments may therefore be the product of the crucial leap in human cognitive and symbolic capacities during the Upper Paleolithic. It is probably no accident that instruments appear in Europe at about the same time as cave painting and sculpture. This is not to say that we can rule out musical activity by Homo sapiens or even pre-sapiens humans prior to 35,000 years ago. The sophisticated instruments that have survived must have been preceded by earlier, less developed forms, perhaps in more perishable materials, not to mention yet earlier periods before musical tools were used at all. As always in palaeoanthropology, the incomplete archaeological record gives us only very partial information. Limited to material traces and what can be inferred from them, we can say little or nothing about what melodies were played, about which people in a particular society created music, if and how music was combined with other art forms, etc.
It can be difficult to ascertain whether a particular item that can be used musically was actually made for that purpose. There is ongoing debate about whether or not Neanderthals were capable of creating music and therefore musical instruments, and whether some of the objects were created by humans at all. There is also no way of knowing whether the materials used for making them were simply the most appropriate or had their own symbolic meaning.
Paleolithic humans may have used music for some or all of the reasons that modern hunter-gatherers do, or for other reasons now lost. What does seem certain is that they, like more recent societies, are bound to have created rich and varied musical cultures for a range of spiritual, social and entertainment purposes.
 See for example Adrian Lister and Paul Bahn, Mammoths: Giants of the Ice Age (2007).
 Iain Morley, The Evolutionary Origins and Archaeology of Music, PhD dissertation (2003).
 D’Errico, Henshilwood, Lawson, Vanhaeren, Tillier, Soressi, Bresson, Maureille, Nowell, Lakarra, Backwell and Julien, ‘Archaeological evidence for the emergence of language, symbolism and music — an alternative multidisciplinary perspective’, Journal of World Prehistory Vol. 17 (2003).
 Nicholas Conard of the University of Tübingen, quoted in John Noble Wilford, ‘Flutes offer clues to Stone Age music’, New York Times (24 June 2009).
 Iegor Reznikoff and Michel Dauvois, ‘La dimension sonore des grottes ornées’, Bulletin de la Société Préhistorique Française Vol. 85 (1988).
 Chris Scarre, ‘Painting by Resonance’, Nature (1989).
 Lya Dams, ‘Preliminary findings at the “Organ” Sanctuary in the cave of Nerja, Malaga, Spain’, Oxford Journal of Archaeology (1984).
 Steven Mithen, The Singing Neanderthals (2005).
 The oldest known wooden instruments, for example, a set of six pipes found at Greystones in Ireland, are relatively recent at 4000 years of age.
 Morley, op. cit.