Did our Late Palaeolithic ancestors use stick shuttles for weaving and netting?
Over a century ago, De Maret (1879) observed similarities between the round reindeer-antler sticks, with two deeply notched ends of Middle Magdalenian (16 700–15 700 BC) date from Le Placard (France), with weaving or netting shuttles (French navettes) still in use in nineteenth-century France (Allain et al. 1985).
Then and now, however, practically all stick shuttles are either needle-like or flat, and therefore quite different from the round Magdalenian navettes (Figure 1). Moreover, the term ‘navette’ implied that Late Palaeolithic hunter-gatherers had already developed the technology of weaving or netting. This latter issue explains why later authors dismissed the suggestion that these reindeer-antler sticks might have been used for weaving or netting, even if they continued to refer to these very distinctive artefacts as navettes (Bahn 2001).
As the netting- or weaving-shuttle hypothesis had been rejected, the navettes had to be interpreted differently; for example, as intermediate pieces (‘foreshafts’) connecting bone or antler points (‘sagaies’) to wooden shafts (e.g. Kozlowski et al. 1995), or as handles for end scrapers (Allain et al. 1985). And yet the end scrapers from Maszycka (Poland) are too thick to fit into the notches of the navettes from that location (Kozlowski et al. 1995), and it is hard to imagine the curved navette from Le Placard (Figure 2) as the foreshaft for an atlatl.
There are, however, several reasons to reconsider whether or not the Magdalenian navettes were indeed stick shuttles. First is the discovery in 1995 that weaving and knotted-netting technologies existed from at least the Early Gravettian (29 000 BC; Adovasio et al. 1996; Adovasio et al. 2007). This “necessarily raised the question whether the Upper Paleolithic people who wove fabrics and made baskets and nets used specialized tools to do so”, although “the absence of such tools in the archaeological record hinted that these might not have been recognized to date” (Soffer 2004: 407–408).
The second reason is the documented—if relatively rare—use of round, stick shuttles from the Egyptian Middle Kingdom (1800–1700 BC) through to the beginning of the twentieth century (Figure 2). Examples from ethnographic collections around the world demonstrate that in shape, cross-section, length and relative thickness, there are striking similarities with the Magdalenian navettes (Figure 2). Replicas of Magdalenian navettes, regardless of their shape (straight, curved), length or thickness, might easily replace the historical, round, stick shuttles for weaving or net making (Figure 3). The comparative rarity of round, stick shuttles around the world may be due to the preference of experienced weavers and net makers for needle-like and flat specimens.
Finally, the Middle Magdalenian female depictions from La Marche (Lussac-les-Châteaux, France) and Isturitz (France) showing possible woven hair-ribbons, necklets, wristlets, anklets and belts (De Saint-Périer 1932; Pales & Tassin de Saint-Péreuse 1976) suggest the importance of woven apparel in Late Palaeolithic “domestic settings or base camps occupied by both hunters and weavers” (Soffer & Adovasio 2010: 242). These ribbons and belts might well have been made with the aid of round, stick shuttles and a batten.
There remains the question of why such a useful tool only occurs at a handful of Middle Magdalenian sites. The answer might be found in rapid but irregular drying-out of the antler sticks, a process enhanced by the deep end notches, causing profound alteration to the structure of these artefacts and making them prone to fracture. The dropping of a dried-out navette on a rocky surface might account for the frequently observed longitudinal and transverse fractures. Middle Magdalenian sites with navettes were apparently restricted to cold, dry steppe environments (Allain et al. 1985; Kozlowski et al. 1995) where suitable trees for the manufacture of wooden netting shuttles may have been scarce or absent. Allain (1957: 222) suggested that the Magdalenian reindeer-antler navette was “a rather rare tool” because it “possibly had been more frequently made of wood”. Reindeer antler was perhaps only a second choice material.
Something similar appears to be true for historical, stick shuttles. A rapid survey of ethnographic museum collections suggests that people in treeless Arctic regions used bone, ivory, baleen and reindeer or caribou antler, whereas wood, cane or bamboo have been preferred in more forested areas.
In fact, wooden implements in general—and the same holds for plant-fibre artefacts such as basketry, cordage, textiles and nets—are several times more common than stone tools in both the ethnographic past and present (even in Arctic and sub-Arctic environments), and this may have been true also in the Late Palaeolithic household (Adovasio 2014). Unfortunately, however, wood is preserved only under the most exceptional circumstances at most Palaeolithic sites.
Identical artefacts need not necessarily share the same function, especially if they are separated by a timespan of more than 10 000 years. And yet only the round-stick shuttle hypothesis can be applied to all Magdalenian navettes; this is in contrast to the alternative interpretations of hafting sleeves or tool handles. The recognition of navettes as stick shuttles may not only provide new insights into the hitherto frequently ignored Late Palaeolithic production and use of textiles and nets, but also indirectly into the often forgotten but important role of the ‘invisible sex’ in this prehistoric society (Adovasio et al. 2007).
* Author for correspondence.