The bear necessities

7 mins read

A new view of the so-called ‘Lion Man’

The ‘Lion Man’ in the Museum Ulm. This enigmatic ivory figurine found in fragments in a German cave is a fine example of Ice Age art. But when was it made, and what does it show? (Photo: O Kuchar © Museum Ulm)

Fragments of ivory found in a German cave on the eve of war have been reconstructed to create a magnificent ‘Lion Man’. This figure has been feted as the earliest representation of a god, and a representation of shamanic beliefs, but how secure are these interpretations? Elle Clifford and Paul Bahn investigate the true identity of the Lion Man.

In summer 2019, the BBC broadcast a TV series that looked at three major archaeological discoveries of 1939. The second show concerned the Ice Age figurine from south-west Germany known as the ‘Lion Man’. In the programme, the figure was described as the ‘oldest known representational work of art in human history’ and ‘clearly a shamanistic object [representing] the transformation from the human to the animal’ and possibly ‘the very first representation of a god’.

We strongly disagree with all of these claims, and have argued in a forthcoming paper (see ‘Further information’ below) that the so-called ‘Lion Man’ is nothing of the kind, and probably not as ancient as is commonly believed.

A study in ivory

Hundreds of fragments of an ivory carving were found by geologist Otto Völzing during excavations in Hohlenstein-Stadel cave, south-west Germany, in 1939. Work had to be abandoned the very next day, as war loomed. The ivory fragments were looked at decades later by Joachim Hahn, who found them in a box full of animal bones. Hahn said that he and two colleagues put about 200 fragments (out of 260) together with UHU glue in 1969. They did this within a few days. Hahn called it a männliche Elfenbeinstatuette (‘human-like ivory statuette’). The ivory’s lamellar structure meant that a good deal of the outer surface had flaked off, and there was much uncertainty in this first reconstruction.

The Hohlenstein-Stadel cave lies in modern Baden-Württemberg, Germany. Excavations within the cave in 1939 revealed two key phases of Ice Age activity, in the Aurignacian (c.40,000-35,000 years ago) and Magdalenian (perhaps c.13,000 years ago) periods. It is unclear which era the ‘Lion Man’ belongs to. (Photo: Wikimedia Commons/Thilo Parg [CC by SA 3.0])

The upper body was especially hard to reconstruct: the head was very incomplete, but clearly zoomorphic. The position of the ears led Hahn to see it as either a bear or a feline. Years later, after more pieces had been added in 1982 by Elisabeth Schmid, especially to the head, Hahn claimed that the head was sicher (‘certainly’) that of a lion. Schmid came to the same conclusion, and her chosen interpretation was then enhanced with putty or modelling wax to fill in the gaps.

New excavations to look for more fragments were carried out by Eberhard Wagner in 1983, but he found that the old holes from 1939 had been backfilled with excavation debris. So the original sediments in which the fragments had been found were certainly no longer available. Clearly, when the outbreak of war terminated digging in 1939, the fill was tossed back with the greatest hurry.

Sekhmet, from the Temple of Mut at Luxor. Could this lion-headed Egyptian goddess have influenced the original reconstruction of the ‘Lion Man’? (Photo: National Museum Copenhagen)

The first truly painstaking restoration of the figurine took place from August 1987 to May 1988. Photos of the countless small fragments show what a terribly difficult task it is to do something with this 3-D jigsaw. During the 1987-1988 work, the fragments were numbered and photographed. The glues used earlier were dissolved. Many of the fragments are astonishingly fragile, so they were impregnated with cellulose nitrate to conserve them. However, despite the best efforts of Ute Wolf, the conservator, it was not possible to attach many of the pieces to the figure – additional pieces were assembled into four larger fragments.

Using the same glue as before, about 230-250 fragments were fitted together. During the restoration of the head, there was some change in the ordering of its fragments – this made the head broader, but the gaps in the face were so troubling that they were deeply filled in. Wolf decided to reconstruct the missing pieces using a mixture made from wax and chalk, which could easily be removed again if additional pieces turned up at a later time. By this point, there was almost as much wax as there was ivory in the head.

A prehistoric Baloo? The latest version of the ‘Lion Man’ has a broad and pronounced snout, but no traces of the whisker pads that might be expected on a lion. Such details could certainly feature in Ice Age art, and appear, for example, in a lion engraving from La Vache. (Photo: Neil Harrison / Dreamstime)

Between 2009 and 2013, 575 more small fragments (most only measuring a few millimetres) were found in new excavations and assumed to be from the figure. There now seemed to be a total of 758 fragments to deal with, although microscopic analysis subsequently showed that 139 pieces were in fact bone or antler. Once again, the carving was dismantled, and the old glue removed (as it was now harder than the ivory). Alas, it was only possible to fit some of the new fragments, while many remain ‘homeless’ – perhaps there was more than one carving at the site? Certainly the restorers themselves have suggested this; and they also emphasise what a huge headache it was to try to fit pieces together that were eroded, tiny, fragile, and from a tusk that had shattered into many different layers, so that the lower part of the figure had 28 separate layers of different thicknesses, and it was virtually impossible to assign fragments to any particular layer. One feels tremendous admiration for the efforts of the restorers, faced with such an intractable and Herculean task, but one does wonder just how much of the finished reconstruction, especially of the body, can be seen as 100% valid rather than a best guess.

Löwenmensch – lion or bear?

Does the ‘Lion Man’ figurine look like a lion-man, a lion, a bear or a bear-man? The orthodox, official view of this figurine is that it is a lion head on a human body. But Ice Age images are not photographs – they are stylised artistic depictions – and we argue that the body has a far closer resemblance to a bear than to a human. The ‘Lion Man’ is often described as having the knees and ankles of a human, but bear anatomy is so similar to human that it can be difficult to differentiate their bones.

Why should a human have a lion head? We cannot help but wonder if the original reconstructors of the figurine were influenced, consciously or unconsciously, by the image of Sekhmet, the Ancient Egyptian goddess who has the head of a lioness.

Schmid saw a striking similarity between the Hohlenstein-Stadel figure’s head and a small ivory carving of a lion head from Vogelherd, though she did admit some differences such as the position of the ears. To our eyes, the Vogelherd head looks nothing like the Hohlenstein-Stadel example, and nor do further stone lion-heads from Kostenki. In fact, no certain Ice Age depictions of big cats bear any real resemblance to the Hohlenstein- Stadel figure – most of them appear incredibly different. Moreover, in the most recent reconstruction, the snout was completed, and proved to be broader and more pronounced – in our view, making it even more bear-like. In fact, it now reminds us somewhat of the head of Baloo the bear in the Disney cartoon of The Jungle Book.

The ‘Adorant’, and Ice Age ivory from Geissenklösterle, which shows a figure with raised arms and either a generous phallus or an animal tail. (Photo: Don Hitchcock, donmaps.com)

The most characteristic features of a lion, other than its mane (which either Ice Age cave lions did not have, or Ice Age artists never depicted) are its whiskers, its teeth, and its tail. The ‘Lion Man’ has no whiskers (unlike the lion engraving from La Vache and some paintings in Chauvet, which have whisker pads – vibrissal follicles – clearly depicted); and even though the figure is said to be ‘smiling’, there are no fangs showing. Furthermore, it has no tail – unless its tail is in the several hundred remaining bits of ivory still looking for a home on the figure. Of course, the body is claimed to be human and so may not have needed a tail, but it is instructive to consider the ‘Adorant’ – an ivory bas-relief from Geissenklösterle – which has recently been reinterpreted by Nicholas Conard as another ‘lion man’.

This bas-relief is so damaged that its subject is impossible to identify with any confidence. Hahn – the man who launched the ‘Lion Man’ on the world – saw the heavily weathered Geissenklösterle relief as a human figure. He was uncertain whether it had a long phallus or an animal tail. And virtually every other account agreed with him, seeing it as a human signalling, or dancing, or as a ‘worshipper’. Whatever the subject is, its raised arms and possible tail make it completely different from the Hohlenstein-Stadel figure.

A standing brown bear. Is this what the creator of the ‘Lion Man’ had in mind? (Photo: Volodymyr Byrdyak / Dreamstime)

Lions don’t stand up. So why would a lion be depicted standing up? Bears, however, often stand up. Ice Age humans were certainly very familiar with bear anatomy and bear behaviour, so we think it infinitely more plausible that the 31cm figure represents a standing bear. A third figurine that Conard has called a ‘lion man’, a tiny (only 2.5cm high) and incomplete figure from Hohle Fels, is so vague that it could easily be a standing bear rather than a therianthrope (part animal, part human).

There are several persuasive arguments that suggest a connection between bears and humans, and make it more likely that the ‘Lion Man’ figure is that of a bear. Bears are more like humans than any other species, particularly in their skeleton and their footprints. Bears are adept swimmers, climbers, and runners. The gestation period for bears is 6-9 months, and they breastfeed their young in much the same way as human mothers. They have a wide variety of dietary preferences, although cave bears are thought to have been omnivorous, like humans. Bears used shelters and caves to sleep, to give birth, and to hibernate. Bears sleeping for long periods and then waking must surely have intrigued ancient humans. Palaeolithic people removed the skins, claws, and teeth of bears, and the latter were worn as pendants, and possibly prized in some way.

In short, we argue that bears must have been of huge significance to Palaeolithic people because of their anatomy, their habits, and their behaviour. It is therefore far more likely, in our eyes, that the figurine depicts a standing bear.

Further information
A more detailed exploration of this subject, entitled ‘If the Cat Fits… a new look at the so-called “Lion Man” from Hohlenstein-Stadel’ will be published soon in a special issue of Die Kunde.

This is an extract of an article featured in issue 100 of Current World ArchaeologyClick here for more information about subscribing to the magazine.


  1. Lion Man, Bear Man or bear…. it is still an inspiring artefact! I was lucky enough to see it in London and it was the star of the show….

  2. Amazing price of art , I do think it more likely to be a bear , thank you you for the information, our past is fascinating

Leave a Reply

Your email address will not be published.