The Work of Making Meaning: Why Ice Age Humans Carved Birds at Hohle Fels

Two mammoth-ivory birds recovered from the Aurignacian layers of Hohle Fels are scarcely larger than a thumbnail, yet their makers invested considerable skill in objects that neither cut, scraped nor hunted. Their discovery revives a difficult archaeological question: what did image-making actually do for the first Homo sapiens communities of Ice Age Europe?

In 2025, excavators working inside Hohle Fels Cave in the Swabian Jura recovered two pieces of mammoth ivory from the lower Aurignacian deposits. They were almost absurdly small. One is a nearly complete bird with its body flattened and its head drawn close to the torso; the other spreads its wings, although the left wing is incomplete. Both belong to archaeological deposits dating to roughly 40,000 years ago. They were found about 1.5 metres apart in the same layer, close enough to invite comparison but not close enough to establish that they were made together. They might be the work of one carver. They might equally have been separated by years or generations.

The smaller figure, the bird with open wings, measures only about two centimetres long, one centimetre wide and half a centimetre high. At approximately 1.3 and 0.7 grams respectively, the carvings are among the smallest works known from the Ice Age art of the Swabian Jura. Their scale has not resulted in simplification. The resting bird has prominent rounded eyes. The second has a pronounced beak and parallel incisions on its wings that seem to indicate plumage. The excavators have suggested ptarmigan among the possible identifications, partly because ptarmigan bones occur frequently in the regional archaeological record, although grouse and even birds of prey remain possibilities. The posture of the first bird may represent brooding; the second might be flying, landing, threatening another animal or displaying during courtship. None of those identifications is secure.

That uncertainty is useful. It directs attention away from the temptation to treat Palaeolithic images as pictures with captions waiting to be recovered. What the carvings demonstrate securely is more modest, and in some ways more interesting: somebody observed birds closely enough to distinguish posture, body form and feathering, selected mammoth ivory as the material, reduced it to a minute three-dimensional figure and preserved details that other people could apparently recognise.

The difficult question begins after that.

A cave already crowded with images and sound

Hohle Fels is not an isolated occurrence of early image-making. It belongs to an unusual archaeological landscape in the Ach and Lone valleys, where six caves containing important Aurignacian deposits form the UNESCO World Heritage property Caves and Ice Age Art in the Swabian Jura. The regional Aurignacian sequence spans approximately 43,000 to 33,000 years ago and is associated with the first established populations of Homo sapiens in this part of central Europe. Figurines, personal ornaments and musical instruments have all been recovered from these caves.

Hohle Fels alone has supplied some of the objects around which discussion of early European art has repeatedly turned. A mammoth-ivory waterbird, published in 2003, means that the two new finds bring the number of bird representations from the cave to three. Other early carvings include animal and human-animal forms. The famous female figure usually called the Venus of Hohle Fels came from basal Aurignacian deposits and was carved from mammoth ivory. The cave has also produced flutes manufactured from bird bone and ivory, part of evidence showing that instrumental music was already established in the Swabian Aurignacian.

Elsewhere in the Swabian Jura, the repertoire includes mammoths, horses, cave lions and hybrid beings. The best known is the approximately 31-centimetre-high Löwenmensch from Hohlenstein-Stadel, a figure combining a human-like body with a cave-lion head. The effect of seeing the assemblage together is important. This was not a society that occasionally scratched an idle mark into bone. Image-making, ornamentation and music recur across different places, materials and technical traditions.

The chronology nevertheless needs care. Archaeologists do not possess a radiocarbon date taken from each ivory bird. Their age comes primarily from stratigraphic association with the Aurignacian deposits and the regional dating framework. Improved radiocarbon programmes at sites such as Geißenklösterle have repeatedly refined the chronology of the early Aurignacian and its artistic and musical assemblages. “Forty thousand years old” is therefore a useful shorthand, not a birthday.

The cost of carving something that does not cut

Calling the birds “non-functional” would be premature. Calling them non-utilitarian is safer. They are not projectile points, hide scrapers or wedges. Their value did not lie in performing an obvious mechanical task.

That distinction matters because mammoth ivory is not an effortless medium. It is dense, structurally complex and unforgiving of mistakes. Experimental work reproducing Upper Palaeolithic ivory objects with stone tools has shown the combination of planning, controlled cutting, scraping, abrasion and accumulated manual experience required to work it successfully. Aurignacian craftspeople were not merely exploiting convenient fragments; elsewhere in the Swabian record, sophisticated methods were used to divide tusks and manufacture tools, ornaments, figurines and even components of ivory flutes.

Miniaturisation adds another problem. A failed stroke on a large blank might be corrected. On an object two centimetres long, very little material separates detail from destruction.

Someone therefore spent time making the birds when that time could have been spent on tasks whose contribution to survival seems more immediate. The old archaeological opposition between “practical technology” and “art” tends to make this appear irrational. It need not have been.

A thing does not have to kill an animal to have practical consequences.

Human societies depend on information about relationships: who belongs to which group, who can be trusted, who shares obligations, where alliances extend, what stories and rules are held in common. Such information is often communicated materially. Clothing, jewellery, insignia, colour, music and images can all make relationships visible or audible. Palaeolithic personal ornaments are especially important in this respect because their geographical distributions and stylistic variation have long been investigated as possible markers of social identities and networks. Exact linguistic or ethnic boundaries cannot simply be read from bead types, but the patterned distribution of ornaments is difficult to explain if appearance carried no socially recognised information.

The figurines may have participated in a comparable world of communication.

Art as social technology

This does not mean that the new birds were necessarily badges of clan identity. Nor does it mean that they were religious objects. Those are hypotheses, not observations.

There is no inscription saying “talisman”, no burial association demonstrating that a bird accompanied a particular category of person, and no architectural context that identifies a shrine. A small portable figure could have been worn, carried in a pouch, handled during storytelling, exchanged, given to a child, used in ritual, kept as a valued possession or discarded after having several different roles. Some of those possibilities would leave exactly the same archaeological trace.

What can be argued more securely is that the wider Aurignacian record contains evidence for conventions extending beyond individual acts of craftsmanship.

A 2026 study by Christian Bentz and Ewa Dutkiewicz analysed thousands of geometric marks on more than 200 Aurignacian mobile objects, including figurines, tools and ornaments. Their statistical results do not show that the marks constituted writing. The authors explicitly reject that conclusion. But the signs were not distributed randomly: certain kinds of marks and sequences were preferentially associated with particular object categories, suggesting deliberate conventions transmitted across time. Some statistical properties resemble those of the much later earliest proto-cuneiform sign sequences, without implying that the two systems encoded information in the same way.

That distinction is crucial. What survives is evidence for patterned behaviour, not a deciphered message.

The same point may apply to figurative art. A mammoth, lion or bird could have meant different things in different contexts, but the repeated ability to recognise and reproduce such forms implies shared visual categories. Making an image presupposes more than eyesight. The maker must decide which features are necessary for another observer to identify the subject. In that sense, even a very small bird is a social object: representation only works if maker and viewer participate, at least partly, in the same conventions of seeing.

Music extends the argument beyond vision. The Swabian flutes indicate not simply that individuals could manufacture elaborate instruments but that performances were possible. Music can coordinate people in time in a way an image cannot. Archaeology cannot reconstruct the melodies, performers or occasions, but the technological investment in instrument manufacture makes casual noise-making an inadequate description.

Art, ornament and music may therefore belong less to a category of prehistoric luxury than to a set of technologies for creating shared experience.

Why birds?

The new finds introduce a more local problem.

Across the famous Swabian ivory assemblage, mammoths, horses and large carnivores receive considerable attention. Yet Hohle Fels has now yielded three bird figures, whereas the other principal artistic sites in the World Heritage area have, so far, produced no comparable bird figurines. Conard and colleagues have suggested that individual sites may have developed distinct artistic “signatures”, perhaps reflecting different workshops or local preferences.

That is plausible, although the sample remains small. Archaeological absence is notoriously treacherous. Ivory breaks; excavation histories differ; large parts of cave deposits remain unexcavated; and a few additional finds could alter the pattern quickly. Three birds at one site are enough to attract attention, not enough to establish a bird cult.

Still, the manner in which the new figures were carved may be relevant. They appear to represent behaviour rather than merely species. The resting body, tucked limbs and retracted head of one bird prompted the suggestion of brooding. The open-winged animal captures some more active moment. If those readings are approximately correct, the carver was interested not simply in what a bird looked like but in what birds did.

For hunter-gatherers living in a strongly seasonal Pleistocene environment, such observations were hardly peripheral. Birds move, congregate, nest and migrate. Their behaviour can signal seasonal change, water, carcasses or disturbance. They can also carry meanings entirely unrelated to subsistence. The difficulty is that the same close ecological knowledge could support both practical and symbolic associations.

Archaeology cannot separate them here.

Perhaps it should not try too hard. The modern tendency to divide “ecological knowledge” from “art” and “religion” may itself impose categories that were not meaningful to Aurignacian people. A bird could simultaneously be prey, a seasonal indicator, a remembered encounter, a narrative character and a creature endowed with qualities people found socially or cosmologically important.

The figurine does not tell us which.

This was not the birth of art

For much of the twentieth century, the European Upper Palaeolithic occupied an outsized position in narratives about the emergence of modern human behaviour. The concentration of sculpture, cave imagery, ornaments and musical instruments around 40,000 years ago appeared to support the idea of a relatively sudden “creative explosion” associated with Homo sapiens entering Europe.

The global archaeological record no longer allows so simple a story.

At Blombos Cave in South Africa, intentional abstract markings are much older: a cross-hatched drawing made with ochre on stone dates to around 73,000 years ago, and still earlier Middle Stone Age deposits contain engraved ochre, shell ornaments and evidence for pigment processing. These finds do not look like the Hohle Fels birds, but they demonstrate that externalising information through colour, pattern and bodily display had a much deeper history among Homo sapiens.

Nor is early figurative imagery confined to Europe. Laser-ablation uranium-series dating in Sulawesi has established a minimum age of at least 51,200 years for a scene at Leang Karampuang showing human-like figures interacting with a pig. In 2026, researchers reported an even older minimum age, at least 67,800 years, for a hand stencil on Muna Island in southeastern Sulawesi. These are minimum ages derived from calcite associated with the paintings, rather than direct dates on pigment, but they decisively displace Europe as the necessary geographical birthplace of graphic expression.

Evidence associated with Neanderthals makes the evolutionary picture more complicated still. Pigment use, personal ornaments and other forms of non-utilitarian material behaviour are no longer easily assigned to Homo sapiens alone, although some claims for very early Neanderthal cave painting remain methodologically disputed. It is therefore increasingly unhelpful to treat a figurine as a simple fossil marker for a newly evolved “modern mind”.

What is unusual in the Swabian Jura is the density and diversity of the surviving evidence within a comparatively restricted landscape: representational sculpture, personal ornamentation, musical instruments and repeated geometric marking appearing within early Upper Palaeolithic communities. That pattern remains archaeologically important even though it is no longer the opening chapter of a universal history of art.

Did culture help Homo sapiens flourish?

This brings us to the most ambitious interpretation attached to the new birds.

The Nature feature reporting their discovery asks whether capacities reflected in such objects might have contributed to the success of early Homo sapiens as populations expanded through Europe. There is a reasonable hypothesis behind the question. Humans capable of maintaining larger social networks, transmitting detailed knowledge, creating durable group identities and teaching complex skills could potentially cope better with environmental uncertainty and population dispersal. Material symbols might help sustain relationships between people who were not constantly together. Music and shared performances could reinforce cooperation. Ornamentation could make affiliations visible during encounters between groups.

But there is a large inferential distance between that model and two ivory birds.

No archaeological dataset presently demonstrates that figurative art caused Homo sapiens populations to expand, survive climatic instability or outcompete Neanderthals. Even the language of “success” requires caution. Population histories during the Middle-to-Upper Palaeolithic transition were complicated, involving dispersals, local disappearances, replacement and interbreeding rather than a single advancing population equipped with a decisive cultural invention.

There is also a preservation problem. The Swabian caves are unusually favourable archives and have been excavated intensively for generations. Organic objects made from wood, fibre, hide or feathers have overwhelmingly disappeared from most Pleistocene sites. We therefore compare surviving archaeological records, not complete cultural systems. A society represented today by stone tools may once have possessed elaborate performances, decorated clothing or perishable images that left almost nothing behind.

The safest conclusion is consequently more restrained.

The Hohle Fels birds do not reveal why art began, because art did not begin at Hohle Fels. They do not prove a religion, a clan system or an evolutionary advantage. They do demonstrate that by roughly 40,000 years ago people in the Upper Danube region were prepared to invest highly developed craftsmanship in miniature representations whose value cannot be reduced to mechanical utility. They observed animals carefully, worked within traditions of ivory manufacture, inhabited a world already containing figurative images, ornaments, music and recurring graphic signs, and passed at least some of those practices between people and probably across generations.

That combination is archaeologically more informative than simply calling the birds beautiful.

The unresolved problem is not whether Aurignacian people were capable of art. That question disappeared long ago. It is how images, sounds and marked objects operated inside the networks of knowledge and relationships that allowed small hunter-gatherer communities to remain recognisably connected across landscapes and generations.

Two birds weighing less than two grams cannot answer that question. They make it considerably harder to ignore.

Selected Sources and Further Reading

1. Conard, Nicholas J., Alexander Janas & Mohsen Zeidi. 2026. “Hochwertige Funde aus dem Mittelpaläolithikum und zwei aurignacienzeitliche Vogeldarstellungen aus dem Hohle Fels bei Schelklingen.” Archäologische Ausgrabungen in Baden-Württemberg 2025. The University of Tübingen bibliographic notice lists the contribution at pp. 46–52.
2. Callaway, Ewen. 2026. “40,000-year-old bird carvings provide clues to how ice age humans flourished.” Nature 655, 1120–1122. DOI: 10.1038/d41586-026-02312-y.
3. Conard, Nicholas J. 2003. “Palaeolithic ivory sculptures from southwestern Germany and the origins of figurative art.” Nature 426: 830–832. DOI: 10.1038/nature02186.
4. Conard, Nicholas J. 2009. “A female figurine from the basal Aurignacian of Hohle Fels Cave in southwestern Germany.” Nature 459: 248–252. DOI: 10.1038/nature07995.
5. Conard, Nicholas J., Maria Malina & Susanne C. Münzel. 2009. “New flutes document the earliest musical tradition in southwestern Germany.” Nature 460: 737–740. DOI: 10.1038/nature08169.
6. Higham, Thomas et al. 2012. “Testing models for the beginnings of the Aurignacian and the advent of figurative art and music: The radiocarbon chronology of Geißenklösterle.” Journal of Human Evolution 62: 664–676. DOI: 10.1016/j.jhevol.2012.03.003.
7. Hein, Wulf. 2018. “Tusks and tools – Experiments in carving mammoth ivory.” L’Anthropologie 122: 437–446. DOI: 10.1016/j.anthro.2018.05.001.
8. Vanhaeren, Marian & Francesco d’Errico. 2006. “Aurignacian ethno-linguistic geography of Europe revealed by personal ornaments.” Journal of Archaeological Science 33: 1105–1128. DOI: 10.1016/j.jas.2005.11.017.
9. Bentz, Christian & Ewa Dutkiewicz. 2026. “Humans 40,000 y ago developed a system of conventional signs.” Proceedings of the National Academy of Sciences. DOI: 10.1073/pnas.2520385123.
10. Henshilwood, Christopher S. et al. 2018. “An abstract drawing from the 73,000-year-old levels at Blombos Cave, South Africa.” Nature 562: 115–118. DOI: 10.1038/s41586-018-0514-3.
11. Oktaviana, Adhi Agus et al. 2024. “Narrative cave art in Indonesia by 51,200 years ago.” Nature 631: 814–818. DOI: 10.1038/s41586-024-07541-7.
12. Oktaviana, Adhi Agus et al. 2026. “Rock art from at least 67,800 years ago in Sulawesi.” Nature. DOI: 10.1038/s41586-025-09968-y.

Suggested keywords

Hohle Fels; Aurignacian; Ice Age art; Palaeolithic art; mammoth ivory; Swabian Jura; Homo sapiens; symbolic behaviour; portable art; figurines; prehistoric music; cultural transmission

Editorial summary

The two newly published Hohle Fels birds matter now not because they push back the first appearance of art, but because they add unusually fine-grained evidence to a rapidly changing archaeological debate about how images, music and shared symbolic conventions functioned within the social worlds of early Homo sapiens.