News Feature: What was the first “art”? How would we know?

Recently discovered cave paintings and bone carvings offer new perspectives on long-held questions about art’s origins—not to mention the nature of art itself.

Archaeologist Adam Brumm recalls the moment in late 2017 when his phone buzzed with a WhatsApp message that included a rather astounding image: three little pigs leaping across the limestone walls of Leang Tedongnge cave on the Indonesian island of Sulawesi.

Researchers believe this 45,500-year-old Indonesian cave painting, apparently of a pig, is the oldest known depiction of the animal world. It’s among several recently unearthed prehistoric images that are shedding new light on the dawn of art. Image credit: Maxime Aubert (Griffith University, Nathan QLD, Australia).

It was a few months after the field season, and Brumm sat in his office at Griffith University in Brisbane, Australia. His field team had remained in Indonesia. Their notes to him that afternoon were nonchalant, Brumm recalls, along the lines of “Oh by the way, we found this spectacular cave painting.” He was quickly scrolling through the messages on his iPhone, when he saw the first images of the pigs flash by on his screen. “I nearly had a heart attack,” Brumm says. “They were absolutely incredible. I replied, at 3:58 pm: ‘Holy hell!!!!! Amazing pig paintings!!!’” At least 45,500 years ago, a human hand had painted the pigs in ochre, making them the oldest known examples of figurative art by at least several thousand years—and, by some standards, the oldest artwork in the world (1).

But which painting, drawing, or carving deserves the superlative “oldest” is debatable. The moniker depends, in part, on how archaeologists define art itself. The Sulawesi pigs are certainly the oldest known figurative, or representational, art. That entails work depicting objects from life, Brumm explains, such that an average observer would glance at the paintings and recognize them as pigs rather than abstractions. Representational art is common in art history, from Greek Hellenistic marble goddesses to First Nations masks of killer whales and ravens in the Americas.

And then there’s nonfigurative, or nonrepresentational, expression. Examples abound in modern art, from Mark Rothko’s color blocks on canvas to Anne Truitt’s colorful minimalist pillars. The earliest abstract markings appear hundreds of thousands of years earlier than the Sulawesi pigs, when humans and other hominids began etching parallel lines, grids, and circles into shell and bone. But archaeologists disagree as to whether these are in fact the earliest glimmers of artistic expression.

The large limestone cave known as Leang Tedongnge is nestled in a valley on the Indonesian island of Sulawesi. The “three little pigs” painted there leap across a back wall in what appears to be a depiction of some social interaction, perhaps mating or fighting. Image credit: Ratno Sardi (photographer).

Recent discoveries of both figurative and abstract images, including 51,000-year-old etchings in bone made by Neanderthals in Germany, are prompting researchers and others to ask the question: When did art truly arise, and in what species? Adding to the archaeological evidence are modern day cognitive experiments using photographs of ancient line carvings. The work could help researchers recognize whether the carvings were originally intended as “art”—images and etchings purposefully created to stimulate the visual senses.

By Hand and Firelight

“There’s a big problem in studying Paleolithic art, in that we don’t know what art is,” confesses evolutionary cognitive archaeologist Dietrich Stout at Emory University in Atlanta, GA. When amateur archaeologist Marcelino Sanz de Sautuola stumbled across the first gallery of cave paintings at Altamira, Spain, a discovery he published in 1880, archaeologists assumed that the art was forged (2). Even decades later, when the paintings were dated to a minimum of 13,500 years ago, Stout says that researchers didn’t spend much time on definitions of art; the sense was, “you know it when you see it.” Here were paintings requiring technical skill, a range of materials, and probably torchlight for the artists to see while painting. And they looked like paintings a 19th century European gallery would display, heavily favoring the figurative over the abstract.

All this suggested to early archaeologists that European cave painters must have been cognitively fully modern, undergoing an abrupt leap from crude to relatively sophisticated, which allowed them to take images from the real world, held in the mind’s eye, and then create a physical representation of those images by using lines on rock. Despite a lack of direct evidence, archaeologists at the time assumed these artists must have had all kinds of other “modern” characteristics as well, such as language, culture, abstract thought, and religion, Stout explains. Art, as those early archaeologists imagined it, came out of a creative revolution coinciding with the earliest European cave paintings, dated to about 30,000 years ago at their oldest in Chauvet, France.

This engraved giant deer bone, made by Neanderthals and found at a cave site in Germany, suggests that the origins of art are earlier than many contend. Image credit: Lower Saxony State Office for Cultural Heritage/Volker Minkus (photographer).

But these would turn out to be hasty conclusions. “That’s a problematic concept,” Stout says, noting that brain size hasn’t changed much for Homo sapiens in the last 500,000 years. Also, plenty of modern cultures make both figurative and abstract art. So there’s no good reason to assume that European figurative artworks are evidence of a cognitive leap that enabled art-making, as opposed to simply heralding the arrival of new groups in Europe that showed up with representational artistic skills. “I’d say the key point is that the idea of a ‘revolution’ was an artifact of looking only at the archaeological record of Europe,” he says. Indications of migration in Europe had created an impression of sudden revolutionary change, which is now known to be more gradual elsewhere. “Around the world today we see lots of diversity in the expression of ‘art’ by modern Homo sapiens,” he explains. “There is no real reason anymore to see figurative art as a key indicator of a broader package of ‘modern’ traits.”

Today, Stout sees several different camps in archaeology, adhering to slightly different definitions—all of which get to the heart of a decidedly modern-day conundrum: What is art? The most common criterion for what’s considered art is behavior without any apparent practical use. Take the red pigment ochre, for instance, which humans used to paint the Sulawesi pigs. The pigment, also found at older sites pre-dating figurative art, may have been used artistically as face paint or other body ornamentation, but that’s hard to prove. Ochre also has practical uses, for example in processing animal hides. Still other archaeologists would like to see stronger evidence that the art was actually intended to convey some kind of aesthetic principle or meaning, Stout notes. Beads, for instance, are decorative but can also signal group identity. But again, it’s hard to rule out practical uses.

In their 2020 study, Tylén and collaborators showed modern participants engraving patterns, or “experimental stimuli,” distilled from the original outlines of engravings on rock artifacts that were made over a period of 30,000 years. Younger artifacts had more attention-grabbing, intentional, and memorable patterns than older ones, but they did not necessarily contain more distinguishable patterns. Reprinted from ref. 7. Photographic materials adapted from refs. 11 and 12, with permission from Elsevier. Original outlines reproduced from ref. 13, with permission from Cambridge University Press.

Still, there are some growing areas of agreement. The notion that modern human art began in Europe some 40,000 years ago has been “a crumbling edifice,” Brumm notes. The Sulawesi pigs firmly usher it out; the figures are representational and pre-date any equivalent figurative depiction in Europe easily by 5,000 years, he adds. And the pigs may not even be the oldest figurative art. Cave scenes of hunter-gatherer life from India to China could possibly be even older.

One big challenge, Brumm says, is that most cave paintings are difficult to date. In Indonesia, for example, archaeologists first reported cave paintings in the same region as Leang Tedongnge in the 1950s, created by blowing a mouthful of pigment around an outstretched hand, pressed to a cave wall. The assumption until recently was that the hand stencils couldn’t be very old, Brumm says, because Indonesia is so hot and humid. The paint should quickly erode.

But in 2011, Brumm and colleagues noticed popcorn-shaped growths formed over the top of some Indonesian paintings in limestone caves, which would have formed sometime after the paintings appeared. The growths turned out to be precipitated calcite deposits, similar to stalactites, which build up in hundreds of hardened layers over centuries. By measuring the ratio of uranium to a product of its decay, thorium, Brumm and colleagues dated the oldest layers of the popcorn deposits. In 2014, his team dated popcorns on a hand stencil to at least 40,000 years ago (3). The subsequent discovery of a hunting scene in another Indonesian cave in 2019 pushed back the oldest figurative art to 43,900 years ago (4). And then, in 2017, Brumm’s team found the pigs and dated them to 45,500 years ago by using a calcite deposit, making that work, published in 2021, so far, “the oldest representational painting in the world,” Brumm says. The discoveries in Sulawesi could imply that representational art began in Asia, but more likely, Brumm says, it’s just part of a trail of representational art through human history. He expects the oldest rock art will eventually turn up from before Homo sapiens’ diaspora out of Africa.

Marks and Meaning

Other interpretations for what constitutes “art” could suggest a different origin story—one that doesn’t necessarily begin with our species. Evidence of abstract images dates as far back as 500,000 years ago, when Homo erectus etched zig zag lines into a seashell in Java (5). And just this year, archaeologist Dirk Leder discovered 51,000-year-old abstract triple L-shaped patterns carved in deer bone, placed between a cave bear skull and two deer shoulder blades, in a Neanderthal cave dwelling in Germany (6). Microscopy and CT scans revealed a three-dimensional picture of the bone in which the engravings are precisely spaced, with cuts at neat angles, suggesting they were made deliberately rather than as accidental hacking marks from a tool. The carving’s estimated age of 51,000 years is based on radiocarbon dating of collagen in the bone itself, which, combined with tools characteristic of Neanderthals at the site, implies that it was made several thousand years before Homo sapiens arrived in Europe.

The takeaway is twofold, says Leder, who is at the Lower Saxony State Office for Cultural Heritage in Hanover, Germany. First, Neanderthals were capable of constructing deliberate symbolic expressions, a notion that has been debated in the past. And second, he says, art’s origins should be pushed back “not just to 45,000 years ago, but a much longer timeframe.” Leder expects that both Homo sapiens and Neanderthals were making expressive and communicative carvings and drawings—even if not representational—off and on throughout their history, and that although debate will likely continue, artistic expressions made by Neanderthals, and possibly even earlier hominins, will be recognized more and more.

If Leder’s and others’ definitions of art seem subjective, they are. But cognitive scientists are now trying to offer a more objective assessment by pinpointing the origins of symbolism—which some treat as a synonym for the origins of art itself. Some archaeologists argue that each scribble denotes some meaning. A circle for instance, might mean “horse,” although the shape looks nothing like the animal. Others see the carvings as aesthetically interesting, but otherwise meaningless, decorations.

To help resolve the debate, cognitive scientist Kristian Tylén, at Aarhus University in Denmark, collaborated with archaeologists beginning in 2017, with a resulting study published in 2020, using up-to-100,000-year-old abstract carvings from South Africa’s Blombos cave and other sites to probe the beginnings of symbolic behavior (7). Tylén and collaborators designed a series of cognitive experiments for modern-day human subjects, testing, for instance, how memorable and how discriminable each carved pattern is. If there were some adaptive pressure to make patterns mean specific things over time, for instance evolving from simple grids into pictograms or words, archaeologists would expect the patterns to become more memorable and easier to tell apart over thousands of years.

The researchers showed subjects the ancient engravings, for instance flashing the images on a monitor for a few seconds and then asking participants to redraw the pattern they’d just seen from memory. Consistent with the evolution of symbolism, younger artifacts were easier to recall than older ones. However, in another set of experiments in the same study, participants matched identical patterns as quickly as possible. Here, the researchers did not find meaningful differences in the discriminability of younger or older patterns. “We should think about these scribbles as proto-art,” Tylén says. “They were made to be pretty and stimulate the visual system, more than to be communicative signs that denote a particular meaning.” Derek Hodgson, a neuro-archaeologist now semi-retired from the University of York in England, had published a 2019 review proposing essentially the same conclusion (8). He suspects that the first intentional markings, often parallel lines or crosses, resulted from a bias in the visual cortex, in which neurons are particularly sensitive to horizontal and vertical lines. If the marks had greater symbolic meaning, Hodgson adds, he would expect to see more variation between them from one culture to the next, just as languages vary. But instead, the earliest grids, V-shapes, and lines turn up in a limited number of configurations around the world, suggesting to Hodgson that they were visually interesting but not explicitly meaningful.

Hodgson’s theory, however, was met with swift disagreement in a 2019 paper by archaeologist Francesco d’Errico and collaborators (9). D’Errico pointed out that he’d used functional magnetic resonance imaging (fMRI) in a prior study to identify the brain areas stimulated by a variety of images, including 540,000- to 30,000-year-old engravings, as well as landscapes, objects, words with no meaning in alphabetic writing, and fragments of Linear B ancient writing systems, as well as the scrambled versions of all these stimuli (10). D’Errico and coauthors found that the scrambled versions of all the stimuli were processed in participants’ primary visual cortexes, indicating simple visual perception without further processing by the brain. But the engravings activated brain regions in a similar pattern to how objects are perceived, suggesting that they are processed as organized visual representations and may have been used to attach symbolic meaning.

One limitation of modern studies is that they use modern human subjects. In d’Errico’s work, for example, modern brain activity in areas associated with complex forms is interpreted as evidence of symbolic meaning thousands of years ago. “The problem is, modern humans have learned to read and write, so we have highly symbolic ability attached to that area of the brain,” Hodgson says. Whether that brain area behaved similarly before people had written language is impossible to tell, he adds. D’Errico plans to address the limitation of working with modern humans in forthcoming unpublished work, by comparing patterns of brain activity in archaeologists to those of non-experts. Both the experts and non-experts will look at a collection of real engravings and unintentional, natural markings that look like engravings. The archaeologists should quickly spot the unintentional markings by using a motor area of the brain that remembers the hand motions required to make a real carving. Comparing the responses in the brains of modern experts with those in the brains of modern non-experts is fairer, d’Errico says, than using the response of modern participants to infer symbolic meaning in the past.

On the phone from the University of Bordeaux in France, d’Errico is careful to note that “art is a very ambiguous notion.” He, Tylén, Hodgson, and their collaborators are not explicitly debating the dawn of art, per se. Rather, they’re arguing over the emergence and evolution of symbolic material culture. The San of the Kalahari use hatch marks to denote, as one example, ownerships of their arrows. And although those markings have a symbolic role, the San would consider them to be craftsmanship, not art. Still, when symbolism emerged is a relevant question to art’s beginnings, as some archaeologists argue that images must be symbolic to be artistic.

D’Errico personally does not subscribe to that view. Symbolism and art aren’t necessarily related, he says. Even if researchers arrive at some agreement on the definition of symbolism, they’d still be hard pressed to reach a consensus on the definition of art itself—and such a consensus could still deviate from the beliefs of the artists themselves. For his part, without proof that a given society had an independent role for the artist, d’Errico would be hesitant to say that group made “art” per se. “I think art begins in a society when someone gets a societal role as artist, independent of what they’re producing,” he says. “The society is recognizing someone has special training and taste that others cannot do.”

So when was the earliest art? The answer depends on which definition of art an archaeologist adheres to and which specific research questions they’re asking. The conceit that abstract art is primitive, and then evolved into representational art, seems increasingly flawed. Tylén, for his part, suspects that abstract art and figurative work such as the Sulawesi pigs could have independent origins. In Europe, some Paleolithic animal figurines, which appear after the dawn of representational art, are decorated with abstract engravings similar to much older patterns. That suggests two simultaneous forms of expression: One that’s purely aesthetic and abstract, and one that’s symbolic, representational, and perhaps requires more complex cognition.

Looking ahead, the Sulawesi pigs may not be the oldest figurative art for long. Brumm is now applying for funding to explore the islands between Sulawesi and Papua Indonesia, following the path of those who first migrated down the Indonesian island chain and onto the northern tip of Australia some 65,000 years ago. “We hope to find earlier rock art in these unexplored islands east of Sulawesi,” Brumm says. If paintings there are indeed as old as the first inhabitants of Australia, it would suggest that wayfarers have carried artistic expression for at least twice as long as some 19th- and 20th-century archaeologists had thought.

It’s these big questions—and historical revisions—that are piquing the interest of archaeologists and others. “It’s one of the biggest questions in archaeology,” Brumm says. “When did our ancestors or close relatives start to produce markings or forms on the continuum of art, and why did they do it?”