Gargas

Why this is extraordinary

On 11 June 1905 a doctor, Félix Régnault, recorded hands sprayed onto the wall of a cave at Aventignan, in the Neste valley of the Hautes-Pyrénées. The other French rooms are on the France page. Counts of the hands have climbed ever since, because faint ones appear and vanish with the damp. The first published tally was 80. A 1976 inventory reached 231 negative hands: 143 black, 80 red, 2 bistre, 5 ochre and 1 white. Of those, 124 are clear enough to read. Ten are complete. One hundred and fourteen are not. The single most common incomplete pattern, 59 of the 114, has the four fingers reduced to one joint and the thumb left whole. For a long time the cave was called the cave of the mutilated hands. That name already chooses a reading.

A comparison with living hands read the stencils as both sexes and every age from infant to adult. That is a size reading, not a child buried in the floor. The paint was blown around the hand in almost every case. One hand is white paste pressed on. Some of the blown hands were touched up afterward with dry black. In the same sanctuary, stencils of thumbs are shown bent, in profile, in two series, five and three. Those are a digit folded on purpose and then sprayed. They share the wall with the shortened hands.

Three readings of the shortened fingers have been in print for a century, and they do not share a winner. One is injury. Cold, frostbite, and a circulation illness, Raynaud's disease, were all proposed, because an Ice Age winter can wreck a finger and because many of the hands are small. The medical objection is specific. Frostbite tends to take every digit, thumb included, and the thumb is the digit the clearest hands keep. In the account used against Raynaud's, the illness eats the tip and rarely the middle joint, while these stencils often lack the middle joint and leave the next finger whole. A second reading is ritual cutting, known from later peoples who shortened a finger as a vow or in mourning. A 2018 survey of those customs keeps the comparison alive. The objection from the wall is the pattern itself. The severe four-finger version is nearly half the clear hands, the combinations run past twenty, and the same combination turns up on hands of different sizes. If the small hands are children, the cut includes them. A third reading is that the fingers were folded, as a sign. Spraying pigment around a living hand with the fingers tucked reproduced the common shapes, including a finger that looks like two joints when it is only bent. A 1967 reading took the folds as a hunting code, or as a code for stories told to initiates. A later check of how a hand can actually bend found that the shapes on the wall are ones a hand can hold in the air, and that shapes which need the rock to prop a finger up are absent. That fits a sign. It does not recover what any one hand meant, and a real cut could land on a shape a hand can also make. The catch is the silhouette. The wall records an outline. It does not record whether the finger was gone, or tucked, when the paint went on.

The date everyone quotes is not on the paint. A splinter of bone, one of several stuck into cracks beside the hand wall, gave 26,860 ± 460 radiocarbon years before present, laboratory number GifA-92369. The number is uncalibrated. It is not a calendar year, and it dates the bone. If the bone and the hands are one gesture, the hands are that old. Dates on the Gravettian floor fall between 25,000 and 27,000 uncalibrated years, the same broad window. The bone sits where a late Aurignacian and the earliest Gravettian overlap on the European lists. In southwest France a result that old is more often filed as Aurignacian. The hands have usually been called Gravettian because of the floor. Both labels are published. Neither was scraped from a finger.

The other European cave long paired with Gargas for shortened fingers is Maltravieso, in Cáceres, and the pattern there was described as a different one. A 2018 uranium-series minimum on calcite over a red hand stencil at Maltravieso is older than 66,700 years. Uranium-series reads the age of cave crust from the decay of uranium in the calcite, and a crust on top of paint is a minimum: the paint is older than the crust. That paper grouped Maltravieso with two other Spanish caves and argued the art is older than 64,800 years, at least 20,000 years before modern humans are placed in Europe, which the paper reads as Neanderthal authorship. A reply says that allowing more dirt in the sample still leaves a minimum of 64,900 years for the Maltravieso crust. The criticism is the one already open at El Castillo. Uranium can leave thin calcite and push the clock older, and the same study produced minimum ages tens of thousands of years apart on parts of one motif at another of the three caves. Gargas does not inherit that age. Its clock is the bone.

A 2007 survey of French decorated caves groups the mouth of Gargas with openings that can admit the winter-solstice sunset, the sun's furthest-south setting of the year. At this latitude that sunset stands in the southwest, and a passage facing it can take the light when the horizon is clear. The same survey sorts almost every cave it studied onto some solar threshold, summer or winter, sunrise or sunset. The summary list does not print a separate azimuth for this mouth, and the survey does not aim the hand panels at a star. A mouth that can see a sunset is a direction. It is not yet a rite, and the hands are inside the hill.

Hands elsewhere on this map: the colour stack at Cueva de las Manos, the crust minima at El Castillo, the drowned passage at Cosquer, the spotted horses at Pech Merle, the pigs under calcite at Maros, the minimum and the maximum at Lubang Jeriji Saléh, and the paint drops at Monte Alegre. What this changes: if the fingers were folded, the wall is a record of a sign, made with adult hands and with small ones, and the old medical name is a choice. If they were cut, or lost to cold, a Pyrenean group carried that pattern into the sanctuary, thumbs spared, in a shape the illness accounts have trouble matching. Either way the bone is not the paint.