For decades, paleoanthropologists pointed to tiny grooves between ancient hominin teeth as proof that early humans and Neandertals used sticks or bone splinters to clean food from their teeth. A new study of 531 wild primates across 27 taxa now challenges that interpretation head-on, showing that the same groove patterns form naturally through ordinary mechanical wear in animals that have never touched a tool. The finding forces a reassessment of behavioral claims built on some of the most celebrated fossils in the human evolutionary record, including specimens from Olduvai Gorge and Middle Pleistocene sites in China.
Why reclassifying ancient tooth grooves changes the fossil record
The stakes are straightforward: if grooves between teeth do not require deliberate picking, then a long chain of inferences about early human behavior collapses. Researchers have previously used interproximal grooves on Neandertal teeth to argue for systematic oral hygiene and even complex social care. A separate study linked grooves on the OH 62 specimen from Olduvai Gorge in Tanzania to intensive meat consumption, reasoning that fibrous animal tissue lodged between teeth would have prompted early Homo to pick at them. Grooves on Middle Pleistocene teeth from Yiyuan in Shandong Province, China, were similarly read as direct evidence of interdental tool use based on their shape and fine parallel striations.
Each of those interpretations assumed the same thing: that the marks could only be made by a foreign object deliberately inserted between teeth. The new primate data, published in the American Journal of Biological Anthropology, dismantles that assumption. Non-carious cervical lesions, or NCCLs, appeared across the comparative sample in locations and with surface textures that closely match the marks previously attributed to toothpicking. Because wild primates have no access to tools, the lesions must arise from other forces, chiefly abrasion from dietary grit and the micro-motion of adjacent teeth grinding against each other during chewing.
The hypothesis that NCCL frequency scales with dietary silicate intake across primate species offers a testable mechanism. Animals feeding in sandy or dust-rich environments would draw grit into the mouth during swallowing, and that grit would abrade the cervical margins of teeth in patterns indistinguishable from supposed toothpick grooves. If confirmed across additional taxa and fossil assemblages, this relationship would predict groove formation in any hominin living in a high-grit paleoenvironment, with no need to invoke tool use at all.
How 531 primates rewrite the toothpick hypothesis
The core evidence comes from a comparative sample of 531 individuals across 27 taxa of wild primates, assembled by a team based at Indiana University Indianapolis. By documenting NCCLs in species ranging from small-bodied monkeys to great apes, the researchers established that groove-like lesions are a common feature of primate dentition under natural conditions. The shape, depth, and location of these lesions overlapped with the diagnostic criteria previously used to identify toothpick grooves in fossil hominins.
That overlap is damaging to the toothpick interpretation because the diagnostic criteria were developed precisely to distinguish deliberate tool marks from incidental wear. Earlier work on Neandertal anterior and posterior teeth described “incipient toothpick grooves” and set morphological benchmarks for identifying them. The Yiyuan study similarly relied on groove morphology and fine parallel striations to argue for toothpicking behavior in eastern China. If wild primates produce the same features without tools, those benchmarks cannot reliably separate deliberate behavior from background wear.
The idea that grit causes interproximal grooves is not entirely new. A classic argument published in the 1980s proposed that sand and soil drawn into the mouth during swallowing could produce approximal grooves, challenging the then-dominant toothpicking and sinew-processing explanations. Separate tribology-focused research using 3D surface texture analysis has since shown how contacting tooth surfaces and micro-motion can shape facets and textures that mimic tool-made features. The new primate study brings these threads together with the largest comparative dataset yet, grounding the mechanical-wear explanation in direct observation rather than theoretical modeling.
Crucially, the authors did not just tally how often lesions appeared; they mapped where, on which teeth, and with what orientations. NCCLs clustered near the cemento-enamel junction, frequently on the mesial or distal surfaces where adjacent teeth make contact. Many took the form of elongated, shallow grooves with smoothed internal surfaces and tapered ends-precisely the configuration that earlier hominin studies had treated as diagnostic of toothpicks. In some primates, multiple grooves intersected or overlapped, echoing the complex patterns seen on heavily worn fossil teeth.
The team also considered alternative behavioral sources of the lesions in wild primates, such as habitual object gnawing or idiosyncratic self-grooming. Field observations, however, did not reveal any consistent behavior that could plausibly involve inserting foreign objects between teeth across so many species and habitats. This absence of a shared behavioral explanation strengthens the argument that a universal mechanical process-grit-laden chewing and tooth-on-tooth movement-is responsible.
Revisiting iconic fossils and behavioral claims
If NCCLs can arise without tools, then some of the most vivid reconstructions of early human life need revisiting. The Olduvai OH 62 specimen, long cited as evidence that early Homo used slivers of bone or wood to dislodge meat fibers, now looks less like a snapshot of a specific behavior and more like a predictable outcome of living in a dusty, open environment. Similarly, Neandertal teeth that once underpinned narratives of routine oral hygiene and caregiving may instead record the cumulative toll of a gritty diet and heavy occlusal loading.
The Middle Pleistocene Yiyuan fossils are an especially telling case. Their grooves were interpreted as signs of interdental cleaning in a population presumed to have at least occasional access to small tools. Yet the new primate data show that similar lesions occur in species with no such access, including forest-dwelling taxa whose diets are rich in fibrous plant matter rather than meat. This convergence suggests that environmental and dietary context, not just technological sophistication, must be front and center when reading behavior from dental micro-damage.
None of this means that early humans and Neandertals never used toothpicks. Ethnographic and archaeological records include clear examples of splinters and small implements that could have served that function, and some fossil grooves may still represent genuine tool marks. The point is narrower but consequential: without independent corroboration-such as embedded foreign material, distinctive micro-striations clearly inconsistent with tooth-on-tooth motion, or direct association with dental tools-interproximal grooves alone can no longer carry the weight of complex behavioral stories.
Gaps in the evidence and what to watch next
The study does not close every door. No quantitative 3D texture data directly comparing the new primate NCCL sample to the specific Neandertal or OH 62 fossils have been published alongside the findings. That kind of head-to-head comparison, matching scan files and lesion coordinate measurements from the 531-specimen repository against the fossil specimens most often cited as toothpicking evidence, would strengthen the case considerably. Without it, defenders of the toothpick interpretation can argue that subtle morphological differences between primate NCCLs and hominin grooves still exist but were not captured at sufficient resolution.
Another open question concerns variation within hominin groups. Even if most grooves on Neandertal or early Homo teeth turn out to be mechanically produced, some individuals may show patterns that deviate from the primate baseline-deeper, more localized lesions, or orientations that do not match expected chewing trajectories. Identifying such outliers will require integrating high-resolution imaging, experimental replication with known tools, and careful attention to the surrounding archaeological context.
Future work is likely to move in three directions at once. First, expanding the comparative primate dataset to include more species and environments will test whether the observed NCCL patterns hold across broader ecological gradients. Second, applying standardized 3D texture metrics to both primate and hominin lesions could reveal whether any consistent micro-wear signatures distinguish tool-made grooves from grit-driven wear. Third, re-analyzing classic fossil collections with this new framework may downgrade some long-standing claims while highlighting a smaller number of more robust cases.
For now, the message is clear: teeth are remarkable recorders of life history, but they are also subject to powerful, non-behavioral forces that can mimic the traces of deliberate action. By demonstrating that ordinary chewing and environmental grit can sculpt grooves once heralded as evidence of early dental care, the new primate study urges greater caution in drawing sweeping conclusions from small marks. The fossil record of behavior is still there to be read-but it may be written in a subtler script than the toothpick hypothesis allowed.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
