The fossil record continues to grow. And with it, so does the contradiction at the heart of the study of human evolution. Each year brings new hominin fossils—jaws, skulls, infant mandibles—that confuse our picture of early human evolution. Instead of revealing a coherent lineage shaped by adaptive refinement, the evidence points to mounting morphological diversity among hominins that shared time, space, and ecological strategies. This is a problem for the theory of human evolution, as typically framed, because natural selection should reduce such variation—not preserve or multiply it—especially once a powerful adaptation like bipedality or tool use takes hold.

Two recent studies sharpen the problem. In Nature Communications (June 2025), researchers published an analysis of three infant Homo fossils—two from South Africa and one from Ethiopia—dating to around two million years ago. What is striking is that even at this early stage of life, the mandibles already show distinct morphologies. These are not subtle variations of a common type; they reflect different developmental trajectories, pointing to divergence from birth—not as a result of environmental adaptation later in life. These specimens are assigned to different species—Homo habilis, Homo erectus, perhaps others—but the fact that they coexist in time and space suggests that they belong to overlapping populations that did not displace each other.

Meanwhile, a fossilised jawbone from Swartkrans, previously misclassified, has now been identified as belonging to a new species of Paranthropus—gracile and small-toothed, unlike the robust chewing machines that usually define the genus. This adds a second Paranthropus species in southern Africa and, once again, points to parallel lineages, not a tidy succession.

Taken together with other finds (e.g. Homo naledi, Homo floresiensis, Homo luzonensis), this growing list of coexisting species challenges the assumption that evolution within the Homo lineage followed a competitive, progressive logic. Instead of a single, gradually adapting lineage, we see a proliferation of forms—many of them apparently capable, bipedal, and tool-using—spreading not only across Africa but deep into Eurasia and Southeast Asia.

The Ecological Inconsistency

So how do palaeoanthropologists explain this? Most still turn to the logic of branching speciation. Populations become isolated by geography, adapt to local conditions, and evolve along separate paths. This model works reasonably well in ecology. But in this case, it begins to fall apart. Why? Because these groups were not isolated. Many shared the same environments. They used similar tools. They walked upright. Yet their differences persisted and deepened. Even as some lineages remained small-brained or retained arboreal traits, they were not eliminated. Diversity increased instead of narrowing.

This presents a deeper inconsistency. If hominins were already occupying a niche that rewarded bipedality and tool manipulation, then the proliferation of multiple distinct forms within that niche becomes biologically puzzling. In many other genera, physiological and behavioural competence tends to reduce diversity: the best-adapted form outcompetes the rest (see Hardin’s Competitive Exclusion Principle). This is arguably what happened among the Pan lineage, which may have been displaced by Homo across much of its former range. By contrast, the great apes that survive today—Gorilla, Pan, Pongo—are largely allopatric and maintain well-differentiated morphologies suited to distinct ecological roles. They do not blur into one another the way hominins do.

Moreover, many hominin species display a curious mosaic of features—a blend of arboreal adaptations (curved fingers, climbing shoulders) with terrestrial traits (bipedal legs, tool-use morphology)—even when they share the same habitat. If ecological adaptation were the dominant force, we’d expect functional streamlining over time, not persistent mosaicism.

Hominins, however, appear to have proliferated side by side, much like different human ethnicities do today, without consolidating into a single dominant form. We find Paranthropus and Homo sharing space, and multiple Homo forms persisting in parallel. That such morphologically capable beings coexisted rather than converged suggests that we are not looking at different species adapting to different ecological roles. Something else is preserving the variation.

Even leading paleoanthropologists have begun to acknowledge this puzzle. Chris Stringer, once a proponent of the linear ‘Out of Africa’ model, now speaks of a reticulated pattern—a ‘web of populations’—connected by gene flow yet morphologically divergent. While he does not discard the evolutionary framework, his model no longer treats speciation as a clear-cut outcome of ecological pressures. Instead, he concedes that what we are observing may not be traditional species, but something more fluid.

Domestication as an Alternative Explanation

The continued presence of multiple ‘species’ within the same environments suggests that we may not be looking at separate species in the classical sense, but rather breeds or varieties within a socially differentiated population. What we call Homo habilis or Paranthropus robustus may reflect lineages maintained through reproductive control, not ecological separation. The variation is not eliminated because social cohesion is maintained culturally—not by geography, but by group identity, possibly grounded in maternal lineage and early caregiving patterns.

I do not base this on direct evidence from early hominins, but on the way reproductive choices are culturally regulated in Homo sapiens today, and how this pattern produces morphological and behavioural difference today. If this pattern has deep evolutionary roots, then interbreeding among early hominins, while biologically possible, may have been shaped by similar social mechanisms. In this light, distinct lineages could have been created and persisted like specialised varieties in a shared system—diverse, yet maintained by social rather than ecological pressures.

Chimpanzees and bonobos both show strong maternal bonds. In bonobos, mothers remain closely attached to their sons and even influence their sons’ mating success. In humans, these relationships go further: grandmothers and others begin to play roles in managing reproduction. This can only happen when offspring remain in the group, when dispersal is suppressed. And to suppress dispersal, you need attachment (love?).

Attachment between mother and adult child is not a biological necessity. It is a social innovation. The original dependency is turned into an asset—not just for survival, but for recognition, bonding, and group continuity. Over time, this asset becomes a powerful selective force. In such a system, children are not merely tolerated; they are encouraged to remain children. This social pressure selects for individuals who remain juvenile longer—both in form and in behaviour. This helps explain why neotenous traits appear in the fossil findings as stable, heritable features. They are hallmarks of a system where care, recognition, and inclusion become selective forces in their own right.

This is why we must take seriously the symbolic potential of early hominin behaviour. When Homo naledi buries its dead deep in a cave, without tools or fire, we are not seeing instinctual hygiene or shelter behaviour. We are seeing the extension of attachment. The dead body resembles the dependent infant: motionless, helpless, inert. If mothers carried pebbles that resembled babies, as the Makapansgat cobble suggests, then the face of the infant was already being seen as meaningful. And when the infant dies, the bond does not vanish. It finds form—in burial, in ritual, in memory.

Conclusion

All of this—bipedality, neoteny, mosaic traits, suppressed dispersal, symbolic behaviour—forms a coherent trajectory. It is not a trajectory well explained by ecological adaptation. It is explained by active self-domestication—not in the narrow sense of reduced aggression, but as the domestication of the child by the parent, and of the group by itself, through patterns of care, recognition, and exclusion. In this light, what we call speciation may simply be social sorting over time. Morphological variation is preserved and reinforced because it matters to the group, not because it increases survival per se. That is why the fossil record keeps multiplying—not because nature is confused, but because we are still reading social processes as biological divergence.