The relative contributions of biological versus cultural aspects of language cognition depend on its evolutionary antiquity. If languages have a shallow prehistory (less than one hundred thousand years), we can expect little correlated biological restructuring of cognition as a result, except insofar as required to get this capacity off the ground. In this case, most of its influence will be traced through cultural processes. If languages have a deep prehistory (on the order of a million years), however, then we can expect that human cognitive and emotional systems have been substantially shaped by its ubiquitous presence in all aspects of human social life. This also should correlate with the extent to which human ethical and spiritual sentiments have become a part of human nature, as opposed to mere cultural overlays on ape nature.

Assessing the origins of these abilities is complicated by the fact that no direct consequences of language use are preserved in the fossil record. Paleolithic archeological evidence for symbolic expression that may signal well-developed linguistic and spiritual activities is well known from European cave paintings and carvings and Australian rock paintings, and from evidence of intentional burials (possibly including Neanderthal burials, as well as the burials of anatomically modern humans).

Though the creation of icons and burial of the dead are not guarantees of shamanistic or religious-like activities, they do suggest the existence of sophisticated symbolic reasoning, and this is a crucial correlation. The first sculpted and pictorial forms can be dated to no earlier than about sixty thousand years ago, and the most well known date to within thirty thousand years ago. This is quite recent, considering that hominids have been on a separate evolutionary track from other African apes for at least five million years, that members of species similar enough to be included in the genus Homo have been around for 1.8 million years, and that the human species Homo sapiens is at least two hundred thousand years old. In general, these earliest samples of expressive symbolism must be understood not as evidence for the initial evolution of symbolic abilities but rather for their first expression in durable media. They likely had long been incorporated into conventionalized social activities by that time. The origins of the symbolic traditions that these works express in material form could easily anticipate this data by an order of magnitude.

To get some idea of the possible extremes of this range of possible dates consider the following. The earliest direct archeological evidence of language is, of course, in the form of early forms of writing, which are all less than ten thousand years old, and most considerably more recent (about five thousand years ago). Since not even the most radical theorists among archeologists and paleontologists would date the appearance of modern languages more recently than about fifty thousand years ago, this late externalization of language offers a curious challenge: Why did it take so long for this most important means of communication to exhibit direct external expression? The same question can be asked of the first evidence of pictorial and carved forms, which date back about sixty thousand years in Europe and Australia and possibly earlier in Africa (though this African evidence is currently less well known). Assuming some comparable difficulties in externalizing these different modes of symbolic expression, we might suggest that, most conservatively, the corresponding distinctively human symbolic communication must be at least ten times as old; that is, 5,000 to 50,000 years for modern language, and 50,000 to 500,000 years for some form of language.

At the other end of the spectrum, there is a series of apparently linked paleontological transitions evident between 1.6 and 2.4 million years ago in Africa that suggest that the beginnings of symbolic communications in some form may date to this fossil epoch. The first clear evidence for the regular production of stone choppers, at a site called Gona, can be dated to about 2.4 million years ago. These are associated with fossil species of the genus Australopithecus (possibly A. garhi). Australopithecines exhibited ape-sized brains, relatively large jaws with heavy dentition (evidence of a vegetarian dietary adaptation), relatively modern bipedal locomotion, and also a characteristic sexual dimorphism (males much larger on average than females), which is indicative of male competition over females in a polygynous mating system that is fairly typical of monkeys and great apes. By 1.8 million years ago a number of fossil sites begin to demonstrate hominid species with larger brains and reduced dentition, correlated with extensive stone tool assemblages. These features have prompted paleontologists to cite this as the point where our genus, Homo, begins. By 1.6 million years ago members of our genus, with brains beginning to cross into the low end of the modern range, had left Africa to spread into Asia, Southeast Asia, and possibly throughout more temperate Asian regions as well, taking with them more sophisticated tools.

Given these unprecedented features, there can be little doubt that some significant changes in communication and cognition also are contemporaneous with these transitions—the first forms of crude symbolic communication—though it is likely that the evolution of modern forms of linguistic communication took much longer to develop.

If symbolic communication has been around in some form for as much as two million years then we can expect it to have had significant consequences not just for human culture but also for human brain function. The evolutionary biological effect of a behavioral adaptation such as this may be usefully compared to that of dam building in North American beavers. The evolution of this ability has changed the niche in which beavers mature and live, and this has changed the natural selection forces affecting beaver physiology and behavioral propensities in succeeding generations. Thus, beavers exhibit extensive aquatic adaptations as a feed-forward result of beaver behaviors. This evolutionary process has been called niche construction.

The effects of human symbolic communication and culture can also be understood as a form of niche construction, though symbolic culture is in many ways a far more all-encompassing niche than a beaver pond. This niche likely favored the evolution of certain cognitive capacities and social predispositions relevant to symbolic learning and communication, but also, as in the case of beavers, there may be many special features of this artificial niche that are idiosyncratic to it. Thus, there is good reason to expect that human brains have been reorganized in response to language, a reorganization that included changes affecting emotional, social, and communicative tendencies, as well as mnemonic, attentional, and motor capacities supportive of symbolic communication.

Anatomical hints of this effect are evident in the changes in regional brain proportions (e.g., disproportionately expanded prefrontal cortex), cortical vocal control (unprecedented among mammals), and lowered laryngeal position. Hints from behavior are even more extensive. These include the convergent contributions of many systems to this capacity, its robustness in the face of variations in learning conditions and the effects of early brain damage, its highly predictable developmental progression, the remarkable universality of many of the structural features of languages, and its unprecedented efficiency. These effects need to be understood also with respect to the complex cultural dynamic of language change, which itself is a kind of quasi-evolutionary process. The ways different languages carve up the meaning and reference “space” and the syntactic systems that organize linguistic expression clearly change and evolve over historical time, and probably with respect to these biological predispositions and abilities as background.