For a long time it was believed that the concept of evolution was linked to adaptive characters. But other completely random phenomena have also statistically determined the history of what we are
Usually, people automatically associate natural evolution with adaptive trait selection. In reality, the concept of evolution, as its etymological root suggests (evolatio, i.e. the act of unrolling a papyrus) is simply connected to the change of a certain population, and more precisely to the change in the percentages with which the different types of individual are represented in that population. One of the mechanisms that produce this change, natural selection, is adaptive: changes in an organism's characteristics make it more or less suited to its environment, and the number of offspring it leaves varies accordingly. However, some changes in a population (even in an entire species) are not the result of selection, and can even be completely random.
Among the most important mechanisms that can produce non-adaptive evolution, i.e. change in the absence of selection, are genetic drift and gene flow. The first of these phenomena predicts that, where a certain trait is neutral, the corresponding genes (with neither a positive nor a negative effect from an adaptive point of view) will increase or decrease in each generation with equal probability. Given long enough, however, the mathematics of probability show that those genes will either disappear or be present in 100% of the population, after which, of course, the trait specified by those neutral genes will be stable in the population. In the absence of selection, that is, the purely stochastic effects of genetic drift therefore eliminate variability from a population over time, until all members of a species possess that trait, and, a posteriori, it will be possible to measure its evolution with the random affirmation of a certain trait. The second of the phenomena mentioned, gene flow, is the flow of genetic variants into or out of a population. This flow may be due to the migration of different single organisms that immigrate and reproduce in new populations, to hybridization between different species and to other phenomena.
A limited gene flow promotes the divergence of a population due to the "loss of reproductive contact" with the rest of the species, eventually leading to the birth of a new species, even if the population in question is not subjected to natural selection, due to the simple accumulation of random variations. There are still further processes of non-adaptive evolution (for example, mediated by the integration of parasitic genomes), but I believe that the first two examples are enough to understand a very important concept: evolution does not stop at all in the absence of natural selection, contrary to what many may think, and even when a given trait asserts itself more and more in a population, to the point of even changing the morphology of a species, this is by no means an indication that this trait is adaptive, but could be a pure product of chance (always, of course, that selection does not can act, i.e. until this trait has no effect on reproductive success). All right, one could concede, but these are still marginal processes. In the vast majority of cases, a population evolves under the pressure of some factor of positive or negative selection. This idea is much less founded than it may appear.
Consider, for example, the morphology of the skull of our species. Conducting an analysis of fossil skulls of Homo dating from 2.8 million years ago to tens of thousands of years ago, it appears that some characteristics of the skulls showed a strong adaptive signal, such as the shape of the jaw. This means, in accordance with the more widely held idea about how natural evolution worked, that the shape of the jaw of early Homo probably changed through natural selection, adapting to a changing diet in the various forms of Homo. However, when the shape of the braincase in our genus is examined instead, the effect of a non-adaptive process is noted, namely genetic drift. Not only that: in fact, the results show that random processes can explain the diversification of Homo species for most of the osteological traits examined, including neurocranial diversification, and in all time periods. These results show that genetic drift and possibly small population size were important factors shaping the evolution of our genus and many of its new traits, while at the same time some specific traits, related for example to diet, were sculpted by natural selection.
Staying with the morphology of the skull, other studies have shown the effect of genetic flow as a random determinant of the evolution of individuals in a population: in hybrids between dog, coyote and wolf, disparate cranial morphologies are affirmed, stably transmissible over the generations, which obviously are not the product of an adaptation, but of hybridization between different species. This diversification can persist up to the effect of selective factors, leading to the random evolution of populations; and, also in this case, the phenomenon is much more extensive and widespread than one might imagine. In short: many, very many traits can vary in a population in a well-defined and unpredictable temporal direction, and, at least in complex organisms, i.e. capable of expressing a multitude of redundant traits, the fraction of truly adaptive traits, subjected to the scrutiny of natural selection so as to cause Darwinian evolution, can be much smaller than what is commonly understood, can vary over time (a trait can "free itself" from the selective pressure due to an environmental change) and in any case may not be immediately obtainable from the simple variation of the morphology of individuals, in the absence of precise genetic information over a long period of time. Like viruses and perhaps even more so, our evolution is also the result of chance, and not only because selection acts in a random direction, but because other phenomena have statistically determined the history of who we are.