There are many ways of determining dates for different elements in the origin of speech, but they all rely on circumstantial evidence. This blog has leaned toward a very old date (2 or 3 million years) for the first steps with a long evolutionary tail leading to full speech (sentences, stories, and myths), but there is an alternate interpretation that has a recent date (75 thousand years ago, plus or minus 25 thousand years). In which full speech pops up very quickly. A couple of new papers from Stony Brook's anthropologist/archaeologist John Shea provide new arguments that support the older dates.
A second late date theory is more archaeological. It refers to the rise of "modern behavior." For language issues, modern behavior means the use of symbols—body decorations, jewelry, artistry, etc. More generally, modern behavior refers to tools that are lighter, smaller, and more heavily modified than older tools, reflecting a more sophisticated tradition. In the 1980s a new term, "the human revolution" entered the story. By its very name, it says something happened around 75 thousand years ago that produced modern humans.
These recent dates are all remarkable because Homo sapiens is normally dated at around 200 thousand years ago. In biology, a species is usually a species, but this appeal to a revolution or leap forward divides the species into modern and archaic forms. The difference between the two periods is behavioral, but if it is to be meaningful there also had to be a biological difference.
The Chomskyan approach is straightforward: there was a biological mutation that changed the way our brain works and ultimately gave us the ability to speak. The human revolution theory is more vague but seems to say that there is an important biological difference between archaic and modern forms of Homo sapiens.
John Shae's paper in the latest issue of Current Anthropology is titled, "Homo sapiens Is as Homo sapiens Was," and stands as a direct counter to the proposition that modern people are biologically different from archaic people.
Complicating the issue are the greatly different cultural lineages scattered throughout the world. The difference between today's societies based on information technology and the societies of people living in the forest of the upper reaches of the Amazon are much more dramatic than the differences in fossil and archaeological traces left 200 and 75 thousand years ago. Yet we do not sort modern races biologically.
We know from many examples that if, say, an Amazon infant is taken at birth and raised in San Francisco it will become a normal participant in the modern world, and if the experiment is reversed a San Franciscan baby can grow to become a normal participant in the Amazonian world. This Amazonian experiment argues that the relevant difference between the groups have to do with the cultural lineages and are not biologically significant.
The Chomskyan great leap forward idea denies that the Amazonian experiment would work if we hopped in a time machine and switched babies with an archaic Homo sapiens. The modern infant should grow up to be an unusually advanced member of its group while the archaic infant will grow to be sub-normal, developmentally disabled, and incapable of participating in contemporary society. Human revolution thinkers are more vague but lean toward the Chomskyan expectation than to Shea's position.
Shea seems to believe that a time-travelling archaic Homo infant would do just fine if raised in the modern world. He bases his reasoning not on symbols but on the tools. His position challenges "the assumption that there were significant differences in the capacity for behavioral variability between the oldest known H. sapiens and recent humans" [p. 10].
The focus on variability is a clever one because it directly addresses the issue of cumulative culture. Apes have some variability in behavior, as proven by the fact that all chimpanzee troops are not alike. But chimpanzee groups do not vary by much, certainly nothing to compare with the kind of contrast considered in the Amazonian experiment. The challenge, of course, is to compare the variability of the earliest sapiens with later ones.
Shea's method is to focus on stone age tools. He assembles a list of ten sites in East Africa, dating from the West Baringo, Kapthurin Formation in central Kenya (234 to 284 thousand years ago) to Lowasera, in northern Kenya (less than 7 thousand years old). He sorts the tools from this time span into five "modes" and notes that in almost every site we find four of these modes. The oldest site has four modes and so does the most recent. Three of the modes are identical at both sites
In other words there is no increase in variation of stone tools over time. There are four modes of stone tools 250 thousand years ago and four modes six thousand years ago. There has been no revolution or great leap forward during the whole of this 200+ thousand year range.
Stone tools do not an entire culture make, and critics of Shea's approach argue that behavioral "variability" is not much more precise or measurable than the behavioral "modernity" that Shea criticizes. Shea's main defense of stones as reference points is that they are durable and easy to find. Like pottery, long after all other elements of a people's culture has gone into the wind, the stones remain. If we did the Amazonian experiment with the peoples of these Kenyan sites, is there any reason to believe that the 250 thousand year old infant wouldn't fit right in with the 6 thousand year old group?
There is no archaeological evidence to say the experiment would fail, and there is evidence to say that the babies would be normal. This argument is not enough to settle the matter in a criminal case that requires proof beyond a reasonable doubt, but it would win in civil trials that require a preponderance of the evidence. And on this blog I try to build on what we've got.
Shea has a second paper that makes the same point in a more popular form for the American Scientist: "Refuting a Myth About Human Origins." In it he neatly explains why he limited the study to East African sites: "Restricting this comparison to Africa minimizes the complicating effects of geographic constraints on stone-tool technology" . The critical finding is nicely summarized: "One does not see a steady accumulation of novel core technologies since our first species appeared or anything like a 'revolution.' Instead one sees a persistent pattern of wide technological variability" .
The usual argument against the human revolution is that many of the things that were supposed to have resulted from the revolution—language, art, jewelry, funeral rites—long predate the proposed revolution. Shea has flipped that line to say that long after the supposed revolution stone age, people were behaving just as they had long before it.
If I can jump in with my own remarks, I often feel that the biggest confusion in most discussions of human and language evolution concerns the difference between humans and animals, and humans and humans. Traditionally there has been a great contempt for animals. Attributing emotions, intentions, and tastes to animals, let alone reasoning, has been routinely dismissed as anthropomorphism. Then experiments showed that animals do have all those things; so people argued that there was no critical difference between animals and humans and any opinion to the contrary was mere speciesism.
The human revolution argument has long struck me as the inevitable mess that comes from not believing in qualitative difference between animals and humans, but needing to explain how it is that humans have occupied so many different niches around the world and live in ways that are so different from the ways animals live. We need a better way of understanding the difference between groups.
Perhaps the Amazonian experiment can help. The experiment has been done on in which chimpanzees have been raised from birth as humans. They don't become normal humans. Does anybody believe that if they were raised from birth by a Homo sapiens group of 200 thousand years ago that the chimpanzee's behavior would have been normal among archaic humans either?
We know for a fact that by 250 thousand years ago the human lineage was already cooperative, big brained, and using tools to make tools. I have my opinions about their speech level too, but I'm leaving that out. Even so, we know that chimpanzees would not have been able to become behaviorally like the very first Homo sapiens.
How far back do we have to go before the Amazonian experiment would result in a chimpanzee that fits in with the human lineage? It must be more than 2.7 million years ago, because Homo habilis of that time was already cooperating in a community and using tools to make tools. It was also somewhat bigger brained than a chimpanzee.
How about performing a variation on the Amazonian experiment? How far into the future could we take a Homo habilis infant, raise it within the human lineage, and have it become a normal member of that population? I'll spare you my thoughts on this and stick with the larger point: humans became qualitatively distinct from their animal ancestors millions of years ago.
Since that break, there have been a number of evolutionary changes that made humans were different from their human ancestors. Humans are the result of a long evolution in which they competed with other members of the human lineage. It was not just one big change that made us different from animals. We have a long, distinctive history built on many changes that have been going on for millions of years. Evolutionary changes are much more thorough than revolutionary ones.