Languages In the World (48 page)

Charles Darwin's
The Origin of Species
got the evolutionary ball rolling in 1859 (Darwin [1859] 1968). However, it was not until the midtwentieth century when the study of evolution was allied to the genetics developed in the wake of Darwin's contemporary, Gregor Mendel, that the scientific discipline came fully into existence. It is called the neo-Darwinian synthesis. Since its inception, the discipline has been very genecentric, perhaps appropriately so given the complexities of the genetic structure of any population a researcher chooses to investigate. In recent decades, however, scientists interested in genetics and evolution have expanded their search for explanations of how certain organisms are – and have come to be – the way they are, and so now in addition to studying genes, scientists also consider epigenetic networks, behavior, and, in the case of humans, symbolic systems. In other words, there is a second neo-Darwinian synthesis afoot.

Speaking of genetics, it is important to note that the term
genetic code
is a misnomer. In the early days of the neo-Darwinian synthesis, the term was borrowed from information theory. It was used to describe the relationship between the genotype and the phenotype, and the term suggested that all information necessary to produce an organism was ‘in' the genes, thus making the phenotype a ‘read out' of that coded information. In the first step of the transition from genotype to phenotype, the idea of a code is apt, in that a gene is transcribed to RNA and then used in the production of a protein. After this genetic moment, however, all the rest is
epigenesis
, that is, development, and the term
code
no longer applies. It is sometimes the case that the presence of one particular gene links directly to a phenotypic result. For instance, the presence of the gene for the neurodegenerative disorder Huntington's disease means that the person carrying it has 100% chance of acquiring the disease. Mostly, however, phenotypic results are caused by networks of genes, and individual genes may participate in a wide array of phenotypic results.

Evolutionary biologists used to speak of evolution in terms of changes in gene frequencies, as if all of evolution was driven by genetic change. More recently, evolutionary biologists have taken a renewed interest in the encounter of the organism and the environment, which involves a reappreciation of the importance of behavior. Now, evolutionary biologists are more likely to talk in terms of organisms inheriting both genes and environments, and to think of evolution in terms of developmental systems with longitudinal stability. One of the stabilities to have entered the developmental system of the human way of living is the presence of language. Just as when inorganic matter was transformed into organic matter, and a new sphere came into existence, namely the biosphere, so the creation of language has brought a new sphere into the world. Some call it the
noosphere
, from the Greek word
noûs
, which corresponds to the Latin word
intellectus
. We humans now mediate a great deal of our social world, our cognitive world, and our environment through language. We talk about things, we talk through things, and we try to come to understandings of things.

Sometimes, we get our understandings wrong. In Chapter 4, we eliminated the term
race
from our account of the languages of the world because it was incoherent in practice. Now, we explain how the term is incoherent biologically. In evolutionary discourse, two kinds of traits are talked about: connected traits and mosaic traits. A connected trait is a trait you cannot change without affecting a great deal else in the organism. Having two lungs is a connected trait, because lung number is influenced by the genes and by the developmental system giving rise to a bilateral organism. It is also a hugely old trait in the animal kingdom. A mosaic trait is one that can evolve independently of any other trait with little or no effect on the rest of the organism. One such mosaic trait is skin color, and there is evidence, based on changes in one particular gene, that Europeans lightened up rather recently, say, 12 to 6 kya. In fact, it turns out that over time, the same (or similar enough) population can go from light to dark to light again, depending on the circumstances.

Mosaic traits are often what are called
anthropometric traits
. These include skin color, nose length, eye shape, and hair type, which are under strong selection by climate. Lighter skin is necessary at higher latitudes where there is less sun, and people living in those latitudes need to absorb more sun in order to produce Vitamin D. These higher latitudes are also colder, and a longer nose allows the air to warm up a bit before reaching the lungs. The epicanthic fold at the eye also provides an extra bit of protection from the cold. Straight hair keeps warmth on the head, while kinky hair keeps sweat off the head. These traits vary with latitude. By way of contrast, as population geneticist Luigi Cavalli-Sforza notes, “genes [rather than anthropometric traits] are considerably more useful as markers of human evolutionary history, especially migration. They vary more with longitude” (2000:65). For the purposes of our discussion about race, the genetic shuffle, which gives rise to these anthropometric traits, simply does not line up into categories with characteristics useful to biologists. Races are gerrymandered results of variable attitudes attached to visible characteristics.

Modern genetics is fascinating, and learning about genetics is like blowing open a door on what makes us human. The effort of the preceding section was not to downgrade
genetics, but only to assign it its proper place in the second neo-Darwinian synthesis. In order to tell our story, we need all the genetic information we can get.

We start by identifying a woman named mitochondrial Eve who lived about 190 kya (with a probability interval of 300 to 150 kya) and who is the mother of our species. The name is provocative, and we hasten to add that she was not the only woman alive at the time. Rather, she is the woman from whom all current human mitochondria descend. Her lineage is the one to have made it through whatever bottlenecks early modern humans passed through before leaving Africa. A
bottleneck
occurs when a population's size is reduced for at least one generation, which also thereby reduces genetic variation.
⁶
Our hearty Eve is complemented by Y-chromosomal Adam and, no, it is not likely they knew one another. The age of Y-chromosomal Adam is difficult to pinpoint, and a suggested range is that of mitochondrial Eve.

One of the most important genetic changes to be introduced into the human genome involves the
FOXP2
gene found in a region of 50–100 genes on chromosome 7. In 2001, a group of geneticists at Oxford University and the Institute of Child Health in London showed that mutations in this gene cause a wide range of language disabilities based on a family known as KE, a number of whose member exhibit severe disruptions to grammar and to speech production. That is, they have trouble controlling the fine muscular movements in the lower half of their face.
FOXP2
is found in all mammals for which a complete genome has been sequenced. In mice, it seems to serve some role in vocalization.

A team of researchers at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany sequenced the
FOXP2
genes of a chimpanzee, gorilla, orangutan, rhesus macaque, and mouse, and then compared them with the human sequence. They discovered that there have been only three changes in the gene's protein sequence of 715 amino acids in the past 70 million years, that is, since the last common ancestor of humans and mice. Two of these changes have occurred in the last six million years, namely after the human and chimp lineages split. The team estimated that the
FOXP2
gene became what we might call fixed in the human population – that is, all humans could be determined to harbor the last amino acid substitution – between 200 and 120 kya.

Note that this gene is not
for
speech production or grammar. The
FOXP2
gene codes for a type of regulatory protein, known as a transcription factor, which is involved in modulating the expression of other genes, that is, it participates in whether other genes are turned on or off. The protein produced by
FOXP2
belongs to a subclass of transcription factors known as forkhead proteins, and many members of the forkhead family are known to be key regulators of embryogenesis. Of note is the fact that
FOXP2
seems to be important in regulating key pathways in the developing lung, heart, and gut, such that to call it a language gene makes no more sense than calling it a lung gene. Many forkheads are, furthermore, critical for the normal patterning of the central nervous system, and there is evidence to suggest that
FOXP2
has effects in subcortical structures that participate in the human reiterative ability in domains as different as syntax and dancing. Clearly, bodily changes are allied to genetic changes. As we have just seen, one of these changes in humans, minute in size, has had large consequences. The amino acid substitution of
FOXP2
has helped to make human language a connected trait from the soles of our feet to the tops of our heads.

The first split in the human lineage is between Africans and non-Africans who are descended from the East Africans who spread out over the world and inhabited all continents by 40 kya. When Cavalli-Sforza sampled 2000 populations worldwide, which clustered into 42 groups on the basis of 110 genes, he found that the six African populations he sampled differed more from one another and then the rest of the world than any of the other 36, with the people the most genetically removed being the people of Central Africa represented by the Aka, Efé, and Mbuti. Of interest is the fact that every population is made of genotypes originating in more than one continent, and that includes the Berbers, one seventh of the African population, whose genome has a Eurasian character. Cavalli-Sforza speculates that either today's Berbers are descendants of a group with a branch who settled in Europe or yesterday's Berbers were a mix of North Africans, Europeans, and people from the Middle East. He notes, “The two hypotheses are not mutually exclusive and might both be true” (2000:88).

Because humans have been in Africa for such a long time, it has been difficult to determine the language stocks with their relatively shallow time-depth of 10,000 years from the macrofamilies with their far greater time-depths.
⁷
In the 1950s and 1960s, historical linguist Joseph Greenberg made a four-way classification for African languages: Afro-Asiatic, Niger–Congo, Nilo-Saharan, and Khoisan. He made these classifications using differing criteria for classification and knowing they had differing time-depths, and he put all languages with
click consonants
into Khoisan despite the fact that the languages were very different in other respects. Please see the Language Profile for !Xóõ, this chapter, for a discussion of clicks.

Ethnogroups speaking languages classed in the putative Khoisan stock/ phylum tend to be relatively small populations of hunter-gatherers, with the groups quite isolated one from the others and surrounded by languages from different stocks. Many of the original hunter-gatherers in Tanzania, for instance, have been absorbed or displaced over the last 4000 years by successive migrations into the territory of herders and cultivators speaking Afro-Asiatic languages (Cushitic family), pastoralists speaking Nilo-Saharan languages (Nilotic family), and agriculturalists speaking Niger–Congo languages (Bantu family). In order to get a sense of the potential relatedness of the click languages, Tishkoff et al. (2007:2191) compared mtDNA and Y-chromosomal variation between click-speaking populations found in Tanzania and those found in Namibia and Botswana. They discovered that these populations diverged on the order of 55–35 kya, and they speculated that among the factors contributing to their isolation could be the long dry period in southern Africa at the height of the last glaciation about 24–17 kya. They make the assumption that click phonemes arose only once and speculate that: “click phonemes arose on the order of tens of thousands of years ago in sub-Saharan Africa” (Tishkoff et al. 2007:2193). The designation of Khoisan as a phylum seems reasonable.

Although Greenberg's classifications have become traditional in many accounts, we update the picture for present-day Africa through Nichols's linguistic geographic work. In addition to Afro-Asiatic, which is the only proven stock, she proposes the following areas: (i) Macro-Sudan (centered on Greenberg's Niger–Congo), whose languages have features such as vowel harmony, complex tone systems, and SOV order; (ii) Kalahari Basin (based on Greenberg's Khoisan) whose languages have clicks and head-marking; (iii) Chad-Ethiopia which is disproportionately Afro-Asiatic; (iv) the
Berber spread zone in the Sahara, now undergoing shift in the face of Arabic; and (v) the Bantu spread zone now engulfing the Kalahari Basin (Nichols 2010:363–364).

Once we leave Africa, we can trace the trail of the Y-chromosome through the work of Rootsi et al. (2007) in an article whose title tells it all: “A counter-clockwise northern route of the Y-chromosome haplogroup N from Southeast Asia towards Europe.” The NO haplogroup has two branches: N and O. The O branch includes the vast majority of men in East and Southeast Asia, as well as those in Oceania, and it looks to be about 30 kya old. The N branch is newer and postulated to begin a spread northward and then westward perhaps around 17–12 kya. Given that this branch is currently present in Siberia but absent in Native Americans, the N branch must have split after eastern Paleosiberians moved into the Americas. However, it could also be the case that the genes of the Paleosiberian men with the N-innovation did not make it into the Native American pool by one of two ways. Either these men did not produce offspring who survived long enough to reproduce, in which case the lack of the N-innovation in the Americas is a result of a lack of genetic fitness; or these men all happened to die by accident, say, by falling off ice floes in Berengia, the vast ice age land bridge, as they made their way to the New World. In the case of accident, the lack of the N-innovation in the American population would be called
drift
. In either case, the N-innovation did not make it into the Americas.

However, the Q-group did, and this group is widespread in the oldest populations to inhabit North and South America. When a small group inhabits a new space and then multiplies over time, the genetic profile of the larger, downstream population reflects a founder's effect of the original settlers. Another way of describing the founder's effect is by saying that those lucky few who are first at bat significantly determine how the rest of the game is played. As for the N-innovation-carrying men who produced offspring who produced offspring,
⁸
their haplogroup is found most at high latitudes, and it spans the Far East to Eastern Europe. The biological evidence is thus consistent with the linguistic evidence for language families and stocks, their homelands, and their westward spreads across the Eurasian steppe outlined in Chapters 7 and 8.

At the same time, there is no equal mtDNA story. The mtDNA haplogroups characteristic to Southeast Asian populations occur in, for instance, Baltic-speaking countries with a total frequency of less than 1%. These contrasting stories give us a picture of a world with men on the move and women largely stationary. However, when it comes to the genes, Cavalli-Sforza (2000:82) points out that Y chromosome mutations are highly clustered geographically, which suggests that men move very little genetically. For genetic mobility, what counts is where people settle for marriage, and in most traditional societies around the world, it is the women who change residences more often than men, making women more genetically mobile. Of related interest, in the click languages study cited above, male–female migration patterns were complementary. There was higher female migration from hunter-gatherer groups into nearby pastoralist or agricultural groups, while there was higher male migration from pastoralist or agricultural groups into hunter-gatherer groups (Tishkoff et al. 2007:2192).

A recent archeologic find gives a fascinating new glimpse into the picture in the northern part of Eurasia before the start of the counterclockwise northern and western tour of the N haplogroup. The genome of a Siberian boy who lived 24 kya has now been sequenced. His Y chromosome belongs to haplogroup R, and his mtDNA
belongs to haplogroup U. These haplogroups are found exclusively in people living in Europe and regions of Asia west of the Altai mountains, and so it is not terribly surprising to think he belonged to a lineage of people who either later moved west or came from the west. What is a surprise is that a portion of the boy's genome is shared only by today's Native Americans and no other groups. A further surprise is that the boy's genome shows no connection to modern East Asians. Because DNA studies strongly suggest that Native Americans are related to modern East Asians – perhaps Siberians, Chinese, or Japanese – it is a puzzle to think the boy could be related to Native Americans and not also to East Asians. However, there is no doubt that the boy's genome represents ancient Native American roots. It has also cleared up a previous puzzle: why the 9 kya Kennewick man (Columbia River, Kennewick, Washington) had some European-like features. Until this boy's bones were investigated, Kennewick man was considered to be an anomaly, and the common idea was that any of the Eurasian features found in Native American genes were a result of postcolonial mixing. Now, it seems that Native Americans have very deep genetic roots with Europeans (Balter 2013a, 2013b).
⁹

We take a final look at the kind of light DNA can shed on linguistics in the intriguing hypothesis advanced by Dediu and Ladd (2007). This hypothesis involves a possible relationship between the distribution of tone languages in the world, namely those in sub-Saharan Africa as well as Southeast and East Asia, and the presence or absence of certain alleles on two brain-development genes,
ASPM
and
Microcephalin
(
MCPH
). Both
ASPM
and
MCPH
help determine the size of the brain and skull by being involved in cell-cycle regulation, such that if there are deleterious mutations to these genes, not enough cells will divide in the right way, and the brain and skull will not grow to proper proportions. Both of these genes have a pair of favorable mutations that are called
derived
(
D
), such that now, in addition to
ASPM
and
MCPH
, there are the variants
ASPM-D
and
MCPH-D
. In other words, both genes now have two haplogroups, and their ages are estimated at 5.8 kya (14.1–0.5 kya) and 37 kya (60–14 kya), respectively.

Both haplogroups show signs for positive selection and geographic distribution. Taking a geographic survey of the world,
ASPM-D
has high frequencies in Central and Western Asia, Europe, and North Africa, and very low frequencies in Asia, Europe, and the Americas. It has moderate frequency in North and East Africa, and Southeast Asia. It is very rare in Central, Western, and South sub-Saharan Africa. For its part,
MCPH-D
, being older, coincides with the introduction of anatomically modern humans into Europe about 40 kya (see
Map 10.1
), as well as the shift in the archeological record indicative of modern human behavior, such as art and the use of symbolism. This variant is widely distributed across the world, but not very frequent in Africa (Evans et al. 2005). Returning to
ASPM-D
, it has shown strong positive selection in primates leading to
Homo sapiens
, especially in the past six million years in which
ASPM
acquired one advantageous amino acid change every 350,00 years. Thus, its role in human-brain evolution seems clear. The recent appearance of the new allele
ASPM-D
suggests that it is continuing to evolve in modern humans (Mekel-Bobrov et al. 2005).