Are There Genes for Language?
Incorporating a critical review of four journal articles: –
1. Savage-Rumbaugh et al 2001, Language, Speech, Tools and Writing – A Cultural Imperative, Journal of Consciousness Studies, 8, No. 5 – 7. 2001, pp. 273-92.
2. Fisher and Pääbo et al 2002, Molecular Evolution of FOXP2, a Gene Involved in Speech and Language. Nature, Vol.418, 22 August 2002.
3. Pääbo et al 2003, Selection on Human Genes as Revealed by Comparisons to Chimpanzee DNA, Genome Research, 13:831-837, Cold Spring Harbor Laboratory Press.
4. Pääbo et al 2007, The Derived FOXP2 Variant of Modern Humans Was Shared with Neanderthals, Current Biology 17, 1908-1912, November 6 2007, Elsevier Ltd.
Introduction
Are there genes for language? In short, Yes. There is a gene for language – FOXP2. (Vargha-Khadem et al 1995, Fisher et al 1998, 2008, Pääbo et al 2003, 2007) Geneticist Simon Fisher states that FOXP2 is not the only language gene, merely the first to be discovered. (Fisher 2008) The isolation of a language gene is an exciting advancement. And the long journey of psycholinguistic investigation, climaxing with the identification of FOXP2, has travelled through more twists and turns than the most torridly tempestuous of tongue twisters…
Summary of Scientific Background
Binocular vision, bipedalism, and opposable thumbs have all contributed to our species’ successful evolutionary leaps. (Diamond 1991) But it is the human ability for speech that sets humans apart from the other animals. (Darwin 1871) All primates are social and when compared to other mammals highly communicative. (Savage-Rumbaugh 1988) Our unique cerebral cortex is proportionally much bigger in primates than in any other order of animal. Unlike other animals the primate brain continues to grow after birth (Dunbar 1996). We seem especially programmed to understand ourselves and others – particularly in the context of verbal communication. (Miller 2000) Research has shown that, universally, babes in arms not only intuitively understand, “Motherese”, but infants actively direct carer’s use of motherese even if it is spoken in another language. (Gopnik, Meltzoff, Kuhl 1999) There is evidence that some areas of linguistics and intelligence have been Sexually Selected. (Miller 2000, Fisher 2008) We are attracted to raconteurs, verbose others who amuse us with their witty, fast-paced banter. This is known as, The Scheherazade Effect. (Miller 2000) Other theorists believe Natural Selection has played a greater role. The talkative female, who could organise herself and her fellow females into cohesive childminding groups had descendents more likely to survive than females who did not. (Dunbar 1996) In 1871 Charles Darwin considered language to be adaptive in humans, stating no child, “…has an instinct to bake, brew or write…” Darwin believed the instinct for speech was seen in, “…the babble of our young children.” It would take another 130 years and endless arguments over Darwin’s “dangerous idea” before The Wellcome Trust’s Sanger Institute for Genetics and The Human Genome Project’s discovery of FOXP2 would finally prove Darwin right.
Goodall and friend share their thoughts
Today, genetics reveals a 1.2% difference in DNA sequences between man and chimpanzee. (Chen and Li 2001) But before the 1960’s and Goodall’s pioneering field work at the Gombe, Tanzania, psychologists had to rely on the fossil record and scant behavioural studies on captive apes to ascertain how close we are to our closest living relatives. The 1930’s saw the birth of ape language studies. Early experiments attempted to see if chimps, when socialised into a human family, would pick up language. (Savage-Rumbaugh 1994) After a few years and fewer husky words it finally it dawned on the scientists that without a larynx positioned deep in the thorax, as a human’s has evolved to be, spoken speech, akin to ours, is physically impossible for chimps. (Marcus and Fisher 2003) (I will return to this specific point later in the essay.) In the 1950’s and 60’s Chomsky was championing human linguistic supremacy. Like Vygotsky and Piaget before him Chomsky drew our attention to children’s innate ability to pick up the rules of language. (Pinker 1994) Chomsky observed children were uniquely good at re-combining words over and over to make new sentences. Chomsky saw syntax as innate in humans and believed without syntax language could not exist. (Savage-Rumbaugh 2001) The cases of feral children have revealed a juvenile has a critical period in which to acquire language. (Pinker 1994) Linguistic stimulation must begin before age six, otherwise the opportunity to develop language and appropriate use of syntax and grammar diminishes. (Pinker 1994) More evidence to support this theory is emerging from the contemporary case of the incarcerated Ffitzl family in Amstetten, Austria. The cellar-bound children, “...appear to have developed a language of their own, consisting of mumbles and grunts…” reported Leopold Etz, the chief inspector of the regional police in Lower Austria. (Guardian Newspaper 28.04.08) A combination of nature and nurture are essential to language acquisition. (Ridley 2004)
Oxana feral child from Ukrain
By the 1960’s primatologists were asking, could ape cognition be wired for language – however primitive? After the failure of the 1930’s ape language studies different methods were required. In 1966, Washoe became the first infant chimp to learn American Sign Language (ASL). (Dunbar 1996) Not long after Washoe’s success with ASL another ASL study was launched with a chimp called Nim Chimpsky, his name intended to poke fun at Noam Chomsky. (Chomsky 1972) Chomsky and more recently Steven Pinker, have both thrown scorn over ape language research claiming the chimps are merely cued circus performers, or the apes have no grammar and therefore do not have language, or, apes do not demonstrate syntactical competence and therefore can not possess the ability for language. (Savage-Rumbaugh 1994, 2001). Over the years primatologists have taken criticisms on the chin, and with their apes in tow have risen to increasing challenges. A troop of simian signers have come and gone over the subsequent decades. Methodology has seen various T.o.M. tests, innovative Game theory experiments and ASL over-taken by the development, by Savage-Rumbaugh, of a lexigram based language for chimps called, “Yerkish”. But it has been Savage-Rumbaugh’s work with bonobo, Kanzi, that has pushed the scientific boundary forwards. (Dunbar 1996) In the 1980’s Savage-Rumbaugh was working on language acquisition with Kanzi’s mother; the baby ape was merely a silent observer. One day the infant jumped up and answered the test Savage-Rumbaugh had set his mother. (Savage-Rumbaugh 1994) It then became evident to Savage-Rumbaugh that just as young humans have a critical period in which to acquire language, so do apes. Now, where possible, all intelligence and language projects with apes begin in the animal’s infancy and where possible the infant ape is kept with its mother. (Savage-Rumbaugh 2001) To help refute accusations of cuing Savage-Rumbaugh has given Kanzi headphones to wear and spoken to him via a radio from another room. Psychological experiments have been employed to successfully exhibit ape intelligence and linguistic capability refuting critics. (Premack and Premack 1983) (Dunbar 1996) (Savage-Rumbaugh 1994, 2001)
Savage-Rumbaugh and Kanzi pose for the camera
Today Kanzi is a loquacious ape, he understands thousands of spoken words and communicates using over 500 lexigrams. (Fields 2008) Kanzi’s voluntary communication consists of short, novel sentences, comprised of grammar. Kanzi understands pronouns such as “it”. If you say to Kanzi “I left it over there, will you get it?” he understands and collects the object you want. He easily comprehends and complies with odd, novel requests and will, for example, put the toy snake in the fridge if you ask him to. (Savage-Rumbaugh 2007) Savage-Rumbaugh states after all these years and so much evidence we should accept apes do possess language ability and move on to now ask, “How does language shape social relationships and culture in apes?” (Savage-Rumbaugh 2007) Goodall has rhetorically asked, “Why does a [wild] male chimpanzee cover his mouth when he involuntarily makes excited food barks – unless he doesn’t want others to hear him?” (Goodall 1986) Once a sceptic of ape T.o.M., and language ability David Premack is now an advocate for ape language research. After years of work with chimpanzee Sarah, Premack now questions our possessive self-obsession with human language and preoccupation with its structure. “A human,” Premack observes, “is not simply a chimpanzee to which language has been added. It is advanced intelligence coupled with a special factor which permits language to emerge.” (Premack 1983)
Back in 1983 that “special factor” remained tantalisingly ephemeral, however, in 2008 might that “special factor” be FOXP2?
FOXP2 – a Gene for Language
In Britain, in the early 1990’s, half the members from three generations of the “KE family” suffered with a developmental disorder resulting in a disruption of speech and language skills. (Vargha-Khadem 1995) Articulation is significantly impaired, particularly to the fine motor control of the larynx and tongue (verbal dyspraxia) and impairment of speech and verbal comprehension (dysphasia). The orofacial (lips, jaw, tongue) control is slowed and when sufferers speak they are unintelligible, even to other family members. Sufferers of this deficit can’t pronounce tongue twisters or multi syllable word such as, “hippopotamus”. (Fisher 2008) In addition there are also deficits in written comprehension, grammar and syntax. A mutated gene, passed from one generation to the next, causes an impediment similar in presentation to Broca’s aphasia. (Fisher 2008) The inheritance of the problem was observed to be straightforward and clear-cut, suggesting one or very few genes were responsible. (Fisher 2008). By studying the DNA of affected and unaffected KE members the specific mutation was located amongst 69 other genes on chromosome 7. The mutation caused a simple alteration of the sequences to just one gene from the FOXP2 pair resulting in devastating consequences. (Fisher 2008) By chance, at the Oxford Radcliffe Hospital, a young boy, known as “CE”, genetically unrelated to the KE family, had a similar deficit. When the child’s chromosome 7 and FOXP2 gene was cross-referenced with the KE family’s FOXP2 it was clear the mutation was the same. It is believed this specific mutation on FOXP2 results in retardation in early brain development. FMRI tests on the affected members of the KE family revealed deformity to the Basal-ganglia, a region of the brain relating to movement. With a genetic impairment of this magnitude existing from birth any therapy administered must be given as early as possible to fully exploit the 0 – 6yr critical period. (Fisher 2008) Fisher has said, “It is unusual that such a simple genetic defect, namely one letter in FOXP2, can cause such a damaging speech disorder – it’s weird it’s so clear cut.”
The extraordinary discovery of FOXP2 begged the question, how has FOXP2 evolved? In 2002 the Sanger Institute collaborated with the Max Plank Institute for Evolutionary Anthropology. The method to examine the evolution of FOXP2 was to isolate DNA sequences and amplify the polymerase chain reaction (PCR), by sequencing the overlapping fragments of FOXP2’s coding region. (PCR enables researchers to produce millions of copies of a specific DNA sequence in just two hours.) (Fisher, Pääbo, et al 2002) It was discovered FOXP2 proteins are identical in chimps, gorillas and rhesus macaques and these three species have two differences from the human FOXP2 protein. Orang-utans, the first ape to branch off the evolutionary line, and Asia’s only great ape, have three differences in FOXP2 proteins from humans, one of those evolving in orang-utans independently since their evolutionary separation. And indeed, the vocal calls and facial movements of the orang-utan are quite unique from other apes, an example being the male orang’s “long call”. (Galdikas 1995)
Two of the differences in human FOXP2 proteins evolved after archaic ape-man separated from archaic chimpanzee, 4.6 - 6.2 Million years ago. (Chen and Li 2001) At least one of these changes has functional implications. (Fisher, Pääbo, et al 2002) There has been a significant, positive selection for amino acid changes in FOXP2 during human evolution. If the two changes to the amino-acid coding found in the human version of FOXP2 were selected for during the last 4 – 6 million years of human evolution there should be traces of a “selective sweep”. A selective sweep happens when a new mutation occurs in a gene that increases the fitness of the carrier relative to others. (Fisher, Pääbo, et al 2002) A distinctive sign of a selective sweep is a higher number of low frequency alleles than is expected to occur via random mating. (Allele – competitive, alternative forms of a gene. Dawkins 1999) Using “Tajama’s statistical model”, it was found the probability of the changes to FOXP2 developing via random mating was 0.002. Therefore the development of FOXP2 was extremely likely to be adaptive. The speculation from the Max Plank Institute and the Sanger Institute, in 2002, was the orofacial movement in humans, endowing us with, amongst other things, fine motor control and co-ordination of our mouth and jaw, thus eliciting fast dialogue and articulation of tongue twisters, was caused by one or both of the amino-acid changes to FOXP2 since we split from apes.
Birute Galdikas and young Orang-utan
If the changes to FOXP2 were indeed adaptive a talkative, expressive archaic human would have evolved – which of course they did. Talking enhances group co-operation, organisation, tactics, culture and it allows for group sizes to enlarge. (Dunbar 1996) FOXP2’s development could have contributed to our species’ expanding population. Apes do not have the same speed or precision of their orofacial apparatus as humans do. (Fisher 2008) This particular discovery underscores and consolidates the failure of the 1930’s language studies with apes mentioned earlier. And, the theories mentioned earlier, specifically of language evolving via Natural Selection to aid the fitness of communicative females, or of a Sexually Selected desire for loquacious others spawning descendents, both sit well with this genetic discovery.
Julia Roberts and George Clooney use their orofacial apparatus
In 2002 researchers dated the fixing of the amino-acid variant at sometime within the last 200,000 years. The fossil record for this pre-historic period shows anatomically modern humans had evolved and as FOXP2 has influence over anatomy, this finding was again concomitant with that. (This point is important and will be returned to later in the essay.) Closer analysis of a normal FOXP2 in humans has revealed the gene contains changes in amino-acid coding and a specific patterning of DNA sequence, (5’UTR and 3’UTR) indicating the gene had been selected for during recent human evolution. (Fisher, Pääbo, et al 2002)
In 2003 Pääbo et al, continued the comparative analysis of chimp and human FOXP2. Firstly by comparing human polymorphism (polymorphism – an occurrence located at the same gene of discontinuous forms of phenotype, Dawkins 1999) versus divergence between human and chimp species. Pääbo aligned the two species’ DNA for 5055 sequences. Pääbo started with the assumption, if most differences between humans and between humans and chimpanzees are neutrally adaptive then the relative differences across the gene should be the same for both within species and between species. (Pääbo et al 2003) But this was not the case. Pääbo found that both diversity and divergence differ at both the 5’UTRs and 3’UTRs sites. More diversity within the human species was found than was expected. 5’UTRs differed amongst humans by 0.04% and by 1.18% between species; there is an excess of divergence at 5’UTRs. 5’UTRs are made up of CpG sites (CpG - dinucleotide cytosine-guanine) these sites are methylated, and methylated sites are known to be hot-spots for transition. (Shen et al. 1994). The excess of divergence at 5’UTR was unexpected and suggests that 5’UTRs are under positive selective pressure in both chimps and humans. This 2003 finding was a significant development from Pääbo et al 2002 research. 5’UTR affects behaviour and cognition, it is beneficial and adaptive in humans and chimps. It is possible that DNA affecting levels of gene expression at both the transcriptional and transitional stages are targets for selection. (Pääbo et al 2003)
But in 2007 the plot thickened again and Pääbo et al, had to revise even more of their 2002 data. Pääbo et al took DNA from Neanderthal remains found in the El Sidron cave in Northern Spain. The bones were removed under sterile conditions in 2006, frozen and transported to the Max Plank Institute. Back in Leipzig nuclear DNA sequences were retrieved from Neanderthal remains. It was discovered our closest extinct relatives shared with humans the two selected differences to FOXP2 that differ between humans and chimpanzees. (Pääbo et al 2007) And not only were the shared Human and Neanderthal DNA alleles confirmed, but a new allele not seen in either humans or chimpanzees was discovered. If, as Pääbo et al have now shown, the Neanderthal had the same fine motor control of their orofacial apparatus, they could certainly express themselves in as sophisticated a way as our ancestors could. But as the Neanderthal had evolved another unique mutation to FOXP2 it is possible their language skills outstripped ours. This is a fascinating discovery for a number of reasons. Firstly, it has long been assumed by science the Neanderthal could not speak fluently like Homo sapiens, but could only grunt. (Liberman and Crelin, 1971) It was also assumed Neanderthal was cognitively inferior to man and this inferiority contributed to their extinction. These assumptions of Homo sapiens superiority were made in spite of Neanderthal paleontological remains revealing the Neanderthal were bigger, stronger and had larger brains than Homo sapiens. They also cared for their sick and elderly, buried their dead and possessed a ritualistic culture. (Dunbar 1996) (This supremacist assumption harks back to Premack’s comments, quoted earlier, regarding our subjective, linguistic possessiveness influencing approaches and theories in science and prejudicing the ape-language studies.)
Svante Paabo with a reconstructed Neanderthal skull. - Image copyright Frank Vinken
Secondly, as both Neanderthal and ancestral humans share the same selective sweep for FOXP2 adaptation the date for this selection needs to be put back from Pääbo’s et al earlier 2002 estimation of 200,000 years, to 300,000 – 400,000 years ago. The time when the fossil record indicates Homo sapiens and Neanderthal joint ancestor, Homo antecessor, was alive. (Pääbo et al 2007) The fossil record has many gaps and controversies; some believe Homo heidelbergensis, not antecessor, to be the ancestor of sapiens and Neanderthal. (Stringer 2006). From Pääbo’s et al 2007 research we know that our direct ancestors, who ever they were, were talking and sharing culture for far longer than previously thought. The Neanderthal became extinct 30,000 years ago, this date co-insides with the arrival from Africa into Europe of Cro-Magnon at approximately 50,000 years ago. As Neanderthals were clearly not grunting oafs their demise may have been more to do with living a sedentary, peaceful lifestyle (Dunbar 1996) than lack of intelligence or language. It probably wasn’t the arrival of “talkers” that wiped out the Neanderthal, but rather the arrival of xenophobic nomads. (Dunbar 1996)
Future Research
The last six years of research have revealed some of the structural evolution to FOXP2 is shared between chimpanzees and man and all of the changes to FOXP2 are shared between man and Neanderthal. The discovery of FOXP2 and its significance has brought us genetically and linguistically closer to our living and extinct closest relatives. For scientists advocating Homo sapiens linguistic supremacy this is a set back.
A logical next step would be DNA analysis from older hominoid remains, such as Australopithecus, an ancestral bipedal species that existed approximately 4 Million years ago, and had a cranial capacity similar to contemporary chimps. (Diamond 1991) Could early mutations to FOXP2 co-inside with bipedalism evolving? As the necks of our ancestors elongated so the larynx migrated downwards. (Dunbar 1996) Bonobos, like Kanzi, are more gracile than chimps and more prone to upright mobility, (Zihlman et al 1997) their necks are a little longer than chimps and in ape language studies they out-chat chimps. (Savage-Rumbaugh 2002) Pääbo et al may well see differences between the bonobo and chimpanzee FOXP2. Something else to consider is the discovery in 2006 by David Reich et al, at MIT. Ape and Human DNA has shown ancestral humans began their speciation away from archaic ape approximately 10M years ago. But the jointly evolving species continued to mate with each other for another 4M years. (This process created hybrids that mated with both ancestral chimps and ancestral humans.) Our ancestors finally split away from apes less than 5.4M years ago, perhaps as recently as 3M years ago. (Reich et al 2006) 4M years is a long period of hybridisation and speciation. Could the Sexual Selection of FOXP2’s 5’UTR have fuelled this continuous interbreeding between archaic man and ape? How far back into the belching primordial soup will scientists need to go to locate the origins of the selective sweep for tongue-twisting-talkers and beatboxers?
Lake Baikol in Southern Siberia, largest freshwater lake in the world.
Tasmanian Aboriginies
Yali man Papua New Guinea
Blond hair is frequnetly seen in Melanesian children
Denisovan/Melanesian admixture in a Polynesian princess
European "wildman"