Language is in the genes
It is accepted as a scientific fact that the prerequisite for the unique human aptitude for language and speech must be in the DNA of Homo sapiens. Yet a single “language gene” does not exist. Researchers in the Language and Genetics Department at the Max Planck Institute for Psycholinguistics in Nijmegen are getting to the bottom of the genetic traces of the human ability to communicate.
“Are you still studying this one gene? Shouldn't you be finished with it by now?” Every now and then, Simon Fisher encounters this puzzled enquiry from his parents. When telling this story, the Max Planck Director and Head of the Language and Genetics Department at the Max Planck Institute for Psycholinguistics in Nijmegen has to laugh. To him, it’s obvious that he’s a long way from letting go of this gene, even 15 years after his great coup. After all, the secret behind FOXP2, the gene that the media described as “the language gene” in 2001, is still far from being resolved –- just like the genetic background to the unique human talent for communication. “We can be sure that the basis for language and speech lies partly in the genome”, says Fisher. “But it was clear from the outset that a single gene cannot explain this outstanding ability.”
But let’s start at the beginning. In the 1990s, Simon Fisher, who was then still at the Wellcome Trust Centre for Human Genetics at Oxford University, worked with his colleagues to examine the genetic material of a unique British family. The so called “KE family” came to the attention of the scientific community when it was noted that many members, across several generations, suffered from serious speech impairments. The affected individuals experienced difficulties in clearly articulating certain sounds, but also had problems with many aspects of language, including sentence construction and grammar, although their overall intelligence seemed normal.
Following years of meticulous and detailed work, Fisher and his colleagues managed to genetically narrow down the family's problem. “This was before the first Human Genome had been fully sequenced, when the search for the genetic causes of certain phenomena was still very time- and labour-intensive”, says the Max Planck researcher. Finally, they found the problematic area in the genome: they discovered that a single letter of DNA is modified in one gene – FOXP2 – in those affected by this speech disorder. This leads to modification of an amino acid building block in the protein that the FOXP2 gene makes, damaging the way it works and disrupting the language development of the people who carry it.
However, FOXP2 is not a uniquely human gene. It is widely distributed among vertebrates, for example it can be found in non-human primates, rodents, birds, and even in reptiles and fish. And it is most active in comparable regions of the brain in many different animals. In-depth studies on various species suggest that the gene has important roles to play in the networking of brain cells, i.e. in how they connect up with each other and how these connections change during learning. Songbirds, for example, need FOXP2 to be able to learn to sing – in a sense, they need it to be able to communicate. Mice, on the other hand, need the gene when they learn to make sequences of movements.
Moreover, FOXP2 is highly conserved, meaning that the gene makes a very similar protein in species that are distantly related. This indicates that it must be an ancient gene. For instance, humans and mice differ only in three amino acids. Interestingly, Svante Pääbo at the Max Planck Institute for Evolutionary Anthropology in Leipzig, working in cooperation with Fisher in 2002, discovered that two of these amino acid differences happened in human evolution after our lineage separated from that of chimpanzees. Subsequent work by Pääbo's team showed that Neanderthal FOXP2 made an identical protein to that found in the modern human. This is in line with a growing appreciation that – contrary to earlier assumptions – our extinct cousins may have already had some speech abilities. Perhaps one of the crucial genetic foundation stones for language and speech was laid early in evolutionary history: after the separation of humans from chimpanzees but before the emergence of Neanderthals, thus at least 300,000 to 400,000 years ago.
Speech and language impairments cannot be traced back to one single modified gene
But is the key to our unique ability to communicate actually present in the human version of FOXP2? Presumably not – at least not only there. There are countless individuals who have difficulties with speech and language whose FOXP2 gene in their genetic make-up is faultless. As Fisher's more recent work with new DNA sequencing techniques shows, this can sometimes be due to damage to a gene other than FOXP2. However, in most cases, the issue is so complex that it cannot be traced back to only a single modified gene. That applies even more to the diversity of speech and language abilities in the general population.
“We can view this diversity as a multifactorial phenomenon, similar to body size, blood pressure or weight”, says Fisher. This means that an entire array of genes determine how well an individual can speak – in the same way that numerous genes affect how tall or heavy an individual is or how susceptible a person is to having high or low blood pressure. Individual genes therefore do not necessarily need to be mutated. In fact, the issue centres on polymorphisms, as they are known: variants in the genome, which often affect only one single DNA letter and which are widely distributed in the population. A single polymorphism usually doesn’t have a major effect by itself. Instead, several of them act together and lead to some people growing very tall, for example, while others remain small.
A very similar principle can be envisaged when it comes to language - as here, too, there is variation from one person to another. Some people are particularly eloquent or find it easy to learn other languages, while others find it difficult to articulate clearly or can only formulate and understand simple sentences. Of course, there are environmental factors influencing individual speech and language abilities. ”But we know that variation in language skills is also highly heritable, meaning that it is strongly influenced by genetic make-up”, explains Fisher. ”"In contrast to our colleagues who study size, weight or blood pressure, we face a very specific challenge. We are dealing with phenomena that aren't easily measurable.” While anyone who has a folding rule can take his or her physical measurements, the gradations between “good speech” and “poor speech” can be difficult to define. In addition, because the effects of polymorphisms are small, scientists need to measure their traits of interest in very large cohorts of people, preferably thousands of individuals.
“This means that we could indeed hunt for polymorphisms in the genome, something that is very easy to do with modern molecular biology methods”, says Fisher. “But we can’t yet reliably interpret their effects on language.” With this in mind, the scientist is currently working with colleagues from other Departments at the Max Planck Institute for Psycholinguistics to more clearly define the various skills that underlie aspects of speech and language and make them easier to measure in a standardised way, for example by using online or app-based testing. This could one day lead to a better understanding of the many small effects of various gene variants on language and speech.
Thus, it’s not a matter of finding “the” language gene but of decoding the genetic and neurobiological networks underlying speech and language. Polymorphisms have a role to play here. They are, as it were, the tiny adjusting screws that determine the fine-tuning of linguistic ability. But there are also key areas in each network that determine whether the interaction works in principle or whether individual areas, or even the entire network, are malfunctioning. And that is ultimately leading Fisher and his colleagues at the Max Planck Institute in Nijmegen back to FOXP2.
Orchestra of language and speech
As is now known, FOXP2 is a “transcription factor”. As such, it regulates the activity of up to 1,000 genes. This can be imagined as an orchestra in which the conductor controls the interaction and thus combines the cacophony of chaotically blaring instruments into the harmonious melodiousness of a collective piece of music. And it is clear that in the FOXP2 orchestra there are some musicians who are more connected to language and speech than others – at least in humans.
Fisher and his team have therefore applied themselves to trawling the extensive network of genes coordinated by FOXP2 to search for those that are most relevant to speech and language, and study their functions in the brain. For instance, in their earliest forays into this area they already struck it lucky: with a gene known as CNTNAP2, which is switched off directly by FOXP2. Defects in this gene correlate with speech problems and autism. And, CNTNAP2 is particularly active during the development of the brain and affects neural connections, especially in the neural networks that play a role in language and speech.
In the meantime, the researchers in the Department of Language and Genetics have their eyes on even more interesting candidate genes. But once again, it is clear: until the extensive network that is coordinated by FOXP2 in its role as orchestra conductor is grasped and understood, a great deal of work remains to be done. It is therefore very likely that the gene that Fisher has been researching for more than 15 years will continue to occupy his attention for a long time to come.