Introduction

Biasing Language

In this chapter, we argue not only that the best approach to understanding the origins and present-day diversity of language is rooted in evolutionary theory, but also that extra-linguistic factors, more specifically biological ones in our genes, may play an important role in shaping language. Likewise, these factors do not act in a void, but interact with multiple constraints and affordances on different scales in parallel. So-called cultural evolution of language (Section 10.1.2) must thus be seen in a rich context (partially) molded by the biological and cognitive entities that ultimately acquire, use, and transmit language – us. Important factors in this context are therefore represented not only by the brain – it has been recognized for a while now that the brain indeed shapes language (Christiansen & Chater 2008) – but also by the anatomy and physiology of the vocal tract and hearing organs.

For those of us interested in the scientific study of language and speech (whom I will call here, for want of a better term, language scientists), keeping abreast with the relevant knowledge and scientific disciplines is becoming more and more difficult. Besides linguistics “proper” (and its classic sub-disciplines phonetics, phonology, morpho-syntax, semantics and pragmatics) now we must know some psychology, we have to be conversant in the cognitive neurosciences, understand something about language disorders and, lately, be on friendly terms with genetics. Familiarity with this latter discipline is increasingly necessary for meaningful discussions about language origins and evolution, its acquisition by children and the design of individually tailored effective second-language learning curricula, the structure of our capacity for language, and to address language and speech impediments, to mention just a few.

Unfortunately, fundamental notions of genetics are not yet part of the standard training in the language sciences, and this means that the interested language scientist must either ignore it at his/her own peril, acquire it piecemeal from heterogeneous popularization sources (with their associated uneven quality, reliability and relevance) or plunge head-on into the dense, confusing and exponentially growing primary literature. Another possibility would be to read one or more of the existing excellent introductions to genetics, genomics, biochemistry, population genetics, evolutionary theory, etc., but these are in general too broad, they address a very different audience and cover much too much material, most of it uninteresting and not directly relevant for the language scientist.

This book aims to fill this gap by offering an introduction to selected aspects of modern genetics and genomics, tailored for scientists involved in the study of language and speech. It tries to provide a condensed selection of relevant topics, briefly introducing the needed concepts, methods and results, and using – as far as possible – examples directly related to language, speech and hearing, while constantly pointing the interested reader towards important papers and recent developments and trends in these areas.

We saw in the last section (8.7) of Chapter 8 that humans certainly were not free from strong selective pressures acting mostly on local populations after they left Africa to conquer the whole world. Thus, it seems that, despite widespread ideas claiming otherwise, our culture did not completely shield us from the rigours of biological evolution, at least until relatively recently. Here we will continue this list of examples, but we will focus on selective pressures that were generated by our own culture, and that fed back on our genomes, shaping them. Sometimes, this moulding of our genomes through selection has favoured the cultural practices that generated those selective pressures in the first place, or even allowed the development of new cultural practices, creating a full co-evolutionary spiral between biological and cultural evolution. For an up-to-date and comprehensive overview of cultural evolution see Richerson and Christiansen (2013); gene–culture evolution is nicely introduced in Richerson and Boyd (2008) and Feldman and Laland (1996), while more recent developments also covering niche construction (discussed in Odling-Smee et al., 2003) are described in Laland et al. (2010), Richerson et al. (2010), and the special issue (number 366) “Human niche construction” of the Philosophical Transactions of the Royal Society B (Biological Sciences) from 27 March 2011.

The concept of linkage disequilibrium (or the tendency for some loci to be transmitted non-independently) is extremely important in our hunt for the genes influencing phenotypes. We will discuss here the two main methods that use it, association and linkage studies, introducing also some statistical concepts fundamental to properly conducting such studies and interpreting their results.

What is linkage disequilibrium?

As we saw in Section 4, the sweet pea flower colour is controlled by a locus with alleles A (dominant, purple flowers) and a (recessive, red flowers). Another locus studied by Bateson, Saunders and Punnett in sweet pea controls the shape of pollen grains and has two alleles: B (dominant, long grains) and b (recessive, round grains). By crossing pure lines of homozygous plants with purple flowers and long grains, AABB plants, with homozygous plants with red flowers and round grains, aabb, they obtained, in the first generation (usually denoted F1), only heterozygous plants with purple flowers and long grains AaBb, as expected given the dominance relationships at these two loci.

Exome and genome sequencing

How about the new and hot developments in high-throughput sequencing (including the “next-generation” and “third-generation” methods) that promise to make sequencing thousands of whole genomes feasible in terms of time, infrastructure required and costs? At the time of writing there is only one published pioneering study using such methodologies to investigate the genetic architecture of language and speech (Worthey et al., 2013), but it is to be expected that in the near future they will provide a sizeable proportion of unexpected and exciting new discoveries (see also Deriziotis and Fisher, 2013).

In a nutshell, it is feasible to sequence somebody's whole genome (or just the protein-coding part of it covering only the exomes – the wholeexome) and, in principle, we could look at any characteristics of the person's genome/exome and her/his phenotype of interest. However, this soon runs against a possibly unexpected problem, namely that individual genomes harbour a lot of variation, some of it shared by a sizeable proportion of the population (representing the so-called polymorphisms, usually taken to be those variants appearing in more than 1% or 5% of the population, SNPs being one example), but also some appearing only in a very few other individuals (such as other family members), being thus rare, and yet others specific to the individual (de novo mutations).

In diploid organisms, such as ourselves, the vast majority of genetic loci have two alleles, as discussed above. For example, the two loci in Figure 3.8 part (i) in the parent (left) have alleles A and a at the first locus, and B and b at the second, but there are also many loci at which the two alleles are identical. To use a real example studied by William Bateson, Edith Saunders and Reginald Punnett at the beginning of the 20th century (Bateson et al., 1905; Lobo and Shaw, 2008), let's for a moment look not at a human but at a sweet pea (Lathyrus odoratus), and more precisely at a locus that influences the flower's colour, with allele A resulting in purple flowers while allele a results in red flowers.

Dominance and recessiveness

But how can we talk about the effects of a single allele, such as A, in diploid organisms? We can look at homozygous individuals that happen to have two copies of the same allele at this locus and observe that AA individuals produce purple flowers while aa ones have red flowers. But what happens to heterozygous individuals that have different alleles at the same locus? In this particular case, the heterozygous Aa individuals (the order in which the alleles are written, Aa or aA, is irrelevant here) produce purple flowers, just like the homozygous AA individuals.

Phenotype, genotype and environment

It is unquestionable that both “nature” and “nurture” are required for the development of a linguistic human being. Lacking “nature” will limit language development no matter how much “nurture” there might be, as many a pet owner can easily confirm. This is seemingly supported by studies of chimps (such as Nim Chimpsky and Washoe) reared in conditions similar to those experienced by human babies and infants, but which nevertheless fail to go beyond a rather limited level of language usage. On the other hand, having “nature” but lacking “nurture” is equally devastating, as shown by the cases of children who, for various reasons, have not been exposed to language during the so-called critical period for language acquisition (a well-known case being Genie) and who fail to develop full-blown language despite considerable efforts.

Thus, if we denote, in a highly abstract manner, the “nature” as G (from genetics) and the “nurture” as E (for environment), then we can attempt to write down a symbolic equation describing how these two factors relate and interact in producing the phenomenon of interest, P (for phenotype).

We, as biological organisms, are composed of trillions of specialized cells, organized and coordinated through mechanisms of bewildering complexity and beauty, and any proper understanding of what we are must, at some point, face this level of organization. Far from a brutal reductionism or molecularism, we must understand how molecules interact and organize in increasingly complex structures and how these “levels” of organization continuously “talk” to each other and with the “environment” – including culture – because all these “higher” structures – language, art, culture in general – build upon, and form an organic whole with the molecules which compose us. It is emphatically not about explaining, say, the Great Vowel Shift in terms of quantum interactions between carbon, oxygen and nitrogen atoms, but about articulating the appropriate historical, phonological and phonetic explanations in their proper living human context, which imposes specific limitations and creates specific affordances. Ultimately, this will allow a proper understanding of the processes that made culture possible and that continue to shape and remodel it to the present day.

We will introduce here this maddeningly dynamic aspect of genetics as it has already forced massive reappraisals of what we thought we knew and what we were complacently taking as dogma, from the nature of “genetic information” and “innateness” to how evolution works, and how culture and language are shaped by, and shape in turn, our biology.

This book is intended to give a relatively up-to-date introduction and overview tailored to the needs of speech and language scientists who venture into genetics and do not have the time to delve into more comprehensive general introductions such as Brown and Brown (2012), Snustad and Simmons (2010) and Jobling et al. (2013). It tries to cover selected topics using wherever possible examples from speech and language and tries to give a level of understanding that should be enough for working in cross-disciplinary teams and understanding modern genetic and genomic research. The book tries to refer both to up-to-date reviews and to primary research reports, which should allow the reader to deepen his or her understanding on selected topics or to understand alternative or emerging interpretations and views.

The main aim of the book, if one must be singled out at all costs, is to allow language and speech scientists to appreciate the complexity and beauty of their topic of interest's genetic infrastructure and understand the constraints and opportunities such an understanding has on their day-to-day work, be it designing psycholinguistic experiments into the acquisition of non-native phonological contrasts, studying sound change or the typology of negation.

However, given the target audience, the length of the book and (not least) my own competence, I had to be highly selective in the topics covered, leaving aside a number of extremely fascinating issues. Among them, the use of concepts and methods adapted from evolutionary biology to study language change and evolution such as phylogenetics and phylogeography (for introduction, discussion and up-to-date research see for example Dunn et al., 2008; Atkinson and Gray, 2005; Gray et al., 2010; Bouckaert et al., 2012; Dediu, 2011b). Likewise, I must only very lightly mention here the use of genetic and linguistic data to reconstruct population history, whereby usually uniparental markers (mitochondrial DNA and the non-recombining region of the Y chromosome) are combined with historical linguistics (for a good overview see Jobling et al., 2013, and the reviews in the special issue “Global genetic history of Homo sapiens” of Current Biology 2010 volume 20 issue 4).

What we have discussed so far about genes and their effects made relatively little reference to evolution and instead focused on the methods used to discover genes, the proximate molecular mechanisms through which they affect the observable phenotype, and their complex inter-relationships. However, as Dobzhansky (1973) famously argued more than 40 years go, “nothing in biology makes sense except in the light of evolution” and the genetic bases of speech and language are no exception. Quite the opposite, we need to understand the why questions behind the how's, such as: why are there so few genes in humans? Why are the interactions between genes so complex? And why are there differences in interactional complexity among genes?

The next sections will introduce the notions of evolutionary biology needed in order to understand questions relevant for the genetic bases of language and speech but will also briefly sketch, where feasible, the larger context into which these notions, methods and results fit. Therefore, it is not intended to be a general introduction to these topics as there are many excellent textbooks such as Halliburton (2004) and Jobling et al. (2013), nor does it cover all possible aspects or go into great depth, but instead tries to sample a selection of relevant topics, with an accent on recent developments that might help shed light on language evolution and change.