The Neurobiological Foundations of Language

Although we have not yet solved many of the mysteries surrounding the brain/mind (What is consciousness? What is voluntary action?), today we know far more than we used to a century ago about the relationship between language and the brain. The discipline that studies this relationship, namely neurolinguistics, is relatively novel and have its roots in some discoveries on the brain made around mid-1850.

“Tantan”: The Patient that Could not Talk

In 1861, the French physician Paul Pierre Broca met monsieur Lebrogne, a man that, since 1811, could produce only two syllables when speaking: tan-tan. This was the reason why he soon became known as “Tantan”. He had no trouble in comprehending others but, except for those two syllables, his linguistic production was null. This sort of block did not affect other cognitive or mental faculties, so Broca hypothesized that the pathology had to do with a specific impairment of language production. Tantan's conditions soon became worse, and the patient died briefly after in the same year. After Lebrogne’s death, an authopsy was performed. Broca observed, as he had predicted, that the brain presented a lesion on the left frontal lobe, which was evidently the cause of the pathological condition. Broca’s idea that higher cognitive capacities like language could be localized in the brain was not universally accepted at this time. A few years later, another physician, Carl Wernicke, published an essay on the possible localization of functions in the brain (The aphasic symptom complex: a psychological study from an anatomical basis). Wernicke’s relation was based on the observation of study of aphasic cases too. For this reason, the brain’s areas where Broca and Wenicke located the language faculty were called respectively Broca’s Area and Wernicke’s Area (Figure 1). Later developments in neurology showed that the localisation of cognitive capacities is much complex than how Broca and Wernicke imagined. Faculties like languages are not exactly located in one place, but rather they are distributed in a set of different but interconnected networks. In this sense, Broca’s and Wernicke’s areas are now considered not the language’s areas, but rather the areas that the language faculty exploits the most. Although Broca’s and Wernicke’s intuitions were incomplete, their work was determining in hypothesizing that mental capacities have an anatomic foundation in the brain (Cacciari 2011, Moro 2015).

Figure 1. Cortical areas that have been shown to be involved in speech processing (Leuthardt, Pei, Breshears, et al. 2012) .

The Brain Structure

To spot the language faculty in the brain, it is necessary to describe what are the main anatomic features of the brain and how they are structured, in terms of areas and related functionalities. The human brain is divided into two portions called ‘hemispheres’. The left hemisphere is involved in the use of linguistic abilities (phonology, morphology, syntax, partly semantics) and praxical (symbolic and gestural expressions). This hemisphere, including areas more related to language, seems somewhat prevailing in cognitive functions mediated by language. The right hemisphere is mainly involved in different tasks, related to spatial and perceptive elaboration. Thus, capacities like perceiving a melody or spatial relation seem to be related more to the right hemisphere. From this first description, it seems linguistic tasks are exclusively performed by the left hemisphere. This vision is incomplete. Indeed, the language faculty is distributed among the two hemispheres, with the right one being responsible for some components of language, like prosody (melodic features of language) and complex semantics (not related to the meaning of single words). The two hemispheres are anatomically connected by the corpus callosum, a wide, thick nerve tract, consisting of a flat bundle of commissural fibers. Each hemisphere is subdivided into two lobes. The names of the overall four lobes are frontal, parietal, occipital and temporal. The structure of the brain consists of two main elements: neurons (which are responsible for the transmission of nerve impulses) and glia (which supports and feeds the neurons). The frontal lobe hosts important functional areas of the brain: the primary motor cortex, the pre-motor cortex, the supplementar motor area, Broca’s area, the prefrontal cortex. The parietal lobe host two functional areas: the primary somatosensorial cortex and the posterior cortical cortex. The temporal lobe hosts the following areas: Wernicke’s area, the hippocampus, the amygdala. The occipital lobe hosts the visual cortex (Cacciari 2011).

Focusing more on the language in the brain, let us describe which areas are more involved in various language components. The perception and production of spoken sounds is based on a fronto-temporal circuit (auditory and motor aspects) and a fronto-parietal circuit (phonological memory work). The morphological segmentation of words and their interpretation is based on the interaction between the left inferior frontal cortex and the temporal cortex (Marslen-Wilson 2007). Tasks related to the processing of grammatically and semantically plausible sentences activate different locations in the left hemisphere, among which the temporal lobes and the inferior frontal cortex.

Monolingual and Bilingual Brains: How Are They Different?

Is there a difference between monolinguals and bilinguals (or multilinguals) in terms of neural connections? The number of studies on this topic has grown considerably in the last years. The following results have emerged in the literature: i) language is represented differently in early bilinguals (L1 and L2 acquisition before age of 6) with respect to late bilinguals (after age of 13): the effect of age of acquisition seems largely independent from the level of fluency; ii) early bilinguals show an augmented activation of the right hemisphere; iii) bilinguals who have acquired L1 during infancy and L2 in a later stage functionally organize language mainly in the left hemisphere, differently from early bilinguals; iv) left hemisphere's dominance is wider in less proficient bilinguals than in more proficient ones (Hull & Vaid 2007).

The Syntax in the Brain and Universal Grammar

Another area of inquiry focused on the neural correlates of syntax in the brain, under the influential hypothesis that there is a set of principles which are universal for languages. The existence of a Universal Grammar has been formulated by Noam Chomsky (Figure 2) at the mid-60s (Chomsky, 1965). Being considered valid for all possible human languages, the logical step has been to hypothesize that such a set of principles may be spotted at the functional level of the brain. In other words, possible languages (i.e., those which behave according to Universal Grammar) should activate specific areas in the brain with respect to impossible language (i.e., those which violate Universal Grammar).

Figure 2. The linguist Noam Chomsky (b. 1928) claims all languages share a set of constraints called 'principles' (Peters, 1977).

Consider one of the properties attributed to Universal Grammar called structural dependency. According to this property, “no syntactic rule refers to the number of words of a sentence or to the position of a word in a sequence” (Moro 2015). To put it simply, there is no attested language that include a rule like ‘put the verb always in the x position’. It is true that languages may be classified by the way grammatical elements are ordered (e.g., S(ubject)V(erb)O(object) vs. VSO languages), but the position of these elements is never predictable arithmetically. Consider the English sentence the cat that the dog chased ran away, where chased refers to the dog while ran away refers to the cat. This is evident for every native (or enough proficient) speaker of English and it is not based on the number of words or the position of some of them in the sentence, but rather on grammatical relations which tie each name with its verb. This property is specific of human languages and does not affect other activities. Experiments have been designed in order to observe if the way people process languages which undergo the principles of Universal Grammar like structural dependency is different from the way they process structures which violate such principles. The results of such experiments (Figure 3) have shown that the brain processes the two kinds of rules (a rule respecting structural dependency vs. a rule not respecting structural dependency) in a macroscopically different way. In particular, rules involving structural dependency activate brain areas generally related to the faculty of language, mainly Broca’s area (Marcus, Vouloumanos & Sag 2003, Moro 2015).

Figure 3. Functional Magnetic Resonance Imaging (fMRI) shows regions of activation in the brain and it has been widely use in neurolinguistics research, including experiments on possible vs. impossible languages (Washington Irving, 2004).

Conclusion

The point of departure of the study on the relationship between language and the brain have been the inquiries on pathological cases of aphasia and, since then, enormous advancements have been made both in the study of the brain on its own and on the location of areas involved in language tasks. More recent studies have underlined the relevance of theoretical linguistics for neurosciences, showing that principles like structural dependency a) are general to all languages according to linguistic reports and b) seem to have a neural base in the brain, in terms of processing and activation. Of course, such results are not conclusive and raise further questions (Moro 2015). Nevertheless, such convergence between theoretical work in linguistics and neuroscience is surely surprising.

References

Visual Sources