Rhythms in the Brain: Exciting Times for Neurolinguistics
Brain oscillations may help explain how the brain processes language.
Posted Aug 29, 2018
Language is in the brain. This is not a truism, but an important finding after decades of inquiry about the biological nature of language. During its early years, neurolinguistics (the branch of linguistics dealing with how the brain processes language) could only rely on anatomical studies from patients with language deficits. This is how Broca’s area (the part of the brain roughly involved in sentence assembly) and Wernicke’s area (the region where word meanings are stored) were identified. Decades later, the development of functional neuroimaging allows for the sharp pictures of the physiological changes that occur in the brain of healthy and affected people, in terms of blood flux, electrical activity, and so on, in response to diverse linguistic stimuli. Over time, an impressive body of data has been gathered regarding different aspects of syntax, semantics, or phonology processing. As a result, we have been able to draw precise brain maps of language, that show that it activates many areas of the brain in precise and concerted ways, the so-called “language areas” (Figure 1). Overall, this is an impressive achievement. However, these pictures cannot be equated to the representations and computations that are important for language (and for linguistic theory).
As David Poeppel (Figure 2) has put it, mapping is not explaining.
To begin with, language relies on a complex networks of brain components that are not specific to language. On the contrary, they perform basic computations that are also needed for other cognitive functions. Nonetheless, if we wish to explain how the brain processes language we also need to address two important shortcomings of current neurolinguistic studies. First, neurolinguistics makes broad conceptual distinctions (syntax vs. semantics, morphology vs. syntax, etc.), which involve multiple neural components, computations, and representations. Second, “the fundamental elements of linguistic theory cannot be reduced or matched up with the fundamental biological units identified by neuroscience.” Putting it differently, as also noted by Poeppel, we need to spell language “in computational terms that are at the appropriate level of abstraction (that is, that can be performed by specific neuronal populations).” At present, a growing number of neurolinguists think that brain rhythms could be the best level of abstraction. Several reasons account for this.
First, brain rhythms, or patterns of electrical activity, are primitive components of brain function. Accordingly, we expect them to be connected to computational primitives of language too. For example, the assignment of language-relevant features, like "tense" or "case," can be satisfactorily interpreted as the embedding of high frequency oscillations inside oscillations operating at a slower frequency (Figure 3). Similarly, some rhythmic features of speech have been successfully related to specific brain oscillations.
Second, the hierarchy of brain oscillations has remained remarkably preserved within mammals during evolution. Accordingly, we expect that brain oscillations help us as well to better understand how language evolved in the species, as the human-specific pattern of brain activity underlying language has been shown to be a slight variation of the pattern observed in other primates. Likewise, in our research we have shown that differences in the oscillatory activity of the brain can be expected for extinct hominins, and account for the inferred differences in their language abilities compared to us. It is pretty obvious that we cannot track the oscillatory activity of the Neanderthal brain. Instead, we have looked for (and found indeed) signals of differences in the expression levels of genes responsible for basic aspects of the brain oscillatory activity relevant for language (Figure 3).
Finally, we have learnt that cognitive disorders that result in problems with language, such as autism or schizophrenia, exhibit disorder-specific profiles of brain rhythmicity. Incidentally, this might be of particular interest for clinical linguistics, because these profiles might help diagnose disorders earlier and more accurately. Certainly, we still lack confident mappings of abnormal patterns of brain oscillations to language-specific abnormal profiles of disorders. Nonetheless, this is not an impossible task. For instance, in our recent research we have shown that the problems that people with autism experience with speech perception, tone recognition, and parsing phonemic representations seemingly result from degraded γ and θ synergy (two types of basic brain oscillations) in specific brain areas. Ultimately, because language (and language deficits in these clinical conditions) have a genetic basis, we expect to find that this abnormal oscillatory activity correlates with specific gene mutations, opening the door to establish bridging links between genes, brain oscillations, and language, in both pathological and neurotypical populations. We have found promising evidence of this (Figure 4).
Eventually, because brain rhythms connect to aspects of human biology (and of human language) that are known to vary within the species, but also across species, they are expected to help us improve our theoretical models of language, language disorders, and language evolution. Needless to say, we still need to spell out the details, but we are now pretty sure that brain rhythms are one of the protagonists of this fascinating story about what makes us humans. Definitely, these are exciting times for neurolinguistics.
Acknowledgements: I am deeply indebted to Elliot Murphy for our joint research and collaboration over the years.