We are excited to share our latest publication exploring the history of the neurobiology of speech and language.

Reviewing the history of a science is essential to understand its richness, its complexity as well as the current challenges that it faces. The neurobiology of language—the study of the relationship between the brain and language—began in the 19th century. If this science was first associated with the medical field with a significant contribution from anatomy, it gradually transformed to integrate approaches and theories from other health sciences (e.g., speech therapy, neuroscience, geriatrics and medicine), as well as from social and human sciences (e.g., linguistics, psychology, psycholinguistics, and cognitive sciences) and from natural sciences and engineering (e.g., computer science and more recently of artificial intelligence). This richness induced by multidisciplinarity also generates challenges.

Pascale Tremblay, the director of the lab and her colleague Simona Maria Brambati, a professor-researcher in the psychology department at the University of Montreal, traced in this article the history of the neurobiology of speech and language: from pioneering works, centred on the study of the injured brain, to the advent of leading-edge technologies that have significantly improved our knowledge of language networks and changed the way we study these networks (see a summary in Figure 1). The researchers conclude by presenting two current models and the challenges of the discipline. In this text, we present a summary of the article.

Figure 1. Chronology of the study of the neurobiology of language in relation to the main methodological advances

The First Model of the Neurobiology of Language

Since the discovery, in the early 19th century, that the brain is a complex and differentiated organ, research has attempted to understand the organization of the brain and to identify the areas involved in language processing and production. Early work focused on people with brain pathologies. The studies on language were carried out on single cases of people with brain damage. Given the absence of technological means to study the functioning of the brain of a living person, research was carried out post-mortem.

During the 19th century, the work of two doctors resulted in the classic conception of language: language is processed in specific and circumscribed areas of the brain, mainly located in the left hemisphere. The French surgeon Paul Broca was one of the first scientists to associate a specific part of the brain (the inferior frontal gyrus, which would come to be named “Broca’s area”) with speech functions. Following this work, the German doctor Carl Wernicke discovered another part of the brain involved in language, namely the posterior temporal lobe, from the study of the brains of patients capable of speaking, but incapable of understanding language and without intellectual deficits. Wernicke proposed the first neurobiology of language model, which gave rise to this field of scientific enquiry. According to this model, two parts of the brain are involved in language: Broca’s area (inferior frontal gyrus), considered a motor center for speech production, and Wernicke’s area (posterior part of the superior temporal lobe), responsible for understanding of speech. These two regions were believed to be connected by a structure called the arcuate fasciculus. This model is shown in Figure 2.

This first and very simplistic model generated a lot of knowledge still relevant today.

Figure 2. The first model of the neurobiology of language

The Birth of the Cognitive Neuroscience of Language

In the mid-1900s, the study of language gradually moved away from the medical field through the influence of the methods and research questions of psycholinguistics and neuropsychology. But it was at the turn of the 20th century that major changes in the neurobiology of language occurred: the advent of leading-edge technologies gave rise to the cognitive neuroscience of language. Indeed, technological advances have enriched and refined our knowledge of the processes underlying the production, perception and understanding of language (see a summary in Figure 3).

The emergence of new research methods has significantly transformed both the objectives and approaches of language research. Today, the focus extends beyond clinical populations to include the healthy and damaged brain, across the lifespan, from youth to old age. By broadening the scope to encompass individuals without brain lesions, studies leveraging advanced technological tools have provided valuable insights into the intricate relationship between language and brain function. This expanded perspective has greatly enriched our understanding of language processes in both typical and atypical populations. Electrophysiological approaches, such as EEG (to find out more, consult our article by clicking here), have enabled a better understanding of the temporal course of language processing. The advent of brain imaging tools in the 1990s (to find out more, read our article by clicking here), revolutionized our ability to study the brain with higher spatial resolution. These tools enabled scientists to map brain regions by tracking blood flow variations linked to neuronal activity. Among these tools, MRI (Magnetic Resonance Imaging) stands out as one of the most widely used techniques in neuroscience research. MRI has provided groundbreaking insights into acquired language disorders in individuals with brain lesions, such as those with post-stroke aphasia or primary progressive aphasia. Beyond pathology, MRI has also advanced our understanding of brain plasticity—the brain’s remarkable ability to reorganize itself following injury, or even in response to learning experiences, such as acquiring a second language. Crucially, MRI has allowed scientists to study language networks in healthy individuals, revealing that language relies on extensive, dynamically interconnected brain networks, especially in adults.

In parallel, brain stimulation methods, which began emerging in the mid-20th century, have further expanded the horizons of language research. These methods, designed to map language functions in the brain, have deepened our understanding of speech and language mechanisms. Among them, transcranial magnetic stimulation (TMS) stands out as a non-invasive, highly precise technique (to learn more, click here).  TMS operates on the principle of electromagnetic induction and differs from imaging techniques in a fundamental way: it can not only identify brain regions involved in speech and language, but also modulate brain activity. For example, TMS can stimulate specific brain regions to enhance certain language functions, offering exciting potential for both research and therapeutic applications.

Figure 3. The contribution of technological advances to the development of knowledge about language

Contemporary Models and Challenges

The researchers highlight that advancing our understanding of language relies on the complementarity of diverse research methods and theoretical approaches. As knowledge evolves, some earlier findings are reconsidered. For instance, it is now clear that Broca’s area is not solely a motor area nor exclusively responsible for language production. Additionally, both hemispheres of the brain contribute to language perception and articulation, even though the left hemisphere remains more dominant for certain language functions.

The article also discusses two major contemporary models of speech processing:

  1. The Dual-Stream Model of Speech Processing developed by Hickok and Poeppel in the early 2000s suggests that two distinct neural pathways support language functions:
  • The dorsal stream, involved in speech perception and production, and
  • The ventral stream, responsible for word processing and language comprehension.

These pathways engage different brain regions (see Figure 4 for an illustration).

  1. The DIVA/GODIVA Model offers valuable insights into speech production and learning, emphasizing the role of sensory feedback during speech and the sensorimotor system that underpins language production.

These models together provide a comprehensive framework for understanding the neural mechanisms of speech and language processing, illustrating the complexity and richness of this uniquely human capability.

Figure 4. The dual-stream model

The authors conclude the article by highlighting several key challenges that currently shape the field, both in terms of research questions and methodologies. They emphasize the need for comprehensive models that account for language development across the lifespan, from birth to old age, while integrating a broader range of subcortical structures, such as the cerebellum, basal ganglia, and thalamus. A significant obstacle in language research is its reliance on human studies, given the profound differences between human language and animal communication, even among our closest relatives, the great apes. In contrast, research on cognitive functions like spatial navigation, memory, and learning often leverages both human and animal models, leading to major breakthroughs in understanding the underlying neurobiological systems.

However, technological advancements are opening up exciting new opportunities to address these limitations, paving the way for a deeper understanding of language—a uniquely human and profoundly complex function.

Suggested readings :

,