Neural Correlates of Neurotypical and Pathological Language Processing

In 1874, the German neuroanatomist Eduard Hitzig presented his ideas on language and the brain to the Berlin Anthropological Society (cf. Hagner, 2000). He interpreted aphasia as the loss of “motor images of words” very similar to neuronal representations of other types of motor activity in humans and non-human animals. Hitzig was sharply criticized by Heymann Steinthal, a linguist, who had analyzed most of the available aphasiological evidence at that time. Steinthal was convinced that the leading view of language in the second half of the nineteenth century completely underestimated the complexity of language as a psychological function. He concluded that language had to be conceived as a complex psychological mechanism beyond the current view of the leading neurologists and neuroanatomists. Although Steinthal discussed his ideas with many important scientists at the time, the leading theoretical views on aphasia and language prevailed. These views had begun to gain influence after the scientific descriptions of motor aphasia by Paul Broca and of sensory aphasia by Carl Wernicke. Wernicke (1874) incorporated both findings into a model of a motor speech center in the inferior frontal and a sensory speech center in the superior temporal cortex, the two being connected by a massive fiber bundle (arcuate fasciculus). Lichtheim (1885) added a “concept center” to this model and arrived at his famous “house model” of language that supposedly made it possible for all types of aphasic syndromes to be explained. Although the so-called Wernicke–Lichtheim model of language has been very influential as a heuristic for both research and therapy, it is faced with several problems (see also Hickok & Poeppel, 2004, 2015): the idea of a few aphasic syndromes is not sufficient to explain the variety of aphasic phenomena; nor is their association to different anatomical areas as clear as the classical model suggests. Furthermore, the model is (psycho)linguistically strongly under-specified. Today, Steinthal's claim that the complex structure of language is inherently tied to a differentiated brain network has gained much influence. Many neural models of language are inextricably bound to (psycho)linguistic theories. Furthermore, new techniques to measure brain activity in vivo give us an idea of how complex the neural basis of language is and how the different language functions are supported by a distributed network of cortical as well as subcortical areas. Through these new methods, we gain understanding about language being processed in time . When the lexical entry of a word is retrieved, many different types of linguistic information (phonological, syntactic, and semantic) need to be integrated into a sentence representation. This happens very fast, even though the process engages multiple interactions between information types. Thus, a model of language not only has to describe anatomically and functionally distinct language-related areas in the brain. It must also explain when these different areas come into play and interact with each other so that language is produced and understood in time-critical conditions of communication.