Department of Neurosurgery, St Elisabeth-Tweesteden Hospital, Tilburg, The Netherlands
In the late nineteenth century, the theory of localization was strongly supported by experimental evidence from both electrocortical stimulation and cytoarchitectonics (for details, see Chaps. 5 and 6, respectively). Fritsch and Hitzig had demonstrated in 1870 that cortical stimulation at specific sites of the frontal lobe of a dog elicited contralateral muscular reactions, whereas stimulation of other regions did not result in any noticeable response . Brodmann had parcellated the cortex of the human brain on the basis of histological differences in cell population and neuronal architecture and came up with 43 different areas in his now famous cytoarchitectonical brain maps .
Following Broca and Wernicke, several authors published on language models and/or language-related anatomical regions. Ludwig Lichtheim (1845–1928) expanded Wernicke’s diagram and explicitly added a ‘concept centre’ (Fig. 4.1a). This subsequently enabled him to explain why some patients had intact repetition despite a comprehension disorder or non-fluent speech. Lichtheim adhered to the same principles as Wernicke, namely, that speech is a learned behaviour that depends on centres for motor and sensory images and a reflex arc. He also assumed—as did Wernicke—that pure lesions (such as those specified with the numbers 1–7 in his model) would only seldom occur, as damage from pathological lesions was generally non-selective and more extensive. Lichtheim added anatomically distinct centres for reading and writing as a new layer to his model (compare Figs. 4.1b and 2.3c) . He did not support his theoretical assumptions with significant experimental evidence and seemed aware of the hypothetical status of his new classification. Still, he associated his centre with anatomical regions:
(Top left, a) The famous Wernicke–Lichtheim or ‘house model’ in which Lichtheim described four new possible aphasic syndromes. Numbers 1–7 refer to theoretical lesion sites that are explained in the paper. A centre of auditory images, B concept centre, M centre of motor images, a acoustic pathway, m motor pathway. (Top right, b) The model has been expanded with centres to enable reading and writing. O centre for optical images, E centre from which the organs for writing are innervated. (cf. Wernicke’s diagram in Fig. 2.3c). In addition to that, Lichtheim here indicated that a new set of pathways has shown up (dotted lines) after destruction of the original pathway AB, suggesting rewiring and plasticity of the brain. (Bottom, c) The concept centre ‘B’ is anatomically distributed over a wide region of the sensorial space and is not located in one particular region of the brain (Figures taken from Compston, 2006 )
The motor-image center is localized in the ‘[part …] of the lowermost left frontal convolution lying against the Sylvian Fossa’ and the sound image center ‘in the temporal convolution lying on the opposite site.’ The connection between the two centers goes through the insula or directly adjacent regions. The center for conceptions represents the exception: ‘My view tends to assume […] that the concept formation is not linked to a location in the brain but is a common function of the entirety of the sensory areas instead’. 
Although the word ‘centre’ suggests otherwise, Lichtheim did not believe that concepts could be localized to a particular brain area, as he explicitly stated in his paper:
Though in the diagram B is represented as a sort of centre for the elaboration of concepts, this has been done for simplicity’s sake; with most writers, I do not consider the function to be localized in one spot of the brain, but rather to result from the combined action of the whole sensorial sphere. Hence the point B should be distributed over many spots; and the commissures MB and AB would not form two distinct and separate paths, but consists of converging radiations from various parts of the cortex to the point A and M [he refers to Fig. 4.1c; Lichtheim later called this the ‘semantic field’]. This admission does not do away with the possibility of the interruptions in the commissures BM, B1M, B2M, &c; but leads us to expect that any simultaneous break in them must occur close to their entrance into the lower centers M and A. 
Because of this new module ‘B’, the two language centres were not only connected anatomically but also via multiple conceptual representations that were spread throughout the cortex. The famous ‘house model’ was created (in line with previous and similar ideas of Wernicke that were discussed in Chap. 2). Lichtheim was able to add four new categories to the three forms of aphasia that had already been dealt with in Wernicke’s model. Roth and Heilman describe these transcortical aphasias as follows (2000):
When connections from Wernicke’s area to the concept center are disrupted, comprehension is impaired, because a semantic analysis of words cannot be performed. However, repetition is spared, because auditory information can access Wernicke’s area and be transmitted to Broca’s area for production of speech. This disorder is called ‘transcortical sensory aphasia’. In contrast, when connections between the concept center and Broca’s area are disrupted, internally generated speech (spontaneous speech or naming) will be halting and effortful, as in Broca’s aphasia. However, because Wernicke’s arc is intact, repetition is normal. This type of aphasia is called ‘transcortical motor aphasia’. 
Premium Wordpress Themes by UFO Themes
In 1877, 8 years before Lichtheim, Adolph Kussmaul (1822–1902) also presented a patient with a transcortical sensory aphasia (i.e. with a comprehension disorder but intact repetition). His explanatory model resembled that of Lichtheim but differed in the fact that the ‘concept centre’ was only accessible via the phonological lexicon (i.e. Wernicke’s area). Because of this, his model was unable to explain transcortical motor aphasia. But Kussmaul’s ideas on language went beyond that of single word processing; he also considered the lexical and the sentence level and introduced the term ‘agrammatism’ to describe impairments in grammatical formulation. The diversity of symptoms in aphasia had also been recognized by other authors. Bateman (1870), for example, pointed to the fact that aphasias can differ in their degree of severity, and deficits can be modality specific [4, 7]. Furthermore, he noted that speech with an emotional content (‘automatic speech’) can be preserved in aphasic patients and that in multilingual patients languages can be selectively affected. Both Kussmaul and Bateman emphasized the heterogeneity of aphasic phenomena and deliberately refrained from anatomical localizations as they believed that language covered an ‘enormous association area’ in the brain (Fig. 4.2). Kussmaul considered it naive to think that there is a set of language in any particular convolution of the brain and wrote that: ‘Wernicke made the mistake of plotting the centre in specific areas of the brain. The localization of elementary functions of language is not mature enough for this’ . This criticism is not entirely justified, however. We have seen in Chap. 2 that Wernicke proclaimed an associationist view for cognitive functions and considered the entire perisylvian area to be involved in language processing.
Kussmaul’s language model (1877) for word processing. Although the model resembles that of Wernicke, Kussmaul explicitly omitted any reference to brain anatomy. He considered ‘language’ far more complex than word processing alone. d route for speech, r route for writing, B centre for sound images (cf Wernicke’s area), B′ centre for optical images, B″ centre for optical images for (deaf–mute) lip-readers only, C centre for coordination of sound movements into spoken words (cf Broca’s area), C′ centre for writing, J concept centre (Figure taken from Tesak and Code, 2008 )
The Wernicke–Lichtheim model had its flaws and, for instance, could not explain anomic aphasia. Many others adapted or augmented this model, sometimes with reference to specific brain areas. Variations were proposed among others by Bastian (1887), Charcot (1889), Dejerine (1891, 1892), von Monakow (1905), Henschen (1922) and Kleist, to name some of the more famous researchers (see Figs. 4.2, 4.3, 4.4, 4.5, 4.6 and 4.7) [4, 9–11]. It must be said that some of these diagram makers acknowledged the limitations of their models (and frankly admitted that they were best used for teaching or heuristic modelling), but many others seemed to believe that clinical language disorders could be deduced from their ‘box-and-arrow’ models.1 This debate, in fact, is still continuing today. One of the main questions that remains to be answered is to what extent individual language maps (either healthy or diseased) differ from group language maps. The fact that even normal brains can significantly differ in size and shape (e.g. the pattern of gyri and sulci) obviously limits generalization of results into detailed anatomical-based models; see Chap. 5 for a detailed account. Additional difficulty is that brains are constantly changing their functional configuration because we never stop learning, or adapting otherwise to our environment. Under pathological conditions, functions can reorganize to such an extent that the resulting anatomo-functional configuration becomes significantly different from that of healthy subjects . This pathology-driven process of reorganization seems particularly effective in slowly growing brain tumours (Fig. 4.9).
Bastian’s language diagram (1887) for word processing and corresponding anatomical sites and pathways. Note the absence of a concept centre, which was rejected by Bastian. The model has visual (V) and auditory (A) word centres and two sensorimotor centres that control kinaesthetic information from the tongue and hand muscles (hence, their respective names: glosso-kinaesthetic (GL.K.) and cheiro-kinaeshetic (CK.K.) centre (Figures taken from Head, 1926 )
Charcot’s localization of language centres and aphasias (1889). (Top) Charcot’s famous ‘bell’ diagram which is essentially the same as the model of Kussmaul. Again, there are four memory centres which are connected to an association centre. (Bottom) Localisation of aphasias. Some authors, for instance, Bogen and Bogen (1976), have argued that Charcot included T2 in his language regions . IC association centre, CAC general auditory centre, CAM hearing centre for words, CLA centre for articulated speech, CVC general visual centre, CVM visual centre for words, CLE centre for writing, 1 Sylvian fissure, 2 Rolandic fissure, F1/F2/F3 first/second/third frontal gyrus, T1/T2/T3 first/second/third temporal gyrus, O1/O2/O3 first/second/third occipital gyrus, Ps superior parietal lobe, Pi inferior parietal lobe (Figures and legends taken from Tesak & Code, 2008 )
Dejerine’s work became important because of his descriptions of isolated disorders of reading (word blindness or alexia) and writing (agraphia). In 1891 he described a case of alexia with agraphia due to a lesion in the left angular gyrus . In 1892 he described a 63-year-old man with pure alexia (i.e. without agraphia) due to a stroke. The patient died 5 years later due to a second stroke that affected the left angular gyrus (and led to paraphasia and agraphia). This case is shown in the figure and discussed here further. Post-mortem examination revealed that the old lesion (which had caused the reading disorder) occupied ‘the [left] occipital lobe, and particular the circumvolutions of the occipital pole, starting at the base of the cuneus, as well as those of the lingual and fusiform lobules’. Dejerine argued that the extensive destruction of white matter in the left occipital lobe had destroyed the connecting fibres from the right occipital lobe to the language areas necessary for reading in the left posterior (and inferior) temporal lobe. This area is sometimes called the visual word form area, although its existence is disputed . Top figures show the old and newer lesion (dark and stippled areas, respectively). For in-depth description, see works by Geschwind  and Dehaene . (Bottom figure) Four years later, Dejerine (together with Mirallie) described another type of alexia (‘third alexia’) which can occur with Broca’s aphasia in frontal lesions . His explanation for this alexia was somewhat different and originated in his concept of a ‘language zone’. Within this zone there are specialized cortical centres that are functionally integrated. According to Tesak and Code (2008): ‘Cortical lesions in the language zone lead to a disorder of “inner speech” and create supramodal disorders such as alexia in motor aphasia’ . Although Dejerine’s model of pure alexia has strong roots in connectionism, he also moved away from Wernicke’s views when he assigned a specialized role to the angular gyrus in the visual representation of words. For Wernicke, higher functions were the product of connections, not cortical areas (Top figures taken from Geschwind, 1962 ; bottom figure taken from Tesak and Code, 2008 )
The borders of the language regions in von Monakow’s model (1905) are deliberately blurred. He was convinced that language regions could not be drawn with lines, as they probably abated gradually in all directions into the neighbouring gyri, and advocated a wide perisylvian language region [4, 11]. von Monakow was the first to describe the phenomenon of diaschisis, whereby a lesion could affect remote areas via its long-range connections. Part of the functional deficits could therefore be explained by distant effects, and as such he argued against strict localism. The figure shows language areas (dark) and ‘relative fields’ (light) (Figure taken from Tesak and Code, 2008 )
Kleist (1934) was an ‘extreme’ localist. His detailed maps were based upon symptom–lesion studies in hundreds of brain-damaged soldiers from the World War I and more or less followed the cytoarchitectonic maps of Brodmann. He located sensory aphasias to Brodmann areas 42 (perception of speech sounds), 22a (understanding melody and tone) and 22b (understanding speech; understanding phrases). Motor aphasias were located to Brodmann areas 44a (singing), 44b (spoken naming) and 45a (syntactic speech). His maps were so detailed that they recall the maps of the phrenologists. Ironically, Kleist had been one of Wernicke’s assistants (between 1903 and 1905) (Figure taken from Nieuwenhuys, Voogd en van Huijzen, 2007 )
A model is a human construct that aims to predict behaviour of real-world systems. It does so by simplification of various processes. An example of a box-and-arrow model for language is shown in the figure. This PALPA (Psycholingual Assessment of Language Performance in Aphasia) model is based on the assumption that the language system is organized in separate modules that can be selectively impaired by brain damage (Kay 1992) . The more complex models are, the more difficult it is to localize functions, as these are ‘distributed’ across many different boxes and arrows. The visual system of the monkey, for example, can be deconstructed into more than 30 areas and over 300 connections . Strict anatomical localization of function has now become impossible
(a, b) MR images of a 29-year-old patient with a low-grade glioma in the left inferior frontal gyrus (this is a type of brain tumour that typically takes years to become symptomatic). Clinical debut was a seizure; there were no (neurological) deficits. The tumour had invaded classic Broca’s area. The posterior border was the precentral sulcus (small arrow), and cranial border was the inferior frontal sulcus. The large arrow points to the central sulcus. Information from MR tractography in figure (b) is shown with colours. The yellow tract is the inferior fronto-occipital fasciculus (IFOF), which represents the medial functional border of the resection. (c) Intraoperative photograph before tumour resection. Contour of the tumour has been marked with a small cord. Small and large arrows indicate precentral and central sulcus, respectively. Numbered markers indicate sensorimotor and language areas that were found with electrocortical stimulation. Markers 1, 2, 4, 7, and 8 indicate the primary motor cortex. A speech arrest was found at markers 8 and 11. Stimulation at markers 9 and 10 did not yield consistent language errors, and these were not considered critical language areas. (d) Postoperative MR image demonstrated macroscopical complete resection. (e) Tumour resection was performed up to sulcal borders. Asterisk indicates sites where subcortical stimulation resulted in (semantic) language impairments, likely due to stimulation of the IFOF. The patient had transient speech disorders that became clinically manifest on the second day after surgery. Three months after surgery, these had resolved. Such a case demonstrates that the functional and anatomical localization of Broca’s area do not necessarily coincide and strongly suggests that reorganization of function took place prior to—and possibly also after—surgery
WordPress theme by UFO themes