Abstract
In 1880 Carl Wernicke gave this plenary lecture at an annual meeting of German physicians and natural scientists. He used principles from his 1874 aphasia monograph to build a neural model of mental illness. He proposed that the brain keeps a record of experiences in distinct areas of the sensory and motor cortices in the form of memory images, which allows for recognition of objects and the planning of motor acts. He conjectured that imperfections, partial defects and complete loss of such memory images lead, respectively, to mild, moderate and severe forms of psychopathology in sensory and motor realms. The lecture is an early presentation of Wernicke’s system of psychiatry. Several of his concepts have remained relevant in contemporary neuroscience.
Wernicke’s project for a scientific psychiatry
Carl Wernicke (1848–1905) was a 32-year-old physician in private practice in Berlin when he delivered the plenary lecture ‘Ueber den wissenschaftlichen Standpunkt in der Psychiatrie’ at the 53rd Convention of the Gesellschaft Deutscher Naturforscher und Ärzte (GDNÄ) in Danzig in September 1880. The GDNÄ, founded in 1822, was the most prestigious professional society for German physicians and convened annual meetings. In 1846, a psychiatry section of the GDNÄ was created, which organized sessions at the annual meeting, mainly devoted to asylum psychiatry and neuroscience, primarily anatomy and pathology (Schmidt, 1982).
Wernicke’s plenary lecture at the 1880 meeting was published after the meeting (Wernicke, 1880a), and in the same year it was reprinted, with several corrections, as a monograph (Wernicke, 1880b). Simultaneously, the lecture was published in five medical journals in Breslau, Vienna and Prague, ensuring a wide distribution in German-speaking countries at the end of the nineteenth century (Wernicke, 1880c, 1880d, 1880e, 1880f, 1880g). The Classic Text below is a translation of the monograph (Wernicke, 1880b); any variations from the text of the lecture (1880a) are mentioned in our notes.
Wernicke before the 1880 lecture
Wernicke grew up in Silesia, which became a Prussian province in 1815 and subsequently part of the German Empire in 1871. He attended high school and then medical school in the capital of Silesia, Breslau (now Wroclaw and part of Poland). He completed medical school in 1870 and began postgraduate training in ophthalmology. After serving in the Franco–Prussian war (1870–1), he returned to Breslau and, by 1874, had completed his medical training.
As a medical resident, he was taught by Theodor Meynert (1833–92), the inaugural chair of psychiatry at the University of Vienna. During a six-month stay in Vienna, he encountered Meynert’s speculations about the human brain, some of which Meynert presented at the 1873 meeting of the GDNÄ, the society that would invite Wernicke to give a lecture seven years later.
After his return to Breslau, Wernicke published Der aphasische Symptomencomplex. Eine psychologische Studie auf anatomischer Basis in 1874. This brought him instant fame, and in 1876 he took a position at the Charité hospital in Berlin. But after just two years, conflicts with his superiors forced him to leave and he then worked in private practice in Berlin for several years (Geschwind, 1963; see also Engstrom, 2004).
The 1880 lecture
Wernicke delivered the plenary lecture at the GDNÄ convention during his break from academia. At 4,800 words, it is of modest length and is divided evenly into two parts: in the first half, Wernicke reviews his aphasia model; in the second, he articulates a new vision for psychiatry.
Here we review three aspects of his lecture: 1) aphasia as a new model for the understanding of brain function; 2) neural models of memory images (Erinnerungsbilder) and representation of movement (Bewegungsvorstellung); and 3) speculations about neural mechanisms of psychosis.
Aphasia
In the second half of the nineteenth century, scientists and clinicians in England, France and Germany/Austria accumulated overwhelming evidence in support of the idea that distinct areas of the cerebral cortex gave rise to sensory experience and motor behaviour. While the initial evidence was crude, the evidence became increasingly more specific and detailed (Young, 1990; Zola-Morgan, 1995).
In his lecture, Wernicke gives credit to early pioneers of aphasia research (Flourens, Bouillaud and Broca), who had demonstrated the importance of frontal lobe regions for mental processes in general and the ability to speak in particular. But he emphasizes the discoveries of his teacher Meynert, who used novel techniques of brain dissection to show that sensory information is sent via ‘fibres . . . comparable to conductive wires’ (p. 7) to specific areas of the cerebral cortex, ‘leading sensation in a centripetal direction’ (p. 7), and ‘another area conveys the connection of the brain to the musculature, leading the motor impulses in a centrifugal direction’ (p. 7). Wernicke concludes from Meynert’s research that language requires two distinct centres in the human brain: ‘one centre would have to contain the representation of speech movement, and the other, the sound images of the spoken language of others’ (p. 9). During the lecture, Wernicke displays this flow of information processing with arrows, superimposed on a drawing of the lateral convexity of a human brain (the figure was included in the monograph, but not in the various journal publications).
Wernicke (1874) had already proposed, in his aphasia monograph, that afferent/efferent projections provide the centripetal/centrifugal flow of information. In the 1880 plenary lecture, he now extrapolates from a specific model of language function to a more general model of sensory/motor function in the human brain. Wernicke proposes that ‘the memory images of movements, the ‘representations of movements’ (Bewegungsvorstellungen), are indispensable for the understanding of movement’ (p. 10). He boldly claims a new model of human brain function: ‘The analysis of aphasia therefore gives us the paradigm for all mental processes with a concrete content, insofar as certain grouped memory images make up our entire intellectual property, the entire content of our consciousness’ (p. 11).
In order to appreciate the second half of Wernicke’s lecture, in which he uses his novel information processing model to explain clinical features of psychiatric illness, we need to gain a better understanding of his terms ‘memory image’ and ‘representation of movement’.
Memory images in the sensory and motor realms
In his lecture, Wernicke refers to memory images as a ‘very specific psychological term’ (p. 10). His audience might have appreciated the rich history of the term ‘memory images’ (Erinnerungsbilder) in nineteenth-century German philosophy and psychology. However, he does not mention any of it in his lecture, so we provide a brief summary here.
Johann Friedrich Herbart (1776–1841), who succeeded Immanuel Kant as chair of philosophy in Königsberg, asked: how can the human mind reproduce past sensory impressions after the disappearance of a stimulus? In his Psychologie als Wissenschaft (Herbart, 1824), he proposed that sensory images create memory images, which decrease in strength over time, but can be enlivened by a new sensation that matches the memory image (Banks, 2005; Ziehen, 1905). While Herbart’s ‘mechanics of mind’ includes many mathematical formulas, he conceptualized psychology as applied metaphysics, devoid of empirical data (Eckardt, 2010).
This changed with Gustav Fechner (1801–87), an early pioneer in experimental psychology and the founder of psychophysics. Fechner built on Herbart’s concepts in his model of memory, but now included experimental data (Phelps, 2016). In his Elemente der Psychophysik, he contrasted the high fidelity of after-images (Nachbilder), generated by sensory images, with the weaker, less accurate and less permanent character of memory images (Fechner, 1860).
Memory images assume a central role in the work of Wilhelm Wundt (1832–1920), often considered the founder of modern psychology (Carpenter, 2005; Scheerer, 1980). Wundt spent many years refining his concept of memory images, often informed by new results from memory experiments. In 1874, Wundt thought that the memory image ‘simply repeats a prior idea’, but is fainter than the original. However, by 1893 he acknowledged that memory images are not replicas of previous events, but signs for them. By now, faintness was no longer the characteristic feature of memory images, but of incompleteness and a special recognitive feeling (Wiedererkennungsgefühl). For Wundt, memory images did not just refer to the object being perceived, but to the cognitive and affective circumstances present during the past experience with the image (Carpenter, 2005).
Some features of the ‘memory image’ concept that Herbart, Fechner and Wundt had developed in the nineteenth century were carried over into the ‘engram’ concept by Richard Semon (Schacter, Eich and Tulving, 1978), which has remained of great interest to memory researchers up to the present (Josselyn, Kohler and Frankland, 2017).
In 1879, Wernicke briefly commented on the extensive philosophical and psychological literature on memory images during his lecture on consciousness at a meeting of the Berlin Medical-Psychological Society (Wernicke, 1879). But one year later, during his plenary lecture at the GDNÄ meeting, he completely ignored it, since he set out to chart new territory: the neural basis of memory images.
He did recognize the contributions of his teacher Meynert, who had adopted the term memory images (Erinnerungsbilder) for his own aphasia case report (Meynert, 1866), published after Broca’s case (Berker, Berker and Smith, 1986) and before Wernicke’s (1874) case series. Furthermore, Meynert had speculated about the neural correlates of memory images in his 1873 lecture Zur Mechanik des Gehirnbaues (Meynert, 1874; Phelps, 2016; Whitaker and Etlinger, 1993). At that time, Meynert was not quite ready to replace Herbart’s mechanics of mind with his own mechanics of brain. Wernicke’s 1880 plenary lecture provided the first mechanistic model for the implementation of memory images in the human brain. This model became central for Wernicke’s explanation of normal and abnormal mental processes (Lanczik and Arts, 2000; Lanczik and Keil, 1991).
Wernicke made two important contributions to the concept of memory images and by doing so moved the discourse from philosophy/psychology to neuroscience. First, he extended the concept from the sensory to the motor domain. What memory image (Erinnerungs-bild) is for the sensory domain, representation of movement (Bewegungs-vorstellung) is for the motor domain. He had taken the term Bewegungsvorstellung from Meynert and had used it already in his 1874 aphasia monograph, where it allowed him to separate motor from sensory aphasias. But in the 1880 lecture, Wernicke went beyond the narrow concept of language function and articulated the broader concept of psychomotor pathology. He expanded on these ideas in his later writings, especially in his textbook, where psychomotor abnormalities are categorized into akinetic, hyperkinetic and parakinetic forms (Wernicke, 1894, 1896, 1900). Wernicke’s student Hugo Liepmann (1863–1925), who provided the original description of the apraxias, introduced a term similar to representation of movement (Bewegungvorstellung), i.e. movement formulae (Bewegungsformeln). He proposed that such formulae are implemented in the left hemisphere and control purposeful, skilled movements (Pramstaller and Marsden, 1996).
Second, Wernicke proposed a neurophysiological mechanism for memory images. Previous neural models of memory images, i.e. Meynert’s and his own, had been strictly anatomical. In the 1880 lecture, Wernicke was ready to advance a physiological model. At a time when the electrical properties of nerve cells, resulting in excitation and inhibition, had not yet been discovered, Wernicke proposed that excitation occurs before memory images are changed or lost. He speculated that, in the sensory realm, such excitation takes the form of hallucinations, and in the motor domain it appears as psychomotor activation/agitation.
In a bold statement, he extends his neurophysiological model of memory images even further, beyond the sensory and motor realms: ‘Irritation of memory images in the brain region that constitutes personality produces grandiose delusions and in it the germ for the destruction of the personality’ (p. 15). We might ask: why grandiose, but not persecutory delusions? What constitutes destruction of personality? Wernicke does not provide any further explanation. It is one of several occasions during his lecture where he speculates about neural mechanisms without much, if any, evidence.
Neural mechanisms of psychosis
After the review of his aphasia model and the introduction of a neural model of memory images in both sensory and motor domains, Wernicke articulates their implications for a scientific viewpoint in psychiatry: ‘Our task will be to determine, by means of observation, the behaviour of the memory images in the mentally ill and to use them to understand the mental state’ (p. 13).
Wernicke suggests that a gradient of memory loss translates into a severity gradient in psychopathology. In the most severe form of mental illness (i.e. general paresis of insane = GPI), memory images are lost: ‘The real essence of this [GPI] consists in the rapid progressive loss of memory images’ (p. 13). He contrasts this with primary Verrücktheit (in 1880, this referred roughly to what we now call non-affective and non-organic psychotic disorders): ‘. . . these diseases do not always involve a destruction of the cell elements which we have to imagine as the physical substrata of the memory images, but often only a pathological change in them’ (p. 17).
Misidentification delusions result when memory images ‘. . . are changed or falsified by some pathological process, then for the afflicted parts of consciousness the congruence between the external world and the image which is deposited in the brain, has ceased’ (p. 18). Anticipating current concepts of psychosis as a dimension, without a categorical distinction from normal mentation, Wernicke (1880b: 19) writes: If we imagine that it were possible to rob a healthy person unnoticed of a large part of the memory images in which the outside world is reflected in him, or to change their content, we would fully understand all sorts of wrongs that he committed in his perplexity.
The 1880 lecture introduced two concepts which Wernicke would go on to develop further in his 1894 textbook and which have become major contributions to the psychopathology literature.
The first was the concept of perplexity (Rathlosigkeit) (Lanczik and Arts, 2000), which has since been recognized as a core feature in the early stage of psychosis (Humpston, 2014: 20): ‘Perplexity is the state of mind in which most mentally ill people remain while they are still curable’ (Wernicke, 1880b: 20).
The second was the interoception/exteroception dichotomy, which has become central for models of consciousness and selfhood (Khalsa et al., 2018). Wernicke (1880b: 16) wrote: For the brain, in our opinion, the body can be nothing other than a group of memory images, which, however, distinguish themselves from the memory images provided by the outside world through their stable, firmly ordered content and their constant recurrence, and thus to a certain extent provide the solid core on which the rest of the content of consciousness must crystallize.
Building on this dichotomy, Wernicke would go on to distinguish three different types of psychoses: 1) allopsychosis, due to abnormal memory images derived from sensory data representing the outside world; 2) somatopsychosis, due to abnormal memory images derived from sensory data originating in the body; and 3) autopsychosis, due to abnormal memory images of the self (Wernicke, 1894). Wernicke’s contrasting of interoception with exteroception anticipated contemporary models of reality distortion, abnormal interoception, and ipseity disturbance (Sass et al., 2018).
Reception of Wernicke’s lecture
Wernicke was not the first to articulate a scientific viewpoint for psychiatry. In 1874, Paul Samt had lectured at the Berlin Medical-Psychological Society on Die naturwissenschaftliche Methode in der Psychiatrie (Samt, 1874).
But Wernicke went much further in his 1880 lecture. He was bold enough to define a neural model of mental illness, by infusing well-established concepts from philosophy and psychology with new findings from neuroanatomy and some conjectures about neurophysiological mechanisms. He went beyond the speculations of his teacher Meynert and provided a novel, more mechanistic model of brain function.
Wernicke died unexpectedly, in July 1905, from a pneumothorax, after a bicycle accident, and in his obituary Theodor Ziehen (1862–1950) concluded: Already in 1880, he had written a treatise on the scientific viewpoint in psychiatry and foreshadowed the path of scientific psychiatry as he imagined it to be, in contrast to the aberrations of contemporary psychiatry. For many years his early departure from Charité and his anatomical and neuropathological work took him away from psychiatry. . . . Wernicke not only discovered sensory aphasia and its localization and thus added a new component to the theory of localization, but the scope of this discovery goes much further. . . . he showed the decomposition of a cortical process into stages: stimulus - sensation - memory image - association of further memory images - and motor projection. (Ziehen, 1905: 3)
Hugo Liepmann suggested that his mentor emphasized localization of brain function, while Kraepelin emphasized course of illness: Wernicke’s element is the spatial. He is interested in what can happen in the brain, in the parts of a spatial organ. Simultaneous relations in a spatial organ are his object, which is why many of his descriptions are cross-sectional images through the temporal sequence of psychic phenomena. Kraepelin’s psychiatry moves more in the temporal dimension: cause at one end and outcome at the other, in between progression to mental weakness, or intermittent progress, etc. (Liepmann, 1911: 30)
But Wernicke’s lecture was not without its critics. In November 1880, within a few months of the lecture, Eugen Hallervorden (1853–1914) published a detailed review. He cited Wundt and Schopenhauer as the original contributors to the memory image discourse and criticized Wernicke for a category error: memory image is a mental process and cannot easily, if ever, be translated into a neural process (Hallervorden, 1880).
Another critique of Wernicke’s lecture was published in 1892 in Allgemeine Zeitschrift für Psychiatrie und psychisch-gerichtliche Medizin (the most influential professional journal in nineteenth-century German psychiatry and widely read by psychiatric clinicians; Engstrom, 2004). This critique concluded that Wernicke’s project – mapping mental states to circuits in the brain – was warranted, but the reviewer expressed concern that not enough was known about the brain to support several of Wernicke’s claims, including the mechanism of delusions (Dittmar, 1882).
More important for the history of psychiatry is the 1891 monograph Zur Auffassung der Aphasien. Eine kritische Studie by Sigmund Freud (1856–1939). He was then a 35-year-old neurologist in private practice in Vienna, and this was his first book as a sole author. He asked pointedly: Is it justified to immerse a nerve fibre, which has been, over the entire length of its course, just a physiological structure and subject to physiological modifications, with its end in the psyche and to equip this end with an idea or a memory image? (Freud, 1891: 56)
Freud concluded, like Hallervorden, that memory images are psychological constructs which cannot be mapped to the cerebral cortex. He proposed to replace Wernicke’s anatomical model, which localized memory images to distinct regions of the cerebral cortex, with a psychological model, which associates words with memory images of objects (Greenberg, 1997; Guenther, 2013; Henderson, 1992; Schoenwald, 1954). Freud’s detailed critique of Wernicke’s project for a scientific psychiatry is now seen as a turning point in his career: he gave up neuroanatomical research and embarked on the study of mental associations (Schoenwald, 1954).
Direction for psychiatric research
During the second half of the nineteenth century, psychiatry developed into a natural science and established itself as part of academic medicine (Engstrom, 2004; Schmitt, 1983). Several psychiatrists, including Griesinger, Meynert and Wernicke, had hoped to put clinical psychiatry on solid ground, by mapping out the neural mechanisms of abnormal mental states. But at the end of the nineteenth century, psychiatrists became increasingly sceptical, asking: is the available neuroscience knowledge sufficient to guide clinical practice? In contrast, behavioural neurologists and clinical neuropsychologists fully embraced Wernicke’s project of mapping cognitive processes to neural networks (Mesulam, 1998; Sutterer and Tranel, 2017).
So far, Wernicke’s scientific viewpoint has not become the leading paradigm in psychiatry. Some have demanded a paradigm shift, not very different from Wernicke’s call in 1880 (Insel et al., 2010; Reynolds et al., 2009). But the majority of clinicians and researchers still favour a clinical model, which starts with diagnostic categories based on signs and symptoms followed over time. In the larger community of mental health care providers, neural models of psychiatric illness are viewed even more critically. If Wernicke returned for another plenary lecture now, would his message be very different?
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.
