‘Febrile Anxiety’,by Robert James (1745)

Earlier narratives on Anxiety

Prominent in current psychiatric listings, the ‘anxiety disorders’ are considered to be a major problem in clinical practice (Nutt and Ballenger, 2003; Stein and Hollander, 2002) and have become a sizeable source of income for the pharmaceutical industry. ‘Anxiety’ and ‘The Age of Anxiety’ (Auden, 1948) have become descriptive and explanatory clichés in literary, social and cultural studies (Barine, 1898; Drinka, 1984; Oppenheim 1991; Pietikainen, 2007; Radkau, 1998). All this might suggest that the anxiety disorders have been around for a long time. In fact, the opposite is the case as their construction as a diagnosable clinical category goes back only to the beginning of the twentieth century.

Before 1900 the very concept of ‘anxiety disorders’ would have had little meaning for alienists (the earlier name for psychiatrists) and for all other medical practitioners. The clinical complaints currently included in the concept of ‘anxiety disorders’ were then considered as independent states, for example, the dizziness of agoraphobia was dealt with by ENT physicians, hyperventilation by chest physicians, tremor by neurologists, nervosism and neurasthenia by general physicians, etc., etc. (Berrios, 1999). Another member of the group, obsessive-compulsive insanity, was looked after by alienists for it would not have occurred to anyone that it might be an ‘anxiety disorder’.

Historiography

Mental symptoms and disorders are constructs put together by means of convergences. The latter are historical processes that bring together names, concepts and selected complaints (behaviours) in the work of a writer who usually acts as the spokesman or amanuensis for a clinical caucus or for a fashion. Each of the three components of a convergence has its own history and may have participated in earlier convergences: names can be recycled words or neologisms; concepts are the expression of theories of disease available at the time and anchor the convergence onto a specific historical period; the behaviours included in the newly proposed symptom or disorder are selected out of many and are believed to constitute a cluster with biological meaning.

Only some convergences are successful. It is conventional to explain their success as reflecting the ‘discovery’ of a new truth. However, social, economic and political explanations are likely to be as important to the success of a convergence as its truth-making capacity. The word ‘anxiety’ has participated in various convergences; in practice this means that in each convergence it has named a ‘different’ disease. As illustrated by the Classic Text that follows, during the eighteenth century the word ‘anxiety’ named a different set of complaints and was associated with a different explanatory concept. This convergence was replaced by the end of the nineteenth century by the one that is still in vogue now.

The history of anxiety

In historiographical terms, a history of ‘anxiety’ since Greek times to the present (a common pursuit by some clinical historians) makes little sense (e.g. Ghinassi, 2010). Of course, it is possible to write the history of each component individually, but by itself none is the history of the convergence under study. The nearest the historian of psychiatry gets to writing the ‘history of the anxiety disorders’ is by chronicling all those past convergences in which the word, concept and behaviours characteristic of the current convergence have participated. Even then, dead ends are soon reached. On account of this, it is far more profitable to explore the reasons and contexts that caused the dissolution of some convergences and the motivations that led to the construction of new ones. What must be avoided at all costs is the temptation of placing convergences along a progress line and considering the latest as the achieved truth. However, contextualizing convergences is not easy. In addition to knowing well the epistemological tenets reigning in a given period, it is of the essence to know about its social and political history and actor-network properties (Berrios, 1994).

Classic Text No. 97 illustrates one of the convergences in which the word ‘anxiety’ participated. It should be illuminating to explain how and why this medical view of anxiety was transformed into the functional, psychologized convergence that followed: that of the ‘anxiety neurosis’. Here there is space to deal only with the relevant epistemological background: (1) the eighteenth-century notion of Neurosis, as originally constructed by William Cullen; (2) the ontological categories used by alienists to characterize mental symptoms and diseases; and (3) the professionalization and economic expansion of Alienism (Psychiatry) which demanded a distinctive language of description and explanation and a broadening of alienist ‘expertise’.

The concept of neurosis

The notion of ‘nervous disease’ appeared during the seventeenth century to refer to conditions believed to be the expression of pathological changes in the ‘nervous system’ (as this was conceived at the time) (Hare, 1991). Conditions like Hypochondria and Hysteria, which until then had had their own somatic locus, were placed under the unifying view of nervous disease proposed by Sydenham and Willis (Beatty, 2012; López-Piñero, 1983). This marked a fundamental shift in the definition of disease (Rousseau, 2004) and broadened the class of ‘nervous diseases’ which in principle could include those disorders whose symptoms could not be directly referred to the brain.

This widening of the notion of nervous disease was to culminate in Cullen’s concept of ‘Neuroses’ (Doig et al., 1993; López-Piñero, 1983) under which asthma, diabetes and other medical conditions were included:

In a certain view, almost the whole of the diseases of the human body might be called nervous: but there would be no use for such a general appellation; and, on the other hand, it seems improper to limit the term, in the loose inaccurate manner in which it has been hitherto applied, to hysteric or hypochondriacal disorders, which are themselves hardly to be defined with sufficient precision … in this place I propose to comprehend, under the title of Neuroses, all those preternatural affections of sense or motion which are without pyrexia, as a part of the primary disease ; and all those which do not depend upon a topical affection of the organs, but upon a more general affection of the nervous system, and of those powers of the system upon which sense and motion more especially depend. (Cullen, 1789: 120–1)

At the beginning of the nineteenth century the concept of Neuroses became popular in France through the earlier translations by Pinel and Bosquillon (Cullen, 1785a, 1785b). Soon enough, however, the development of the anatomo-clinical model (Ackerknecht, 1967) and the gradual abandonment of the Vitalistic and Neuralpathology doctrines that had inspired Cullen led to a gradual attrition of the class of neuroses. Some diseases were removed from this group because they were found to be correlated with localized organ pathology; others because they were reconceptualized as not being diseases of the nervous system. Indeed, the medical specialism now called ‘Neurology’ was formed around Neuroses considered to have localized pathology. By the second half of the nineteenth century the group of the Neuroses was still sizeable and included nervosism, hysteria, hypochondria, obsessional madness, paranoia, melancholia, mania, etc., and other disorders still considered as functional disturbances of the nervous system.

The concept of Neuroses, like all other clinical categories during the nineteenth century, was caught in the debate between the anatomo-clinical and the physiopathological schools of thought. The former explained disease as changes in structure, the latter as changes in function. As the century progressed, the definition of function became more and more abstract, and by the very end of the century the concepts of anatomical lesion (e.g. atrophy) and physiological lesion (e.g. spinal irritation) were widely accepted, and the notion of psychological lesion was beginning to gain some vogue. Neuroses in which no structural lesion could be found became the ideal examples of diseases caused by a physiological lesion (López-Piñero, 1983).

By the early 1890s, in addition to the European functional disorders, a clinical category called ‘neurasthenia’ arrived from the USA (Bergengruen, Müller-Wille and Pross, 2010; Gijswijt-Hofstra and Porter, 2001). Similar to the French notion of nervosism (Borel, 1894) and on account of its imprecise clinical boundaries, it engulfed many of the conventional functional ‘neuroses’ including those which later on were to be included in the new convergence called ‘anxiety disorders’. In a classical paper, Freud (1894/1953: 97) proposed that ‘anxiety-neurosis’ should include:

general irritability, anxious expectation, anxiety attacks, and [somatic] equivalents such as cardiovascular and respiratory symptoms, sweating, tremor, shuddering, ravenous hunger, diarrhoea, vertigo, congestion, paræsthesia, awakening in fright, obsessional symptoms, agoraphobia, and nausea.

Found in various combinations these complaints, according to Freud, resulted from either a ‘grave hereditary taint’ or a ‘deflection of somatic sexual excitation from the psychical field, and an abnormal use of it, due to this deflection’ (p. 97). Freud’s view that the anxiety states should constitute a separate condition received little challenge. Indeed, important alienists agreed with this. Thus, Hartenberg (1901: 699) only took issue with the proposed ‘sexual aetiology’, riposting: ‘Anxiety neurosis originates in the sympathetic nervous system’ … ‘the term anxiety neurosis is useful to differentiate from neurasthenia a distinct group of symptoms that represent a “primary disorder of the emotions” and which can provide an explanation for the development of phobias.’ This general acceptance of Freud’s view may have been due to the fact that earlier works by Axenfeld (1883), Berthier (1875), Borel (1894), Bouchut (1877), Cullerre (1887), Desmartis (1859), etc., had prepared the ground. Although the notions of nevroses and nevrosisme entertained by these authors were wider than the later concept of ‘anxiety neurosis’, they were also narrower that Cullen’s original definition.

However, disagreements with Freud and the others were also recorded. For example, Pitres and Régis (1902: 250) wrote:

During recent years German authors have described what they call anxiety-neurosis. According to Hecker, this disorder would include all the symptoms of neurasthenia – the latter term being now reserved for simple spinal irritation. On the other hand, Freud considers anxiety-neurosis as an independent disorder characterized, in its pure form, by nervous over-excitement, chronic anxiety and anxious attention, attacks of acute and paroxysmal anxiety with dyspnoea, palpitations, profuse sweating … .

Instead, these authors postulated that anxiety-neurosis was:

only a syndrome and hence may be found grafted, whether acutely or chronically, upon any neuropathic or psychopathic personality … It is associated with neurasthenia and melancholia but can also be seen in other neuroses and psychoses … There is, therefore, no independent disorder called anxiety-neurosis. (p. 251)

This debate prepared the ground for current views: on the one hand, the proposal that the anxiety disorders are independent clinical entities, with both psychological and somatic manifestations and resulting from pathology in specific brain sites; on the other, the view that anxiety is a non-specific, stereotyped emotional response found attached to many psychiatric and physical diseases.

‘Neurosis’ and variants continued to be used in Europe in the inter-war period, and each country dealt with it in a different way. For example, it did not appear in the ‘Mental Diseases’ section of the 4th edition of the Nomenclature of Diseases (Royal College of Physicians of London, 1906). Neurasthenia and Hysteria are listed, but as Diseases of the Nervous System (p. 15). On the other hand, the 7th edition of the same Nomenclature (Royal College of Physicians of London, 1948) does show ‘anxiety’ – simple and with physical complaints – but included under Hysteria (p. 84). Ross (1923) was one of the earlier British psychiatrists to publish on the ‘Common Neuroses’ and their treatment.

‘Psychoneurotic disorders: anxiety reaction’ do appear, however, in the first edition of the DSM series (APA, 1952). ‘Psiconeurosis’ was introduced into Spanish psychiatry after World War I (Fernández Sanz, 1921) and its evolution culminated with a major work on Neurosis by López Ibor (1966). In Germany, Schultz (1928) introduced the concept but as a form of personality disorder.

The ontology of mental disease

All definitions and explanatory theories of disease make assumptions about the nature of illness and about the type of object it is supposed to be. This assumption can be explicit and, for example, as part of the definition it might be stated that disease is a material thing, a natural kind that exists in time and space; or even an entity independent from the body of the subject affected by it (this seems to have been Sydenham’s belief, at least it was the way in which it was understood and criticized by Virchow two centuries later; Taylor, 1979: 11). On other occasions, the ontological assumptions are implicit and must be extricated (Malabou, 2012). Nineteenth-century alienists also had their own views, which they mainly borrowed from general medicine, where they had developed first. In general, whether alienists supported an anatomo-clinical view or a physiopathological view of mental disorder, they considered that madness, in the last analysis, was the result of a brain lesion (which they thought might soon be found) and hence considered the mental disorder itself as deriving all its ontology from the lesion in question. Mutatis mutandi, this reductionistic assumption still informs current neurobiological models of mental disorder (Cooper, 2004; Murphy, 2006).

The expansion of the territory of psychiatry

Throughout the twentieth century, the territory of ‘psychiatry’ has steadily expanded. In addition to the original forms of madness (since then renamed as ‘psychoses’), new ‘disorders’ are being regularly added. This expansion has been presided over by factors philanthropic, scientific, social and economic. New regions of human behaviour have been (and are being) medicalized, and pressures for this process come not only from psychiatrists and others who might stand to profit from it but, rather surprisingly, from the consumers themselves.

The expansion of the territory of psychiatry can also be mapped in the way in which psych-iatry services started to be re-organized at the beginning of the twentieth century. Whether these changes were a cause or an effect of the expansion is unclear, and further research is needed. Changes such as the growth of private practice psychiatry and the gradual penetration of psychiatry into the general hospital and general practice exposed practitioners to a new type of psychiatric sufferer. These patients complained of all manner of mental ills, none of which required compulsory admission or in-patient treatment. By the late 1920s this legion of ‘milder’ forms of psychiatric disorder became a clinical problem and was repeatedly debated in medical meetings. The ‘anxiety disorders’ can be found constituting a sizeable part of these new pathologies.

Soon enough, and in terms both of clinical presentation and speculative aetiology (either neurobiological or psychodynamic), distinctions were made between generalized anxiety, the phobias, and panic attacks. The development during the 1950s of drugs called ‘anxiolytic’ (Berrios, 1983, 1996) allowed the claim that the clinical groups in question could also be factored out on the basis of their selective response to one or other of these new drugs.

In DSM-II, ‘anxiety neurosis’ features as a member of a class called ‘Neurosis’ (APA, 1968). For reasons which remain obscure, it was decided that in DSM-III obsessive compulsive insanity and shell-shock should be included into a new omnibus category called ‘Anxiety Disorders’ which also included panic attack, agoraphobia, specific phobias, social phobias, acute stress disorder, and generalized anxiety disorder (APA, 1980). There was no change in DSM-IV (APA, 1994), but again, for unclear reasons, it was decided that in DSM-5 the obsessive compulsive disorders and PTSD should no longer be part of the ‘anxiety disorders’ (which however have acquired a new member: ‘selective mutism’) (APA, 2013).

The Classic Text

The clinical description of ‘febrile anxiety’ that follows (which deals, in fact, with all the forms of medical anxiety as accepted during the eighteenth century) was included by Dr Robert James (1703–76) in his Medicinal Dictionary (Corley, 2004). James was an eighteenth-century doctor (O’Malley, 1972) whose Dictionary, which also included contributions by Dr Johnson (Brack and Kaminski, 1984), remains a monument to medical scholarship. Translated into French as Dictionnaire universel de médecine (1746–48) by Diderot, Toussaint and others, it became an important work in France until the turn of the nineteenth century. (For more information on James and his dictionary, see Berrios, 2013.)