Beyond Sign-Based Diagnosis: Symptoms as Personal Constructions
Interestingly, semiotic theory discerns a third component in its study of the human use of signs: the signified. The signified is the concept or idea that humans connect to their use of signifiers. Signifieds are created as humans try to make sense of the world, attribute meanings, and build mental representations. Whereas the signifier qua material carrier of information and the referent qua object, event, or action are transindividual, this doesn’t seem to be the case for the signified. Signifieds are created by sense-making individuals, and their content depends on the precise way in which a person interprets the signifiers he uses and perceives. In the DSM the signified connected to the signifiers of distress does not receive attention. It is assumed that, above all, the sense connected to symptoms is a common sense: laypersons and trained professionals are believed to have the same signified in mind when they discuss specific symptoms.
The question is whether such focus on the sign-like quality of mental symptoms and the neglect of the dimension of the signified in psychiatric diagnosis is justified. If this is not the case, the study of how individuals use signifiers in making sense of their own suffering should be more central to the study of psychopathology.
An important characteristic of the DSM’s descriptive classificatory diagnosis is that it merely deals with the human being qua observable object. What it disregards is that people also always have a reflexive relation to their own actions and symptoms. The phenomenologist Josef Parnas (2012, p. 120) articulates this as follows: “the human being is both an empirical object and a transcendental subject, that is a condition of possibility for its own study.” An individual living through a difficult period and experiencing a plethora of symptoms is not simply the sum of these problems. He is also always the relation to his own suffering and to the context within which this suffering came to the fore. This reflexive relationship implies that, invariably, people interpret the elements of the distress that they experience. Interpretation determines the sense or nonsense people attribute to their experience of distress, and guides them in qualifying their distress as a problem or not.
For example, research in the general population indicates that many people report anomalous psychological experiences: In a British sample of 1000 individuals Pechey and Halligan (2012) observed that 48 % of the sample reported occasional or frequent anomalous or paranormal experiences. These experiences were significantly associated with the presence of beliefs of a delusional nature, which were also quite common: 39 % of the participants reported strong belief in ideas of a delusional nature, and 25 % reported strong beliefs in bizarre ideas of a delusional nature (Pechey and Halligan 2011). However, such beliefs and experiences are often not reported to cause distress and thus deemed unrelated to psychopathology (Evrard 2013). Similarly, in various international surveys, Beavan et al. (2011) observed that approximately 13.2 % of people from the general population occasionally hear voices. From a purely descriptive point of view, it might be concluded from these studies that large numbers of individuals have plainly psychotic or attenuated psychotic symptoms. Yet, without taking into account the effect of the voice on the voice-hearer, or of the belief on the believer, such conclusions are unjustifiable. After all, most people from the general population who hear voices are not distressed (Kelleher et al. 2010; Van Os 2010). Neglecting the subjective aspect of the individual’s experiences is ill starred as it leads to a one-sided appraisal of mental health problems (Stoyanov et al. 2012). Taking into account the reflexive relation people have to their own symptoms implies that the subject and subjective experience are crucial to what
a symptom is for a person. There is a real danger that if the subject and subjective experience are not explicitly taken into account, especially for individuals that are not particularly communicative or are easily intimidated by psy-professionals, only a partial appraisal of distress is made.
If symptom experience was more central to diagnosis the mere presence of descriptive characteristics would be less of an issue. As Jim Van Os (2010, p. 308) suggests, such an alternative approach to diagnosis “in part becomes understanding the onset of need for care.” The study of human psychological functioning teaches us that that there is quite some variety in how people behave. Yet variation comes in degrees, making some behaviors and mental phenomena more uncommon than others. Certain ways of behaving and relating that resemble the kind of symptoms often reported by patients might seem peculiar, but should not give rise to diagnosis if they don’t provoke distress. Mental states should only be of clinical concern if they are not (or no longer) manageable for a person. This is reminiscent of Freud’s (1913) suggestion that symptoms come in varieties. In his view, most people, if not all, carry symptoms, albeit some suffer more than others. Throughout time, symptoms can obtain a different status. At some moments a symptom might be part of a person’s habitual mode of functioning, working as a livable compromise between incompatibilities the person is living with, such as conflicting desires or incongruities between desires and inhibitions. At other moments, the balance can lean toward the other side: compromises then obtain a pathological quality and start to produce more distress than tranquility (Verhaeghe 2002). Consequently, the important question a diagnostician needs to ask concerns the factors that made the balance keel to the other side, prompting a request for help. By disregarding symptom experience DSM-based diagnosis neglects such questions.
Against this criticism, it could be argued that in the end the DSM-5 does take into account the reflexive relation people have to their symptoms. After all, it asks diagnosticians to pay attention to symptom severity and valence: aren’t these attempts to grasp precisely how a person experiences distress? Perhaps they are. But obviously people’s reflexive relationships to their symptoms is only of secondary importance in the DSM. In terms of the comment made by Parnas (2012), the DSM above all qualifies mental disorders as empirical objects to which all transcendental subjectivity is subordinated. This bears witness to a naturalistic attitude. Indeed, in the DSM the self-reflexive relation is not considered to be constitutive of mental symptoms; it is a factor that may moderate their expression, but by no means organizes them.
Ivana Markova and German Berrios (2009) formulate interesting challenges for such a naturalistic approach (see also Berrios and Markova 2002, 2006). In their view, the structures or realities underlying mental symptoms and subjective complaints are basically unstable. Material objects such as vegetables or tables have a relatively constant configuration. They always consist of the same constituents that have a constant functional relation to one another. The appearance of such objects might be altered by contextual variables, like climatological circumstances for vegetables or designers’ ideas for tables, but contextual influences don’t change the fundamental qualities of the objects. Such is not the case for mental symptoms. The structure of a self-reported mental state, such as “depressed mood,” is fundamentally fuzzy because it depends on the individual judgment of the person experiencing that state. What one person qualifies as a depressed mood is not the same as what another individual refers to with the same word. Indeed, mental symptoms are not simply signs of underlying disorders, but above all personal constructs: “They are constructs in the sense that subjects create sense or construct a meaning out of an inchoate pre-conceptual and preverbal experience. They are personal in that, although social and cultural influences will help their articulation, the experiences themselves are unique to the individual and inaccessible to anyone else” (Markova and Berrios 2009, p. 344). The authors also argue that observed symptoms and behaviors psychiatrists typically take into account when judging an individual’s state of mind cannot be considered as entities with a stable structure: “This time the constructs involved relate to the judgments made by clinicians and we have seen how these will depend on manifold factors (e.g. experience, personal biases and mood) which, being neither fixed nor consistent, will serve to destabilize the structure itself” (Markova and Berrios 2009, p. 346).
Following this viewpoint it could be argued that by a priori separating symptoms from the lived context of an individual, and not asking clinicians to take into account the personal meaning-making processes, the DSM makes a big leap. The DSM assumes that symptoms and aspects of personal experience do not make up a systemic whole and can easily be separated. In line with Markova and Berrios I believe this is deeply problematic as it neglects the personal-constructional side of a symptom. In terms of the DSM, lived experience is only a minor moderator: it affects the magnitude or quantity of the symptom, but leaves its quality unaddressed. The question as to how a symptom functions in the context of a person’s life, and how exactly or why precisely he is affected in the way that he is, is left out of consideration. Therefore, the DSM decon- textualizes mental suffering because it disconnects the symptom from the personal-construction process in which mental suffering is embedded.
In their Cambridge Model of Symptom Formation, Berrios and Markova (2006) don’t merely treat context as a moderating variable, but as a constructive force in the creation of mental symptoms. In their view, symptoms are made up of both brain signals, which cohere with broader neurophysiological processes, and semantic configuration processes. Central to their conceptual model is the idea that symptoms rest on brain signals that undergo mental elaboration, which are eventually conveyed in clinically observable modes of expression, as we see in self-reported complaints (e.g., “I feel sad,” “I hear voices”) or clinically observable behaviors (e.g., catatonic stupor, agitation). The model discerns four pathways, sequentially termed (a), (b), (c), and (d) along which brain signals are expressed.
Pathway (a) describes the trajectory along which self-reported symptoms are constituted. It assumes that brain signals “resulting, for example, from a malfunctioning pathway or site or indeed nonpathological firing” first penetrate awareness and create a vague pre-linguistic, preconceptual state of consciousness called the “primordial soup” (Berrios and Markova
2006, p. 31). Although they don’t discuss this explicitly, it is quite likely that in order to give rise to symptomatic expression, the primordial soup should provoke feelings of unease and malaise. The model says that in a next step the person interprets his state of mind and configures it “in terms of personal, familiar, social, and cultural formats” (Berrios and Markova 2006, p. 31). The configuration process implies a procedure of transformation along which the primordial soup is encoded in a multilayered world of ideas, images, and meanings, which can be communicated. Obviously, this transformation process is complex and influenced by the person’s habitual ways of speaking about experiences and the body, in which personal, familial, social, as well as (sub-)cultural factors can be discerned. In their discussion of the elements that influence the configuration of the primordial soup, Berrios and Markova (2006, p. 32) state:
“Various factors are likely to be important, for example, the social and cultural background of the individual, past experiences of like events, general and personal knowledge and attitudes, contexts in which it takes place, familial styles of response, and so on. In addition, it is likely that the interpretation and articulation of the resultant symptom will be further influenced by individual language capacity, imagination, decision as to whether to express, and, importantly, the effect of interacting with an interlocutor (often the clinician).”
The net result is that, although based on a brain signal, a symptom can hardly be seen as the direct expression of a neurobiological reality.
Pathway (b) in the Cambridge Model of Symptom Formation describes the way in which clinically observed symptoms that don’t give rise to subjective complaints are constituted. Characteristic of these symptoms is that specific brain signals don’t generate a primordial soup entering awareness, but are immediately expressed in behavioral manifestations, which the diagnostician might interpret as signs of an underlying condition. This time the component of interpretation is not on the side of the patient, but on the side of clinician, who must evaluate whether or not the observable traits in the patient’s functioning signal a disorder.
Pathway (c) is used to describe vague unnamed symptoms. Characteristic of this trajectory is that the brain signal gives rise to the primordial soup, but not to the broader configuration process described in pathway (a). In this case, the patient is aware of mental or somatic distress, but fails to make sense of these subjective experiences, that is, name and classify them in terms of relevant culturally and socially sanctioned modes of expression. The clinician in his turn might learn about such distress via vague complaints in which the experience of unease stands to the fore.
Pathway (d) in its turn describes a path leading to symptoms that are not rooted in a disturbed neurobiological signal. This pathway makes clear that some symptoms are secondary in nature and are provoked by primary symptoms that are indeed an expression of brain signals: “some symptoms may be viewed as secondary constructions. For example, anxiety may develop secondary to a frightening hallucination” (Berrios and Markova 2006, p. 33). The idea here is that primary symptoms may trigger an individual’s configuration processes, thus provoking secondary symptoms. The model indicates that especially disorganized psychological functioning and the presence of aberrant images in the mind might prompt the creation of such secondary symptoms.
What is interesting about the Cambridge Model of Symptom Formation is that, contrary to how the DSM approaches symptoms, it underscores the formative and configuring role of context on symptoms, ranging from personal life history, familial influences, and social factors to cultural determinants. The model does not deny the role of neurobiological factors, but considers them as fundamentally intertwined with context variables.
However, on a different note, it could be argued that the conceptual model of Berrios and Markova could be criticized for being too linear. For example, at the level of pathway (a) they assume that brain signals create a primordial soup in consciousness, which is then configured in terms of relevant context variables, thus giving rise to symptoms. What they don’t seem to model is how important feedback loops might be added between the different pathway components. These would shed a different light on causality. In pathways (a), (b), and (c) of the Cambridge Model of Symptom Formation, brain signals make up the first step in symptom formation, which suggests that in the end most distress is essentially neurobiological. By adding feedback loops to the model, causality could be thought of as more circular in nature. For example, from a sociological point of view it might be argued that perceived inequality in a given society is an important cause of mental disorders (Verhaeghe 2014; Wilkinson and Pickett 2009). In terms of the model, this implies that a feedback loop should be drawn between configuration processes and brain signals: perceiving oneself as disadvantaged in relation to others might provoke brain signals that give rise to distress. Another reason why a feedback loop could be added in between configuration processes and brain signals has to do with important individual differences in how individuals make sense of distress. From a psychoanalytic point of view, it can be argued that in dealing with their distress, some people have a strong tendency to negate and repress, while others are inclined to face and analyze their problems and act upon them. These different styles of managing distress likely have a different impact on the distress as well as on the brain signal that initially gave rise to distress: negation and repression have a worsening effect, while facing and analyzing distress are alleviating. Indeed, more broadly, it could be argued that the effect of psychotherapy, which has a profound impact on brain functioning (e.g., Buchheim et al. 2013; Karlsson 2011), might be thought of as a feedback loop going from configuration to brain signal and distress.
Moreover, the Cambridge Model of Symptom Formation assumes that symptoms are always provoked by biological signals: construction processes transform the brain signal that gives rise to distress but in the end, the biological component is the “prime mover.” This is also true for pathway (d), which describes ancillary symptoms that rest on primary, that is, biologically caused, symptoms. Against or complementary to such a naturalizing view it might be argued that symptoms can also have a psychological or psychosocial cause, which then provokes biological irregularities. For example, experiences such as falling in love, losing a best friend, or being disappointed as a parent can provoke psychological processes that are primal in relation to biological disturbances. In these cases the trigger is non-biological, while its effect will obviously be mirrored in the body. As he makes a plea for recognizing psychical causality, Lacan refers to identification (Lacan 1947) and signification processes in relation to life events (Lacan 1959) as potential triggers for setting illness processes in motion. Although starting from a completely different perspective, behaviorists, in their attention to learning processes driven by external stimuli, also point to modes of causality that are not primarily biological (Moore 2001). By adding psychical causality to the pathways of the Cambridge Model of Symptom Formation this model changes. What is thus created is a Systemic Model of Symptom Formation, which assumes that the causal factor could be situated at each level that the model discerns.
However, my main objection toward the Cambridge Model of Symptom Formation is that it only maps the dimensions that determine symptom formation, thus neglecting the processes by which a symptom is constructed. For example, the model acknowledges that individuals configure the primordial soup they experience in terms of personal, familiar, social, and cultural formats, suggesting the symptom obtains “an onionlike structure in which the primordial soup is wrapped in multiple layers of meaning” (Berrios and Markova 2006, p. 31). Yet, how are these “multiple layers of meaning” constituted? And how do “forms” and “styles of talking” produce their constitutive effect on both the primordial soup and the symptom? It is through the act of speaking, thinking, and fantasizing about who one is and what one lives through. In this context, the verbal and pragmatic embedment of mental symptoms in speech acts should be stressed, which is not the case in the Cambridge Model of Symptom Formation.
In de Saussure’s (1916) structural linguistics, a distinction is made between three constitutive dimensions: speech, language, and individual speaking. In his view, speech (langage in French) concerns the rule- abiding game humans play when communicating via linguistic signs. Language is transindividual, and concerns our ability to communicate and reflect in an orderly way by means of lexical and grammatical conventions and rules. Language (langue in French), by contrast, is a specific tongue, or a specific system of linguistic signs and grammatical rules an individual uses to express something. For example, English or French are languages. Indeed, whereas speech refers to the abstraction of human language use, language concerns the specific idiom one makes an appeal to. Individuals assimilate language, such that in a next step they can use it to write, talk, and reflect. Speaking (parole in French), in its turn, concerns an individual’s specific speech acts, which are expressed in a certain way, with specific stylistic and interactional qualities. Whereas language is a collectively shared social product, which no one masters completely and entirely, speaking is an individual’s product, which can be observed in the practical ways language is used. Language is virtual: it can be mapped in dictionaries and other documents, but even when it is not formally documented it exists. Speaking, by contrast, is actually graspable in specific linguistic expressions. Note that in de Saussure’s work, speaking has a rather broad meaning, referring to all kinds of specific utterances, verbal or written.
What is interesting about the distinction between speech, language, and speaking is that it draws our attention to structural and pragmatic aspects of how individuals use linguistic signs. Speech and language make up structural frameworks individual speakers rely upon. Specific speech acts concretize and actualize the general systems of speech and language, using these as a treasure trove from which elements are selected at the moment an individual speaks. In his work, de Saussure describes speech as a general ability humans have, and language as a social construction and a specifically shaped tongue that a community of speakers relies upon. In his view, language is strongly synonymous with the different languages (e.g., French, English) we discern, meaning that the collective level he concentrates upon is the language community in its very general form. Yet language communities might be thought of as far more local: ICT specialists in Mumbai use English in different ways than inhabitants from a small Scottish village; college graduates use language differently compared people with lower schooling; subcultures employ different slang; and families have their own habitual narrative styles. Indeed, next to cultural, social, and familial contexts, each individual has their own style, which is determined by life events, learning processes, and memories. Indeed, ultimately each individual makes up a language community that is composed of different internal voices, and these styles can be mapped. When language communities are compared, we map general patterns, pointing to structural differences and similarities between the units being compared.
In the Cambridge Model of Symptom Formation the aspect of configuration seems to refer only to such structural and stylistic characteristics.
Consequently, the model strongly presents the symptom as a product of multiple contextual influences (personal, familial, social, cultural), which at the level of language make up different language communities. Yet, perhaps the speech act itself is the most important context in which the symptom needs to be situated. Symptoms cannot be seen as fixed things. Indeed, in their later work Markova and Berrios (2009) stress that symptoms are personal constructs. If this is the case, the way symptoms are constructed in the process of speaking should be more central in how mental symptoms are modeled. When an individual complains of depression it doesn’t suffice to get hold of the social, cultural, familial, and individual meanings she attributes to such a state. We principally gain knowledge of her depressive mood because she talks about it. Thus it is in the speech act that the symptom is articulated and defined. However, the exact form of the articulation will contain variability, such as subtle thematic variations, temporal shifts in style, or vaguely contradictory claims. Apart from this, the speech act also expresses non-verbal characteristics, including intonation, hesitation, and scansion. In the speech act some narrative lines about the symptom might be particularly well elaborated, while others are poor and ambiguous. Moreover, speaking is not only the activity through which the symptom is expressed, it is also the process through which it might be challenged, questioned, and redefined. Indeed, just as neurophysiological and configuring processes modulate the characteristics of the mental symptom, speaking also shapes its characteristics.
Therefore, as an alternative to the Cambridge Model, I present the Triangular Model of Symptom Formation (see Fig. 3.1). This model proposes that the symptom is best conceptualized as an outcome of three systemically connected processes: the act of speaking, within which meaning arises and non-verbal markers are expressed; neurophysiological processes, which might give rise to an explicit occurrence of unease and malaise, or not; and contextual influences that provide a general framework within which experience is configured. At the level of these contextual factors we situate the structural impact of language on mental functioning. Within this model, the symptom is a multidimensional product with certain speech-act specific qualities, biological characteristics, and contextual configuration features covering characteristics that are specific of an individual, family, social context, and (sub-)culture.
Fig. 3.1 Triangular model of symptom formation
-  Sometimes the DSM’s focus on description is characterized as “phenomenological” (e.g., Decker2013). This is confusing since the phenomenological approach to psychopathology, as illustratedby the quote of Parnas, largely exceeds the DSM-like focus on description.
-  Interesting psychometric challenges for the DSM approach to mental symptoms can be found inthe work of Denny Borsboom and colleagues (e.g., Borsboom and Cramer 2013).