Abstract
This text develops three interwoven issues: first, a succinct comparative analysis of medical and psychiatric semiology, which proposes that the lack of referring relations between psychiatric symptoms and brain/psychic dysfunction is a fundamental distinction between medical and psychiatric semiology. Second, the multiple features of psychiatric semiology are reviewed. Third, a new approach to psychopathology is introduced, proposing three different ways to shape symptoms (perception, linguistic structure, praxis); highlighting its role as a cognitive activity that creates intelligibility from undifferentiated experiences; and distinguishing psychopathology and semiology on an activity/product relation basis.
An attempt at comparative semiology 1
The birth of medical semiology
This section will draw on Laennec’s work and its influence on subsequent medical and psychiatric semiology. It may be difficult today to have a clear grasp of the reach and depth of the revolution which broke out in Paris during the first third of the nineteenth century, but it changed Western medicine forever. Canguilhem (1997) goes as far as to consider the conjunct work of Laennec, Pierre Louis and the other physicians of the École de Paris as the origin of a new kind of medical science to which we remain heirs.
Laennec’s classification sets apart signs and symptoms and subdivides each category into two (Laín, 1997, 1998). Symptoms are the alterations of biological functions secondary to any given injury; they may be physical or moral (emotional). Signs are those items which allow the clinician to infer the existence of an injury; they may be physical (i.e. found after a doctor’s manoeuvre on the patient’s body, such as Murphy or Blumberg sign) or vital (spontaneously present, such as jugular pulsation). It is important to note that, from this point of view, signs and symptoms share the same kind of correlation with biological damage. Symptoms are just too unspecific and ambiguous. They belong in the same semiotic family and are the ‘poor relation’ of physical signs, but related nonetheless. The definition of ‘sign’ given in Landré-Beauvais’ Séméiotique (1818: 3) is: ‘Le signe dans son essence est une conclusion que l’esprit tire des symtômes observés par les sens;[…] le signe appartient plus au jugement et les symptômes au sens.’ In other words, some plus is exerted on common material by some intellectual activity to turn symptoms into signs. This should not be understood as a refutation of the traditional sign/symptom distinction, rooted in Boerhaave’s works (Laín, 1997) but as a first-hand hint at the elucidation of some structural analogies.
Canguilhem (1997) highlights three characteristics of physical signs: they are produced by a doctor’s manoeuvre; they may be repeated and counted (‘present or absent’) at will; they need new codifications which may correlate them to subjacent injuries. This new codification was achieved in two steps. The first step, taken by Sydenham, established a semiological thesaurus still unconnected to pathology (Foucault, 1966: 16–41). The second one needed Bichat’s conception of ‘tissue’ and the study of damaged tissues carried on in dissection halls. The connection between these two surfaces (sign and tissue) helped to build modern medicine: signs could announce injuries in the living body, and thus become the foundation of prognosis and therapeutics.
For the purposes of this essay I will bear in mind two main features of medical semiology: (1) semiotic analogy between signs and symptoms, on the basis of (2) a common relationship with injury. This vertical, causal relation between signs and injuries, whereby an injury produces a sign, will be called basic medical semiotic relation.
Psychiatry and semiology
Modern psychiatry was born into the project of establishing a specific semiology cast in this new mould. A good number of intertwining factors was needed. These are, according to Berrios (1984, 1996): the need for descriptive tools for use by the new asylum officers; the availability of psychological theories which might support stable descriptions; the changing notion of sign and symptom in medicine; the introduction of subjective symptomatology; the use of time as context; and the development of quantification.
Let us focus on two of these. New psychological theories played a double role: an up-to-date version of associationism tore apart the somewhat monolithic pre-modern concept of madness. Faculty psychology reassembled the units into a feasible new order. It is debatable whether both psychological theories are to be understood as a context of discovery or as a context of justification. While Berrios seemed to suggest the first hypothesis, Lantéri-Laura (1994) has argued for a less conspicuous role of psychology: a way to tidy up the semiological thesaurus established through clinical work and for clinical practice, rather than a tool needed for the discovery of signs. We need not join in this discussion to observe how associationism managed to break large categories of pre-modern pre-psychiatry into units of analysis that could be regrouped into packets of disturbance which followed the lines of the faculties of the mind: judgement, emotion, will. In doing so, psychiatry entered a complex epistemology which made human beings both the subject and the object of research and made a double bind with human and natural sciences which I will unravel later.
The introduction of ‘subjective experiences’ as legitimate semiotic material has been considered the single most important contribution of the nineteenth century to psychiatric semiology and the source of many of its special characteristics (Berrios, 1996). The intimate side of experience (the vécu) is not directly accessed but is reported by the subject of the experience. But the subject himself gets to know it (makes an intelligible experience out of the vécu) through language. Language, of course, is a public tool, with public meanings. How can anyone tell a psychiatrist about Wahnstimmung, if the very experience of pre-delusional states cannot be allocated into language, for language was never expected to be able to talk about such things? Expression is public, of course, but even more in need of interpretation than language itself. Matching, contrasting, assigning importance to either the public or the intimate sides of pathological experiences will profoundly affect the approach to psychopathology. Paying special attention to this specific semiotic material will bring about interesting novelties: the dialogic exploration of mental states which is now common; the redefinition of mania, melancholy, stupor, paranoia; and the ‘psychologization of the old notion of consciousness’ (Berrios, 1996: 23) In short, nascent descriptive psychopathology emerged as a cognitive system that, through the analysis of behaviour, speech and private experience, and its regrouping into units along the lines provided by faculty psychology, tried to place the province of psychiatric semiology within the realm of general medical semiology (Berrios, 1984). The meaning of each term depends both on the targeted behaviour or experience and also on the other terms of the system. The rules of the language of psychopathology will include both the observer’s judgement of intensity, duration and experiential quality of the symptom and a number of allowed symptom associations. In this text, I shall concentrate on a microanalysis of the manifold ways psychopathology brings intelligibility to abnormal experience, speech or behaviour, and I shall not discuss the influences of economic, social and institutional influences on the development of psychopathology. These forms of logic within psychopathology work as a sort of membrane that both filters and drives the macro forces listed.
I will now describe two important features underlying medical and psychiatric semiology: the heterogeneity of semiotic material (intimate experience, overt behaviour, language, expression) and the influence of the medical notion of sign.
Medical and psychiatric semiology in comparison
Probably the most useful, theoretically sound reconstruction of both medical and psychiatric semiology is that of Lantéri-Laura (1994, 2003).
To distinguish between signs and symptoms, Lantéri-Laura (1994) employs de Saussure’s well-known distinction between signifier and signified. The ‘face signifiant’ (signifier) is built upon the results of percussion, inspection, etc., and is organized through differential traits which oppose ‘dullness’ to ‘resonance’ or ‘pain under pressure’ to ‘pain when relieving pressure’. The ‘face signifiée’ (signified) consists of items such as ‘thickening’ or ‘effusion’. Dullness, for instance, is polysemic. It may mean either thickening or effusion. This polysemy is reduced through concurrent signs and symptoms. Lantéri-Laura highlights this structural property of semiology, the relations of mutual determination entertained by signs and symptoms. Although he does note as the main feature of any sign whatsoever ‘being instead of something else’, he somehow underestimates the outreach of causal relationship, on the basis of the unknown pathology of well-established syndromes. This is true, of course, but it takes semiology back to its previous ‘Sydenham-state’ of lack of connection with damaged tissues. The revolutionary nature of the new conception of medical sign lay in its power to reveal injuries and secure treatment and prognosis. Lantéri-Laura (1994: 423–64) himself comes close to this fact when he addresses the canonic model of illness established by Wirchow, which ties together sign, injury and organ. Unfortunately, the linguistic model of semiology chosen as a tool for the analysis is bound to pay more attention to differential traits than it does to vertical relationship (which in language is either non-existent or inessential).
However, Lantéri-Laura (1994) describes several features of medical semiology whose importance exceeds this shortcoming. It becomes clear in his work that medical semiology is a cognitive procedure which demands: (1) a database of theoretical and practical knowledge, which secures correlation, and (2) a synthetic activity which puts together the findings of physical exploration, their natural history and the other signs present or absent.
This synthetic activity will finally be shaped as a clinical judgement. It needs two parallel epistemic manoeuvres. First, the body of the patient will be solely thought of as a collection of assembled organs, and its expressive, erotic, aesthetic features will be bracketed. Second, within the discourse of the patient only the presence or absence of the items of the semiological thesaurus will be attended. The empirical field of medical semiology is framed by these two epistemic reductions exerted on the body and on the discourse of patients.
Interestingly, Lantéri-Laura (1994: 423–64) argues that these two manoeuvres cannot be accomplished in psychiatric semiology. Both the expressive features of the body and the whole syntactic, semantic, rhetoric or pragmatic characteristics of discourse are materials with semiological relevancy. In a sense, epistemic reduction of the body to the organism may be achieved, but only after its expressive, aesthetic, etc., dimensions have been explored (as in differential diagnosis of mannerism and tics, or as in the relationship that schizophrenic, anorectic or hypochondriac delusional patients entertain with their bodies).
On this basis, four differential traits are more easily grasped. First, both verbal and non-verbal aspects of discourse are dealt with. Second, the context of evaluation may affect the expression of a symptom, and may even shape it (Marková and Berrios, 1995). Third, and above all, semiotic material increases its heterogeneity, including both simple phenomena, akin to apraxia, complex behavioural/experiential phenomena such as delusion or depersonalization, coping/reactive phenomena (i.e. coping with some primary symptomatology, e.g. some ‘social withdrawal’ protects the psychotic patient from cognitive overstimulation, or anxiety leading to delusion which increases anxiety) and personality traits or styles of experience and behaviour. Finally, vertical semiotic relationships are sought after, but, with the well-known exceptions (syphilis and pellagra, which, anyway, have not established semiological correlations but managed to bypass its absence), are neither found nor needed.
So far we have established a few differential traits of medical and psychiatric semiology. Close examination of basic medical semiotic relationships will need a closer examination of the nature of signs.
What is a sign, after all?
A definition of sign close to canonical status may be this one proposed by Eco (2000: 83–5), which I will paraphrase:
Sign is the place where two mutually independent elements meet. These elements belong in different systems and are associated through a codifying relationship which places one of them instead of the other (abdominal pain increased while relieving pressure for peritoneal irritation). The elements correlated may be called ‘expression’ (pain) and ‘content’ (irritation). If both elements share [what Eco calls] ‘matter’, the items are homomateric, if not (as in medical semiology pain: injury) heteromateric.
The correlation mentioned above is a function which goes from ‘content’ to ‘expression’ (if we follow the direction of the cause) or from ‘expression’ to ‘content’ (if we follow the direction of discovery). The nature of the correlating function varies. It may be causal, of course, but it may also be conventional or metonymic or metaphoric and so on. Whatever the type, this synthesis between different elements carried along in signs is always contingent, historical and cultural. This does deny the objectivity of causal relations, but stresses the fact that semiology needs extra components which ensure explicit correlation: database, theory, institutions.
Another feature of general semiotic function, as proposed by Eco (2000: 21–2), is that it does not need to be truthful to have meaning: ‘semiotics allows lying’. The very nature of semiotic function (synthesis between elements which become, in virtue of the relationship, expression and content) makes it possible to posit a member of the function which is not known and which may be non-existent.
The success of medical semiology arises from the fact that choosing an item and denoting it as a ‘sign’ can usually be accompanied by another pole, a referent, from another system, as in the example of pain and irritation given above. However, brain injuries or dysfunction which may be putatively considered referents for signs in psychiatry, such as speech acts, are either too scattered around the brain or rather unspecific. This does not mean there are no such referents, which is after all an empirical matter. But in the current state of knowledge the meaning of psychiatric signs is not backed up by empirical evidence. Therefore psychiatric semiology must rest on some other basis than medical semiology.
The non-referring nature of psychiatric semiology
If psychiatric semiology cannot establish well-defined semiotic relations with specific brain states or mental functions, all we have left (all that psychiatrists have historically worked with) is experience (patients’ experience, psychiatrists’ experience), behaviour and language (patients’ speech, psychiatric semiology). Cognitive meaningfulness and consistency have thus been reached working inside a tricotomic structure which includes:
- Items of experience, behaviour or speech.
- Semiological thesaurus.
- Psychiatrist who employs the words and rules of the thesaurus and the reconstruction of the function from item to context to render intelligible pathological experience, etc.
What is this function from item to context? Let us summarize. If we deprive psychiatric signs of injury/dysfunction as referents and sources of meaning, we are necessarily pushed within the realm of experience, speech, etc. This is another differential trait of psychiatric semiology: its basically homomateric nature. Here we find horizontal, co-determining relations, entertained between elements of similar nature (behaviour, speech, experience) rather than heteromateric (speech and inflammation). Thus, psychiatric semiology will have to take full advantage of a series of procedures (which will be reviewed below) to reduce the polysemy of items of experience without the help of a semantic anchor as in injuries or visible pathological changes.
Furthermore, psychiatric semiology displaces the relation of signs between sign-token and brain/psychic injury to particular disturbances of speech and behaviour and the definitions used to identify them. Thus, the relation between sign-types and sign-tokens becomes crucial. Eco (2000: 276–8) splits signs into two categories: those in which signs include all necessary and sufficient conditions to identify token signs and those where this is not so. The first type is called ‘ratio facile’, the second one, ‘ratio difficile’. In medicine, causal relations exert a constraint in the features of pain or rigidity which must be taken into account in sign-type definition and sign-token identification. This semantic/pragmatic filter enables medical semiology to propose as sign-types traits those causally related to injury and, in case of doubt, check whether sign-tokens were correctly identified. In short, in medicine, sign-tokens refer to organ damage and sign-types are mapped without much trouble onto sign-tokens.
In psychiatry, this process is absent; sign-type definitions and putative sign-token exemplars are all we have to work with. Unfortunately, we cannot assign any particular token to its type if some semantic or pragmatic ‘saturation’ processes are not carried out. This trait is called semantic incompleteness, and has been well established in the semantics of natural languages (Recanati, 2004). (‘Table number three wants a sandwich’ is contextually disambiguated (completed) so that no one can think that the waiter believes tables harbour desires and intentions.)
In order to create the necessary correlation of sign-types to tokens, the features of the sign-type definition are located on a matrix of other information including certain of its own features together with several layers of contextual information. These traits that belong to the token are known as ‘individuation conditions’ (Rejón, 2012, 2013). They include:
- Intensive individuation conditions (IIC) such as the intensity, persistency and frequency of the features included in the sign-type definition and the balance between them.
- Contextual individuation conditions (CIC), such as:
Relations between the content of the utterance and its expressive qualities, where expression works as a micro-context to content.
Other symptoms present or absent.
Patient’s biography and premorbid personality.
Meaning background of experience, as it has been analysed by Husserl, Merleau Ponty, Wittgenstein, Waissmann and Searle (I will not enter into the differences here).
Thus, the sign-token identification process amounts to a reconstruction of a function from item to context. It brings full semanticity to signs and some consistency to semiology. This reconstructive process will be analysed below.
This changes the focus of diagnosis from injury/dysfunction to experience/language/speech and sign definitions. In medical semiology, pain and pain description, for instance, remain close together, and both are ideally transparent enough to render injury visible. This is not the case for psychiatric semiology (although, on occasion, some authors have considered it so). Psychopathology is a reconstructive process which has to deal with the conditions introduced above and which employs a language (psychiatric semiology) to work on it.
Psychiatric semiology is therefore different in kind from medical semiology. It is not a frustrated form of medical semiology. The work of Apel (1994, 2002) and his re-interpretation of Peirce’s work on semiotics leads to a reconception of semiotic relationship from ‘this stands for that’ to ‘knowing something in the world through interpretative mediation in a linguistically structured means’. This change enables us to regard psychiatric semiology as a legitimate enterprise in its own right.
A brief guide to symptom heterogeneity
Before describing in some detail the workings of psychopathology, we have to pay closer attention to the received view of psychiatric semiology as a collection of signs and symptoms which stand in semiotic relation with injuries/dysfunctions. Psychiatric signs and symptoms are actually quite disparate. They may consist of disturbances of movement or of perceptive experience; of altered mood states; of patterns of thought about oneself or others. The empirical field to which the signs may be referring varies widely in content, structure, and legitimate ways to elicit them. These differences have unfortunately been overlooked by most authors, arguably on a pragmatic basis, which will be analysed later. It is therefore all the more important to pay attention to a series of relevant papers by Berrios and colleagues which have been in relative isolation for the past 20 years (Berrios, Marková and Olivares, 1995; Berrios, Marková and Villagrán, 1996; Marková and Berrios, 1995, 2009).
The sources of heterogeneity analysed in these texts can be summarized as follows. An altered neurobiological signal may be consciously experienced by the patient (e.g. sadness or hallucinations) or may be a cause of dysfunction which is not perceived by the patient him/herself (as happens with some thought disorders). It may cause some experience to which the patient reacts (as with anxiety provoked by hallucinations), thus provoking a first symptom related to the anomalous signal and an unrelated second symptom. The nature, intensity, duration and novelty of the biological signal will influence the form and content of symptoms.
However, beyond this, the form and content of these symptoms are also subject to the attitudes of the patient towards the new experiences, together with his/her cognitive, linguistic and symbolic capacities, cultural framework and level of consciousness. On the psychiatrist’s side, the means through which information is elicited (whether by observation, question, even suggestion) depends on his/her own cultural framework, technical and theoretical expertise, etc. Symptom heterogeneity thus springs from:
the signal itself;
both patient’s and clinician’s conceptualization of the new experience;
clinician’s classification and ascribing of diagnostic relevance.
Within this framework, it is important to be able to discriminate symptoms which may share surface resemblances but not deep structures, as they may be caused by different biological signals. Also, as recently proposed by Marková and Berrios (2009), it well may be that no unequivocal signal-symptom relation is found.
Symptom heterogeneity and empirical fields
We can now go one step further. But it is first necessary to refine the terminology. So far, we have used ‘symptom’ or ‘sign’ both for the raw disturbance and for the same disturbance once it has undergone psychopathological reconstruction and labelling. From now on, we will reserve the term for the complex ‘disturbance + name’ and, if necessary, we will refer to the raw disturbance as pre-symptom. This is no whimsical subtlety. It follows a long tradition in logic and philosophy of science (that of distinguishing between formal and material object and scientific and material object) enabling a clearer sight of the processes of symptom formation which depend not on the cause but on the effort on the clinician’s side to make pre-symptoms intelligible.
To begin with, (intimate) experience, (observed) behaviour and (dialogic) speech are obviously heterogeneous in their structure, and they are explored in different ways. Table 1 offers an example concerning schizophrenia. It is no wonder that these four columns roughly match the distinctions introduced by Lantéri-Laura between body and speech which have undergone the semiotic reduction of medical semiology and those which have not. But now we need to consider the differences between observation and understanding which form a contrast between the empirical fabric and the research methods of the natural and social sciences. Habermas (1997: 307–10) summarizes the differences as follows. ‘Observation’ deals with fragments of reality which may be shaped by theoretical vocabulary and assumptions but which, nevertheless, are considered observer-independent. But in ‘understanding’, the researcher is part of a dialogical relationship and intends to understand (rather than describe), and the material of research is symbolically pre-structured. This sets up three dichotomies, each half corresponding to natural or social sciences: ‘perceived reality’ vs. ‘symbolically pre-structured reality’; ‘observation’ vs. ‘understanding’; ‘description’ vs. ‘meaning explanation’. Some years later, Habermas (2003, Vol. I) insisted that, while natural and social sciences share the influence of theory on description and of paradigms on theory, in social sciences there is a need for interpretative understanding prior to theoretical development, due to the symbolic structure of the ‘raw field’.
Symptoms of schizophrenia and procedures of symptom eliciting
usually by family
The observational-descriptive features of psychiatry hardly need further analysis, being now commonplace for both clinicians and scholars. However, the importance of symbolic pre-structure and its pervasiveness in mental symptoms are less well understood. A good description is contained in Eugène Minkowski’s classic Le Temps vécu (1933/1995: 180):
Delusional ideas wouldn’t be the mere product of morbid imagination or impaired judgement, but an attempt to translate an unaccustomed situation to the language of previous psychism […] The patient tries to express this situation by ideas borrowed from her previous life. She thus reaches delusional ideation.
2
Furthermore, it is likely that mood, will and pleasure semiology are even more dependent on this symbolic pre-structure than delusions are. This by no means suggests that psychiatry is a social science, but it seems clear that psychiatric semiology is scattered in two empirical fields which correspond to Habermas’s dichotomy (symbolic pre-structure of the field or its absence). Psychiatric semiology employs both the ‘observation and description’ pole and ‘meaning understanding and explaining pole’, and sometimes it applies both to the same raw disturbance.
But both poles have: different methodologies, different validity conditions and standards, different theoretical and practical interests, different concepts of meaning, different relations with causality and very different predictive power. This is, of course, no novelty. From Dilthey to Carnap to von Wright to Popper to Habermas to Churchland to Passeron, the question remains open. Is epistemic reduction of social sciences to natural sciences amenable and desirable? And should this wait until neuroscience becomes sophisticated enough to render account of qualia, intentionality, consciousness, freedom? Or are all these concepts confounding and obscure, and should be disposed of?
In psychiatric semiology, the dilemma has been defined this way: either the biological or the psychic pole should become the basic empirical field causally related to signs to which the other may be reduced. These reductions have been achieved thanks (partially, at least) to the phantasmatic nature of the semiotic relationship. As long as items of experience or speech are considered signs referring to specific brain/psychic apparatus dysfunction, the differences between them can be levelled without much loss. However, in many instances this type of reduction is not possible. Whenever semiology turns towards one of the poles seeking a clear-cut empirical structure and causal series, the ignored/not-yet reduced pole reappears as soon as the causal chain analysis proves insufficient. This is the basis of the bio-psycho-social model.
The essay by Bolton and Hill (2003), Mind, Meaning and Mental Disorder, might be read as a sophisticated attempt to find common ground for these varied empirical fields. They do so by attributing causal weight to meanings and considering ‘intentionality’ as a trait of the workings of different sub-personal structures of the organism (the liver or the heart, not the sub-personal cognitive procedures of the brain). In order to be able to attribute intentionality to instances other than mental states, they must first translate the intentionality of conscience to informational vocabulary. Then, just as intentional mental states carry information about something other than themselves, the liver and the heart are involved with information fed by circumstances ‘other than themselves’. These packets of information are considered ‘meanings’. As long as these ‘meanings’ provoke changes in the liver and the heart, they are endorsed with causal power. Information about blood pressure is ‘meaningful’ for carotid baroreceptors, which become in this light ‘intentional systems’.
Biology and psychology are thus ‘united in intentionality’ and their differences are attributed to increasing levels of complexity. Due to lack of space, I cannot discuss this text fully, but I do note a couple of problems: (a) this informational reduction of intentionality could fail to distinguish between biological and mechanical devices whose ‘behaviour’ obeys informational states, such as photoelectric cells; and (b) rebuilding the levels of intentionality could be a task just as exacting as deriving intentionality from non-intentionality in a classical way.
On the other hand, other authors as disparate as Lantéri-Laura (2004) and Kendler (2005) have accepted that psychiatry contains some elements of diversity and complexity that cannot be broken down. Indeed, Lantéri-Laura has reviewed the different empirical fields of psychiatry and psychiatric semiology and found this one and only unifying factor: savoir faire, that is, praxis.
Psychiatry is born within medical science, and shares some pragmatic validity conditions with the other branches of medicine: it is supposed to be in charge of healing or alleviating some form of human ailment, and to do so in a rational, scientifically sound manner. But modern medicine is built on and supported by a semiotic relationship, which has not been established in psychiatry. However, this semiology affects psychiatry’s understanding of itself, and thus:
hinders clinicians, scholars and researchers in keeping an adequate balance between descriptive definitions and individuation factors;
neglects the differences between the two empirical fields on which psychiatric semiology is based.
A number of authors, notably Bentall (2003), have criticized this state of affairs, arguing that it has led psychiatry (or clinical psychology) to stagnation, inefficiency, lack of predictive power, conflicting theoretical and therapeutic approaches, etc. There may be something in this, but we cannot expect much improvement from a strategy based on displacing semiotic relations from syndromes to symptoms and from brain to psychic functions.
There is, however, a possible way out of this dilemma. Both Minkowski (1933/1995, 1926/1997) and Lantéri-Laura (1994) have disputed that psychology, for instance, is basic for psycho(patho)-logy. We have seen how Lantéri-Laura understood faculty psychology as a context of justification, rather than a context of discovery, and he was strongly critical of the psychologization of psychiatric semiology: ‘The semiotics employed in clinical practice is hardly derived from general psychology affected by some “pathological index”. It stems from reasoned practice and clinical knowledge.’ (Lantéri-Laura, 1994: 465–500). Minkowski (1933/1995) considered that the basic language of psychopathology could be built not on the concepts imposed by either psychology or physiology, but on some on sort of philosophical anthropology which would render feasible the balance between symptoms (unspecific) and the background which would make them specific (personality, in Minkowski’s approach).
I have criticized Minkowski’s conception elsewhere (Ramos and Rejón, 2002), but its intuition, honed by Lanteri-Laura’s work, remains valid. It is feasible to undertake a rational reconstruction of psychiatric knowledge and practice that escapes this dilemma. For psychopathology need not be (and is not) based on some basic science (or several), but instead it relates to a special way of dealing with pathological experience or speech, and on a special balance between descriptive definitions and individuation factors. But in psychopathology, we need to distinguish between this cognitive activity which renders abnormal experience or behaviour intelligible and the language it employs – tool and product at the same time. From now on I shall call this cognitive process psychopathology and reserve the term psychiatric semiology for the language. But this semiology is no longer referentially compromised within a psychic or biological empirical field. It deals with experience, language and behaviour before they are conceptually structured by either physiology or psychology, mostly as Lantéri-Laura proposed. The traits gathered in the definitions which constitute the semiological thesaurus work rather as descriptive anchors that enable the reconstructive process to begin rather than being specific or close-to-specific translations of brain injuries or psychic dysfunctions.
The logic(s) of psychopathology
If psychopathology is a cognitive activity which brings intelligibility to items of experience or speech acts, we may well call it logic in a strong sense, which includes but exceeds formal logics. It is logic as long as it is a tool for knowledge, an organon. There is, of course, a long and rich tradition backing this stance (Kant, Hegel, Husserl). I will state briefly some different ways in which psychopathology works to bring its objects to knowledge, and the unifying procedure which operates behind them. First, I shall deal with descriptions of behaviour, made either by the patient or the psychiatrist; next, the structural effects of language on the raw pre-symptoms; and finally, the unattended pragmatic functions which underlie signs and symptoms.
Meaning and perception
Descriptive approaches to psychiatric symptoms usually overlook the fact that whatever we choose to describe, and whichever features are considered essential for the description to be valid and reliable, they are picked from an already meaningful experience, mostly perceptive. This meaningful quality of perception has been differently tackled in different epistemic traditions. In analytic hilosophy, McDowell’s (1996: 15–34, 37–92) reading of Sellars’ (1997) critique of sense data as the bare ‘givens’ of experience pointed to the ‘conceptual nature of experience’ as a way to do without the ‘interminable oscillation’ between the attempt to secure thought about the world in some crude givens (understood by Russell or Moore as sense data) and the choice of leaving the application of concepts unconstrained as a consequence of rejecting the ‘idea of the given’. If experience is capable of providing rational constraint on thought, then it should be thoroughly conceptual before judgement. I shall not pursue this thread here. The other major philosophical stance which has argued for the meaningful quality of pre-predicative experience is phenomenology.
There is a long and winding relation between psychiatry and phenomenology. It is necessary, however, to remark on its ties with classic semiology. Phenomenological psychopathology (Binswanger, Blankenburg, Stanghellini, Sass) employs conceptual tools borrowed from Husserl and Merleau-Ponty but does not part with the traditional thesaurus (some trends do: there are several currents and families within phenomenological psychiatry) in order to attain richer, more faithful descriptions of pathological phenomena (Daumezon and Lantéri-Laura, 1961). Of course, a few assumptions are made, the first of which is the meaningfulness of pre-predicative experience and, second, some basic features which underlie this experience: holism; openness of the horizon of possible experiences implicated by current experience, embodiment, spatiality and temporality. Pathological varieties of pre-predicative experience should express changes in one or more of these dimensions. In delusional perception, for instance, the horizon of experiences which should underlie my seeing a train is restricted to just one possible meaning (let us say imminent kidnapping). Phenomenological psychopathology in its current developments is more prone to exchange concepts and findings with empirical sciences than it once was. Nevertheless, its focus remains unchanged. Nor is it necessary to adhere to a phenomenological approach to pre-predicative experience to share the value that phenomenology has accorded to it. Psychopathology is tied to whatever the patient and the psychiatrist experience.
From meaning to sign
Pre-predicative experience is, however, only a component of knowledge. Full knowledge needs linguistic form; either common language or a constrained, specialized language such as psychiatric semiology imposes its form on pre-predicative experience. Thus, symptom description depends not only on the experience but on the tool employed to shape it.
This does not imply that pre-predicative experience is a heap or bundle of indiscriminate information. We have just seen that it has its own structural features. But language is a different means, with different structural features which are not dependent on extra-linguistic, pre-predicative experience. This is the core of Derrida’s early criticism of Husserl’s philosophy of language. As Derrida (2003) argues, there is no such thing such as transparency of linguistic signs. They are not meaningful as long as they express some ideal object which has been previously constituted in conscience. Language is the locus for meaning, and language is there understood as a system articulated through opposition and difference (‘m’ vs. ‘n’, ‘mother’ vs. ‘bother’) which connects a signifier series of elements with a signified series of elements. In that connection lies meaning.
But language does not exhaust other meaningful structures. The really interesting issue is, however, that raised not by the existence of non-linguistic meaning but by the need to render it in linguistic terms (Ducrot and Todorov, 2003). In this contradiction lies a good part of the task of psychopathology. A semiological thesaurus and rules (in other words, a language) is an unavoidable need. But psychopathology points towards the non-codified, individualized conditions which enable descriptive definitions to reach full-blown meaning.
These structural features of language which ensure signification are responsible for an unexpected side-effect. Naïve reconstructions of psychiatric semiology rely unawares on this assumption. If there is a signifier, then there must be something signified. This is trivially true of words (the terms ‘delusion’, ‘delirio’, ‘délire’, ‘Wahn’ point to some experiences and behaviours) but not necessarily true of the brain injury or psychological dysfunction when they are understood as the true meaning of psychiatric semiology, analogous to liver disease as the meaning of jaundice.
To summarize, there are two heterogeneous fields in which meaning lies, namely pre-predicative experience and language. Pre-predicative experience needs language to become fully intelligible, but it is not exhausted by language; language does not need pre-predicative experience to generate meaning, but it is not able to exhaust the features of non-linguistic meaning.
From sign to praxis
There are at least two historical traditions concerning praxis which we need to recall. On the one hand, leading from Hegel to post-Marxist thinking (and, of course, reaching back to Plato) praxis has been understood as anthropogenic. In other words, it is a series of material activities which not only create utensils or books or works of art, but they also create human beings, objectifying needs, desires, plans and knowledge in a panoply of tools and ‘subjectifying’ Nature into the shape of human beings, different from the rest of mammals, able to defer the fulfilling of needs.
On the other hand, from the 1950s on, philosophy of language has elaborated on Wittgenstein’s insights of ‘meaning as use’, leading to different conflicts and agreements with other approaches to semantics and with the hermeneutic approach to meaning. In psychiatry, this approach has been successfully employed by Fulford (1989, 1992, 1993), for instance (drawing from Austin’s work) in his analysis of some key terms such as insight and psychosis. While there are serious difficulties in reaching sensible definitions of such terms, Fulford has shown that their use in common clinical practice is quite precise. Heinemaa (2000a, 2000b) has turned to Wittgenstein in a similar analysis of ‘person’, ‘personality’ and ‘psychosis’. 3 There is common ground between these two traditions. Both of them deny the ontological or epistemological priority of theory. Praxis is not the mere application of abstract theoretical principles. It has its own specific functions, criteria for success or coherency, even its own spatiality and temporality. Bourdieu (1980, 2000) summarizes them as follows. Praxis organizes thought, action and perception following schemes which may render contradictory results when theoretically reconstructed. It subordinates formal logic to ‘practical principles’ which are: efficacy, simplicity, ease in use, and close connection with economic, institutional and social needs; coherence must be contextually assessed, and it can stand a good number of contradictions as long as the results are successful; and, what is more important for psychopathology, there are practical schemes of perception and judgement (this is, they do carry semantic load) which are learned without theoretical awareness and which co-create the perceptual subject as much as they give meaning to the perceived object (they are ‘anthropogenic’). These schemes perform two main cognitive tasks: they group phenomena together and they split groups apart, both of them on the basis of partial resemblance, analogy or partial opposition.
Returning to psychopathology, this approach to the semantic load of praxis furnishes a third layer to the manifold set of features found in symptom definitions. They depend not only on the rendering of pre-predicative experience or on the structural properties of language. They must ensure the proper fulfilment of their practical roles, either grouping or splitting. Let us look at an example of each. Quite soon after the publication of Bleuler’s text on dementia praecox (1911/1960), Minkowski (1926/1997: 35) realized that Kraepelin’s work had had the unexpected side effect of diminishing the importance of isolated symptoms, rendering them more or less non-specific. But, if there was a commonality of course and prognosis, it was reasonable to expect some kinship between the symptoms found in schizophrenia. Bleuler had advanced a first attempt at this logical reduction (Bleuler 1911/1960). He split schizophrenic symptoms into four categories: fundamental (the famous four As, those symptoms needed for diagnosis) and accessory on the one hand; primary (closer to biological substrate) and secondary (which developed from the former). Loose associations were the only symptom both fundamental and primary. But Minkowski found that it was hard to explain the multifarious symptomatology and the peculiar unity of them by the mere looseness of associations. Something else was needed, something which would affect the person as a whole. So he took autism and ‘refunctionalized’ it. To do so he needed another set of descriptive items and the chance to secure this description on some basic instance which he found in philosophical anthropology. Although I cannot start a detailed discussion of Minkowski’s work here, I want to stress that this reformulation of autism was driven by semiological as well as philosophical forces: there was a search for the specificity which could only stem from the ‘trouble générateur’ of schizophrenia (Minkowski, 1926/1997: 96). This search would affect its final description.
Something similar happened to delusional perception. The symptom was first described by Jules Baillarger (Berrios, 1996: 97) and ‘refunctionalized’ by Karl Jaspers and Kurt Schneider. This time the role assigned consisted of splitting apart schizophrenia and cyclothymia. Schneider (1997) rightly considered hallucinations and delusional interpretation as non-specific. But he deemed delusional perception to belong only to schizophrenia (and organic brain disease). To perform as a tool for differential diagnosis, delusional perception needed a convenient description which could set it apart from both hallucinations and delusional interpretation. Schneider found it in the two-fold structure of delusional perception: normal percept to which abnormal meaning is accorded. Later empirical research (which belied this specificity) and theoretical changes in the conceptualization of bipolar disorder would render Schneiderian delusional perception impractical (Rejón, 2002).
Thus, we find a similar process: psychiatric signs stem from the needs of clinical practice, they receive some descriptive formalization and theoretical foundation and are then applied in clinical practice until they are no longer able to fulfil this double role of grouping/splitting so eagerly needed in psychiatric semiology.
Psychopathology as a logical technique
We come to the following conclusion: perception, designation and praxis are the three components of symptom construction in psychopathology. Their respective importance varies with each symptom, but the three of them are the manifold unravelling of a single intelligibility-building process which I will call technique.
Technique (from the Greek téchne) will be employed in this context as a specialized term, coined by Duque (1986, 1995) in an attempt to understand the way human activity co-creates subjects and objects or generates such concepts as world and world-dweller, Nature and civilization, citizens and barbarians. Its ontological foundation assumes that all these kinds of praxis share a matrix of intelligibility which varies in history. Thus, Greek social organization, theoretical thought, art and industry will share a series of features concerning basic concepts about Nature, objectivity, freedom and the like. It is easy to see how our own social organization, industry, communication, theoretical thought and ethical demands are quite differently organized. As long as technique is able, at the same time, to articulate the most basic layer of intelligibility and to refract itself both into social practice and ontological assumptions, I consider it a useful and illuminating loan in a meditation concerned with a process which, at the same time, grants a minimum of intelligibility to some types of human experience and behaviour, articulates itself into distinct components and is performed by a specialized fraction of society on a special fraction of society (the obvious ethical and political strands will not be pursued here)
To summarize, then, psychopathology is a technique which produces symptoms out of raw behavioural/experiential phenomena. Psychiatric semiology is a product of this technique. The semiological thesaurus consists of a series of descriptive/stipulative definitions, but this thesaurus has failed in the historical attempt to build a psychiatric semiology analogous to medical semiology. This is due to: (1) there are not basic medical semiotic relationships between symptoms and dysfunctions, and (2) experience, behaviour and speech, which are the raw material of psychopathology, need individuation-related information to fix reference, but these are lacking in psychiatry. Thus, token-type identification processes are replaced by symptom reconstruction procedures. Such a procedure begins with a definition, and then moves on to include the subsignifier
Conclusions
We have changed the focus of psychiatric semiology from a medical-like referring semiology to processes that create intelligibility in the absence of defined physical pathology. The coherence achieved in medicine by basic medical semiotic relationships is attained otherwise in psychiatry. In particular, there is a different balance between formalized descriptive features and individuation-related features. This in turn has important implications for future research in psychiatry. As long as clinical interviews by experts remain gold standard, any tendency affecting this balance may have consequences in symptom eliciting and, it goes without saying, in the empirical work erected on it.
But there is a larger agenda involved in this change of focus. Causal relations have been employed as intelligibility creation devices. While this has worked reasonably well for medicine, it is not the case for psychiatry, where the semiologic thesaurus has failed to establish a solid gateway to causation. There are two possible ways out of this cul-de-sac. Maybe brain/mind processes are just complicated, and patience and hard work along current lines is what is needed. But maybe we are dealing with different issues here. Campbell (2008) has argued that causal relations should not be expected to render abnormal experience or behaviour intelligible. Whenever and wherever they are found, they are just hard data which should pave ways to treatment and prevention – and maybe it is just so. Thus, psychopathology and clinical semiology should do their best to strengthen their own internal accuracy and consistency, while more sophisticated models or causation are explored and tested. In my view, this does not need to sever psychopathology from empirical research. It just recognizes the different nature of medical and psychiatric semiology, and the need to devise other, more suitable, models of relation between brain, mind, experience and behaviour, rather than continuing to squeeze psychopathology into an inherited framework.