The Many Dimensions of Recovery Definitions, Problems, and Possibilities

ABSTRACT

In recent decades the possibility that people diagnosed with severe mental illness might live a satisfying life outside psychiatric institutions and even independently of psychiatric services has won some acceptance. During the same period the concept of recovery has been given different, contradictory, and/or complementary definitions. Often users, staff, and scientists have taken a stand for a specific definition, clinical, or personal recovery, but do not problematize the other main concept: What is the person in recovery from? In this article we review the different definitions of recovery and the diagnostic categories, and articulate their pros and cons. We propose a critical viewpoint by arguing for transparency and critical reflexivity, putting one's main concepts and choices under scrutiny. This should not be seen as an obstacle, but rather as an obvious part of scientific research, as such concepts and choices will influence the results of the research.

The first step is to measure whatever can be easily measured. This is OK as far as it goes. The second step is to disregard that which can't be easily measured or to give it an arbitrary quantitative value. This is artificial and misleading. The third step is to presume that what can't be measured easily really isn't important. This is blindness. The fourth step is to say that what can't be easily measured really doesn't exist. This is suicide.

—The McNamara fallacy (also known as the quantitative fallacy). Yankelovich, D. (1972). Corporate priorities: A continuing study of the new demands on business.

The existence and probability of recovery from or in severe mental illness (SMI) varies a great deal between different studies. These variations seem to be related not only to the person's state and situation, but also to the definitions of recovery and of the actual diagnosis clinicians and researchers use. If we start by limiting the range of acceptable aspects of recovery and then disregard other aspects and thus other facts as unimportant and even nonexistent, we risk creating an ideological construction that could lead to the reduction of the positive potential of recovery for the mental health field.

Controversies About Diagnostic Criteria

The definition of the main mental illnesses and their eventual connection to chronicity has been the subject of many controversies. For many decades, and still today, the diagnosis of schizophrenia has been associated with a downward-moving process in which the person is stripped of his or her human capacities. Kraepelin (1919/1971) characterized "the schizophrenic" as follows:

Emotional dullness, failure of mental activities, loss of mastery over volition, of endeavor, and of the probability for independent action. The essence of the personality is thereby destroyed. . . . The annihilation of human will. . . . The loss of the inner unity of the activities of intellect, emotion and volition.

(p. 75)

In our own time we can find the same representation. Frith and John-stone (2003) wrote about "the negative features of schizophrenia, the loss of will and the poverty of thought, are associated with a tragic decline in intellectual and social function" (p. 123).

Because of this "tragic decline," Kraepelin (1919/1971) was pessimistic about the possibility of recovery: "We shall be able to pronounce a final judgment about the issue of an apparently cured case only after a very long time" (p. 187).

Both Kraepelin and Bleuler thought that on close examination, a trained eye could almost always detect traces of the illness, even in people who had received no psychiatric treatment for many years and lived a happy and productive life. Patients who have recovered are therefore sometimes rediagnosed afterward to preserve the notion of the natural and degenerating course of schizophrenia (Davidson, 2013).

This pessimistic view of the course of the illness was reflected in the fourth edition of the American Psychiatric Association's (APA's) Diagnostic and Statistical Manual of Mental Disorders (DSM– IV). "Schizophrenia tends to be chronic. . . . Complete remission (i.e., a complete return to full premorbid functioning) is probably not common in this disorder" ( APA, 2000, p. 282).

Against this background, for many years the dynamics of the recovery process were not a central issue in the psychiatric field, though in the 1970s and 1980s several studies were published showing a relatively high probability of recovery for individuals with a schizophrenia diagnosis (Bleuler, 1978; Ciompi, 1980; Harding et al., 1987; World Health Organization [WHO], 1979). Those results opened a new research field studying the dynamics of the recovery process, and also led to controversies about the definitions of, and thus the criteria for, recovery: "Yet across different counties and settings, the term is used inconsistently" (Slade et al., 2012, p. 353). The importance of the definition was underlined by Bellack (2006): "Just as one can define recovery in such a stringent way as to make it an impossible goal, it can also be defined so broadly as to make its achievement unimportant" (p. 437).

Thus, results from studies of recovery are dependent on the definitions of two phenomena: recovery and the current psychiatric problems. Research on recovery is facing a problem usually formulated as the lack of consensus on how these should be defined; and even whether both recovery and SMI, like the diagnosis of schizophrenia, exist at all (Boyle, 2002; Davidson & Roe, 2007; Leamy, Bird, Le Boutillier, Williams, & Slade, 2011). Many attempts to manage this situation have in common that they propose only one definition. This might be in the form of a consensus definition disregarding the important creative conflicts in the different definitions, or of one that disregards the others and even negates their existence.

Another coping strategy could be to adopt a critical reflexive approach ( Bourdieu, 2004). This approach proposes that we should analyze the construction of the core concepts that occur in the recovery field and in our own studies.

In this article, we summarize the construction of the main definitions of recovery and of the problems from which people are recovering. We also describe the problems with and limitations of the different definitions. Critical reflexivity should not be seen as an obstacle to scientific research, but rather as an obvious part of it.

The following review is based mostly on articles published after 2000, but also on a few books from that period and on articles published earlier. We argue that the review presents a fair enough overview of the different available definitions in Western literature of the phenomena it focuses on and of their pros and cons.

The Object of Recovery

Different concepts occur in research about what recovery relates to. Most concepts are on a diagnostic level: schizophrenia, psychosis, and SMI. In some cases, concepts that break with a narrow diagnostic frame such as "the human being" are used. The choice of concept has important implications for the results of the studies and how these are interpreted.

The three aforementioned diagnostic concepts lack commonly accepted definitions. Their meaning and the symptoms included in their definitions vary between countries, psychiatric services, and also psychiatrists, depending on various diagnostic manuals and local practice (Cooke, 2014; Frances, 2013). They also vary depending on the time the diagnosis was delivered, because significant variations may occur between different editions of the same diagnostic manual. Multiple concepts can even occur as synonyms in a single text (Davidson, O'Connell, Tondora, Staeheli, & Evans, 2005 ).

In our review, researchers described the lack of scientific justification behind the adoption of new diagnoses (Young, 1995) and the rejection of established ones ( Kirk & Kutchins, 1992). Davidson and McGlashan (1997) underlined "the lack of clear and broadly accepted criteria for diagnosis" (p. 40). Insel (2013), director of the National Institute for Mental Health, criticized the latest version of the DSM (DSM– 5 ) as lacking validity, adding: "The DSM diagnoses are based on a consensus about clusters of clinical symptoms, not any objective laboratory measure." Few have problematized the consequences of the choices made regarding the phenomenon to be investigated (see, however, Strauss, 2005; Tew, 2013).

Schizophrenia

Schizophrenia occupies a special position in the psychiatric field, as one of the fundamental diagnoses in modern biomedical psychiatry (Kraepelin 1919/1971; Garrabé, 1992). Schizophrenia's central role in the recovery field is based on the assumption that this diagnosis was as a chronic condition that meant a steady deterioration in the condition of the individual up to a stage of dementia.

In today's definitions of schizophrenia, this degenerative-course perspective appears when it is described "as a chronic illness with persisting, relapsing or deteriorating symptoms, and no hope for a sustained remission and recovery of functioning" (Liberman, Kopelowicz, Ventura, & Gutkind, 2002 , p. 256). In the same spirit, Andreassen (1984) wrote that "the 'burned-out' schizophrenic is an empty shell—[he or she] cannot think, feel, or act. . . . She or he has lost the capacity both to suffer and to hope—and at present, medicine has no good remedy to offer for this loss" (p. 63).

In this perspective, the possibility of recovery is considered as negligible; Kraepelin (1919/1971) wrote, "A complete return to premorbid functioning is unusual—so rare, in fact, that some clinicians would question the diagnosis" (p. 185). This tradition has survived in the DSM, which in its fourth edition offered a clear echo of Kraepelin's words: "Complete remission (i.e., a complete return to full premorbid functioning) is probably not common in this disorder" (APA, 2000, p. 282. See also APA, 2013; Liberman et al., 2002).

To handle the case of patients who received a diagnosis of schizophrenia and recovered, specific diagnoses were created such as "reactive psychosis, atypical psychosis, schizoaffective disorder and schizophreniform psychosis" that have "doubtful validity and clinical utility" (Carpenter & Kirkpatrick, 1988, p. 645). Such procedure has significant implications for the research results. Harding, Brooks, Takamaru, Strauss, and Breier (1987) wrote that "finding an outcome of chronic illness may be primarily related to the original selection of patients with a longstanding disorder at the entry criterion" (p. 733).

In contrast to these anticipated chronic developments, several followup studies "demonstrated considerable heterogeneity in the long-term course of schizophrenia" and stressed "the absence of valid disease entities" (Carpenter & Kirkpatrick, 1988, pp. 645 & 649; see also Harding et al., 1987). Strauss (2005) summarized the findings from the U.S. part of the World Health Organization's international follow-up study of schizophrenia (1979): "We suggested that different diagnostic categories were not discretely separate but that many patients actually fell in between diagnostic categories and between the diagnostic categories and 'normal'" (p. 52).

The diagnosis of schizophrenia has been criticized for a lack of reliability and validity (Bentall, 2010; Read, 2004 ). Its very existence has been questioned by Boyle (2002) , who points to the crucial differences existing between what Kraepelin described and current definitions. Carpenter and Kirkpatrick (1988) wrote that it was rather "a clinical syndrome rather than a single disease entity" (p. 645).

Some countries have abandoned the schizophrenia diagnosis because of its stigmatizing consequences ( Sartorius et al., 2014). Even a leading scientific journal, Schizophrenia Bulletin, has discussed abandoning the concept in favor of the term psychosis (Carpenter, 2016).

Psychosis

Psychosis is a generic term for psychiatric diagnoses that, according to Jaspers (see Bentall, 2003) are incomprehensible based on the individual's experience, and are therefore assumed to be caused by biological disorders. Psychosis was used as opposed to neurosis, which was considered to be understandable in the light of the individual's experience. Bentall (2003) mentioned four main groups of psychosis: schizophrenia, schizoaffective disorders, manic depression, and paranoia. Several of these are in turn divided into different forms. Common to them is that the person is considered to have a lack of reality testing. However, the sum of the four diagnoses with unclear and changing definitions could not resolve the problem psychiatry met with a single one. Moreover, a person's problem that one psychiatrist perceived as understandable need not be so for another.

Although even an unambiguous definition of the term psychosis has been difficult to achieve, more general diagnostic-based categories have been developed. In his editorial in Schizophrenia Bulletin, in which he argued for a change of name of the magazine, Carpenter (2016) pointed also to the problem of the psychosis concept and presented proposals that included "psychoses and related disorders" (p. 864), which, however, remained undefined, but could be linked to another concept often used in research contexts: SMI.

Severe Mental Illness

The concept of SMI occurred because of the aforementioned uncertainties in the delimitations of a certain number of diagnoses. A broader concept would avoid this problem of delimitation between specific diagnoses, but different practices have been developed regarding the diagnoses contained in this overall concept:

Severe mental illness is often defined by its length of duration and the disability it produces. These illnesses include disorders that produce psychotic symptoms, such as schizophrenia and schizoaffective disorder, and severe forms of other disorders, such as major depression and bipolar disorder.

(Hazelden Foundation, 2016)

In this definition, we find various forms of psychosis, but also severe forms of depression and other unspecified forms of disorders. Other definitions also include personality disorders. The National Institute of Mental Health (2016) gave a broader definition, in which no diagnosis is mentioned:

A mental, behavioral, or emotional disorder (excluding developmental and substance use disorders); diagnosable currently or within the past year; of sufficient duration to meet diagnostic criteria specified within the 4th edition of the Diagnostic and Statistical Manual of Mental Disorders . . . and resulting in serious functional impairment, which substantially interferes with or limits one or more major life activities.

Here unclear notions such as "sufficient duration" and "serious functional impairment" are the key aspects of the definition. The Royal College of Psychiatrists (2016) , for its part, equated SMI with psychoses. In an information sheet they wrote, under the heading "Severe mental illness (psychosis)," "This leaflet is aimed for the carers of people with Severe Mental Illness (Psychosis)."

SMI can therefore be equal to psychosis, but also includes various other diagnoses. In striving to get away from ambiguities and the varying definitions of specific diagnoses, wider concepts were created that nevertheless seem to suffer from the same problem of lack of reliability and validity that occurred for the specific diagnoses.

The Person

The concepts of schizophrenia, psychosis, and SMI are based on the classification of psychiatric disorders. They have been challenged by people with personal experience and by researchers. Ramon, Healy, and Renouf (2007, p. 111) found that the new definition of recovery "entails also a relook at what psychosis is all about." The common starting point for many critics is that mental distress is not an illness, a disorder, or a disability; a deviation from the socially constructed normality, but an expression of the human being with its complex diversity (Basaglia, 1982).

Deegan (1996) formulated this alternative approach as follows: "What exists, in the truly existential sense, is not an illness or disease. What exists is a human being and wisdom demands that we see and reference this human being before all else" (p. 92).

Strauss (2005) focused on "the person with disorder as a person" (p. 52). Such an approach can be traced back to Bleuler's work: "In fact Bleuler taught us to see 'the person behind the disorder'" (Harding, 2005, p. 44). Timander, Grinyer, and Möller (2015) wrote, "The goal of recovery is thus not to become 'normal.' The goal is to become more deeply, more fully human" (p. 331).

Such a widening of the clinical perspective makes psychiatry's object as only a biomedical deviation from a supposed normality problematic. Against this background, Strauss (2005) mentioned "the most fundamental issue of our field: how can we make the mental health field a human science?" (p. 53).

Criticism

The understanding of mental health problems as diagnostic categories and illnesses and the dichotomous notions (healthy/ill) behind them is problematized in various studies (Bentall, 2004). Borg and Davidson (2007) studied recovery in an everyday life context and wrote that "when the symptoms are not viewed within the context of the person's everyday life they've become simply representations of an illness identity" (p. 129).

This perspective has been developed further in the studies that demonstrated that the person's economic conditions could affect the presence and extent of various psychiatric symptoms (Davidson et al., 2004; Ljungqvist, Topor, Forssell, Smith, & Davidson, 2016; Sheridan et al., 2015).

In many studies, people who were followed up show a complex development. Some may have persisting symptoms, but live a rich and socially integrated life. Others may be asymptomatic, but have a limited social life (Harding et al., 1987; Torgalsbøen, 2005). Strauss (2005) emphasized this complexity as one of the greatest challenges for psychiatry.

Farkas (2007) took a further step:

Former patients and other critics of biological approaches have questioned whether mental illnesses even exist as medical entities and prefer to think of life crises as a normal part of human existence. From this viewpoint, there can be no "recovery" because there has been no illness.

(p. 68)

Summary

There is considerable confusion regarding one of the key concepts in recovery studies. Diagnoses are not based on biomarkers but on clusters of symptoms, leaving space for subjective judgments depending on factors independent of the state of the person diagnosed. Attempts to use wider clusters exacerbate the problems with single diagnoses. To manage these difficulties some researchers have argued for considering the whole person in his or her context. Thus, it is not only concepts about recovery that lack unequivocal definitions; this seems to occur throughout the entire mental health field.

The researchers' definitions of their selection criteria for the people they will follow up will play a significant role in their studies' results. By excluding or including various user groups, the proportion of people who will turn out to be recovered or in a recovery process will vary.

Recovery From or In?

In studies investigating recovery, a split occurs related to the fundamental question of whether people recover from, or in, a psychiatric condition.

Recovery from or recovery in is about whether recovery can be seen as a process that can last while the person is showing symptoms and impairment, and is receiving a psychiatric treatment, or if recovery comprises a recovery from these symptoms and from functional impairments, leading to a condition in which the individual is asymptomatic and independent of support ( Davidson et al., 2005).

We next present these different definitions of the concept of recovery and how they relate to each other, as well as critical comments that have been made for each definition.

Clinical Recovery

Clinical recovery has become a generic name for the definition of the concept of recovery, debuting in the follow-up studies published around the 1980s ( Slade, 2009; Davidson, 2013). Recovery was then divided into total and social recovery (Bleuler, 1978; Ciompi, 1980; Harding, Brooks, Takamaru, Breier, & Strauss, 1987; Warner, 1985/2004; WHO, 1979). In both cases the degree of recovery was based on professionals' assessments of the person's symptoms, care consumption, and social adjustment.

The advantage of clinical recovery has been declared to be that its definition is "scientifically based" (Liberman & Kopelowicz, 2002, p. 245) and "operationalizable-suitable for use in empirical research" (Slade, 2009 , p. 35). Additionally, it has been described as objective when it is judged by "an expert clinician, not the patient," often with standardized instruments (p. 35).

Total Recovery

Total recovery has sometimes been defined as equivalent to "cure" ( Davidson, 2013): the absence of symptoms from the initial disorder or illness, with no current treatment interventions, which during the first decades of the 1900s could be equated with no "hospitalization" (Warner, 1985/2004) at the follow-up or during a specified period (usually at least two years, but up to five; Arvidsson & Arvidsson, 2005), as well as a normal social life, often in the form of independent living, work, or study and a normal social network (Warner, 1985/2004).

Mason (1995, as cited in Torgalsbøen, 2005) summarized total recovery as "no symptoms, no disability and no treatment" (p. 303). Several authors mention a "return to pre-illness state" as a criterion for total recovery ( Davidson & Roe, 2007, p. 462). On the other hand, Liberman & Kopelowicz (2002, p. 246) warn that we should not "confuse recovery with cure," which could reduce recovery just to a notion of well-being.

Social Recovery

Social recovery has been operationalized using the same criteria as total recovery, but on a lower level. Symptoms may occur, if they are not inconvenient for the person in his everyday life, and psychiatric outpatient care can occur, but no periods of hospitalization. The person should also live a normal social life (Warner, 1985/2004). The duration of the various criteria is the same as for total recovery.

Davidson & Roe (2007) gave a modern definition of social recovery as

the amelioration of symptoms and other deficits associated with the disorder to a sufficient degree that they no longer interfere with daily functioning, allowing the person to resume personal, social and vocational activities within what is considered a normal range.

(p. 461)

In addition to the absence of both "positive and negative symptoms and signs," Liberman and Kopelowicz (2002) mentioned

work in the regular labour market or studies at ordinary educational institutions and social activities with others. . . . Independent living without supervision of money, self-care skills and medication. . . . Cordial family relations and contacts. Recreational activity in normative settings. . . . Subjective satisfaction with life. Self-esteem and stable self-identity. Participation as a citizen in voting, self-advocacy, neighbourliness and other civic areas.

(p. 250)

Criticism

Despite researchers' claims of being able to provide objective measurements of a state within the individual, attempts to assess both total and social recovery have resulted in different results.

Warner's (1985/2004) compilation of different follow-up studies of people with a diagnosis of schizophrenia in the United States and Great Britain during the 1900s indicated that between 12% and 23% were judged to have achieved a total recovery, and between 29% and 44% a social recovery. Slade (2009) compiled nine studies published between 1976 and 2001 of people with a psychotic diagnosis, in which between 46% and 68% were deemed "recovered or significantly improved." (p. 36).

The Swedish Council on Health Technology Assessment (SBU; 2012) compiled 33 medium-cohort studies (between 2 and 5 years). They summarized the findings as follows:

There is moderately strong scientific evidence that some people with schizophrenia recover completely from the disease schizophrenia. About 15 percent of persons with first time psychosis had recovered compared with 0–10 percent of people who have been ill for longer. However, it cannot be assessed how sustainable this recovery was.

(p. 284)

In the case of meta-studies and comparisons between study results, Slade (2009) wrote that "results are difficult to interpret due to differences in patient selection, definition of recovery, use of retrospective versus prospective tracking, frequency of repeated measurement, length of followup, location and time period" (p. 36).

SBU (2012) made several critical points about follow-up studies: They are described as having "vague, non-operationalized and non-standardized outcome measures . . . some measures are heavily dependent on context" (p. 291). In terms of methodology, issues mentioned are that "some scientists doubt if it is very meaningful to combine results from studies using highly diverse diagnostic systems" (p. 291). SBU's criticism applies also when various outcome measures are combined in meta-studies. SBU recommends that only studies with "quantifiable and well-defined outcomes" (p. 292) should be included. Their criticism also applies to the study design and management of the attrition rate in the different studies, before they finally note, "The way to diagnose schizophrenia disorder has varied greatly over the past 100 years," with "mutually incomparable diagnosis traditions" (p. 294).

Other scientists have also formulated criticism. Davidson and MacGlashan (1997) take up "sample bias," "lack of control and comparison groups," and "the complicated inter-play of disease processes, treatment effects, and social and cultural context" (p. 40).

Another criticism is about the follow-up period. Various studies have presented a period of 5 years after the "onset" has been presented as when a plateau in the development of the illness occurs and a time when the recovery process can be noticed (e.g., Carpenter & Kirkpatrick, 1988; Davidson & McGlashan, 1997 ). Harding (2005) therefore made a distinction between follow-up studies and long-term follow-up studies.

Despite claims of providing an objective measurement of the individual's condition, the assessment of clinical recovery also depends on external factors that are independent of the individual's mental state. One such factor affecting the results of follow-up studies is the criterion of the clinical recovery requirements. Many definitions of clinical recovery mention as a criterion a return to "the premorbid level." This criterion has been questioned, since this level is rarely known, and many of the people concerned state that they would not want to return to it (Davidson et al., 2005). Davidson and Roe (2007) argued that such a criterion is in most cases impossible to achieve "given the traumatic nature of being treated as a mental patient" (p. 462). Instead, personal recovery is often described, as it "involves growth and an expansion of capacities" ( Davidson et al., 2005, p. 15).

Criteria related to social life risk becoming normative without any scientific basis (Rose, 2014). They might reflect the cultural vision of an ideal life of the people who set up these criteria. Some criteria in the literature are having cordial family relationships or a stable self-identity, and voting in elections. Rose (2014) pointed out that "not socializing with others or breaking the connection with families was not necessarily dysfunctional but a way of protecting one's mental health" (p. 217; see also Andersson, Denhov, Bülow, & Topor, 2015 ; Crawford et al., 2011; Tew, 2013). Ralf (2005) formulated this criticism into a question: "How many goals must be achieved to be considered recovered? For that matter, how many life successes are considered 'normal?'" (p. 5). This normative risk and sociocultural bias is visible in, for example, Liberman and Kopelowicz (2002), when they wrote that self-esteem, choice, and self-determination—often highlighted in the definitions of personal recovery—"may just as frequently be associated with the personal choice to 'feel good' by using social security income to obtain cocaine, while living a homeless and victimized existence on the streets" (p. 247).

Rose (2014) also noted that "certain goals are not permitted" (p. 217). The right to be different, a real individual, ends up in opposition to the dominant discourse of normality. Behind those normality requirements are the Anglo-Saxon tradition of normalization, which focuses on the normalization of the individual (Wolfensberger & Tullman, 1982), in contrast to the Scandinavian tradition, which focuses on the normalization of the individual's living conditions (Nirje, 1985).

Societal factors such as the local structure of mental care services (proportion of inpatient care and open care interventions), housing and labor market policies (having and maintaining a home and job), social insurance (having a decent life, even if one does not work; Standing, 2011), and norms concerning social relations (i.e., about the extent and composition of a social network, marriage versus other forms of partnership) will be crucial for assessing an individual's mental state and for his or her placement in the different recovery categories (Kidd et al., 2016; Topor, 2001).

Ongoing contact with mental health services, which can be seen as a sign of lack of recovery, may also be due to iatrogenic factors, such as drug side effects and withdrawal effects. The continued use of psychotropic drugs may also be due to many psychiatrists' unwillingness to reduce the dose of, or even to end, drug treatments, because of the spread perception that SMI means a lifelong dependence on drugs (Deegan, 1996; Harrow, Jobe, & Faull, 2012; Liberman et al., 2002; Whitaker, 2010). Torgalsbøen (2005) also argued that a continued use of neuroleptics does not contradict the possibility that a person has achieved "complete recovery" (p. 312) and referred to other medical conditions such as heart disease and diabetes. In her own study, she said that for a person to be assessed as in recovery, there should be a cutoff of using less than half of a "defined daily dose" of neuroleptics (Torgalsbøen & Rund, 1998).

The assessment of a person's condition and situation is further influenced by the changes in the institutional landscape in the provision of welfare in general, and in the psychiatric field in particular (Priebe et al., 2008; Topor, Andersson, Bülow, Stefansson, & Denhov, 2015). A significant portion of the support that many people previously had from their informal networks has been taken over by public services. Community support may not in itself be considered a sign of continued illness. In the psychiatric field the extent of hospitalization has fallen sharply, and in several countries has been widely replaced by interventions from outpatient services.

What previously were understood as psychiatric interventions are no longer always mediated by psychiatric organizations, but by municipal social services or third-sector services. Where previously a hospitalization could be seen as a sign that the person had not recovered, the need arises now to assess how the level of care in the person's own home or from other services should be viewed in relation to an assessment of his or her recovery.

The contact with various forms of support and treatment thus need not necessarily be due to continued psychiatric problems. Other factors that may influence this contact include the lack of economic means to enter society's social arenas outside the psychiatric landscape (Topor et al., 2014; Wilton, 2003), to feel "more comfortable spending time in contexts associated with mental illness," contexts in which they could have "a sense of belonging" (Kidd et al., 2016, p. 113).

Another question that has been raised when it comes to the objectivity of assessments is who assesses the recovery rate of a person. If the professionals' assessments are affected by social circumstances and culturally determined norms, it becomes difficult to argue for the existence of a value-and context-free, objective basis. Would it be impossible to replace or supplement professional assessments with the concerned individuals' assessments? "One of the clearest findings of the study was the preference of group members for patient-rated outcome measures," summarized Crawford et al. (2011, p. 343) in their results.

A final issue that should be addressed is the ability for a recovery concept to become "a new tyranny that is intolerant of 'chronicity'" (Pilgrim & McCranie, 2013 , p. 149). Recovery risks becoming a new tool in the professional's hands. The users risk being forced to participate in the planning of their own recovery, and can then be accused of not having implemented it and not reaching their recovery target.

Personal Recovery

In a context consisting of an emerging independent-user movement, inspired among other things by the struggle of Blacks and women for civil rights, and the publication of follow-up studies showing that recovery was a possibility—which confirmed the individuals' descriptions of their experiences—a different definition of recovery has been developed: personal recovery ( Davidson, Rakfeldt, & Strauss, 2010).

Personal recovery is defined and determined by the individuals themselves, unlike clinical recovery (Crawford et al., 2011). "At its heart, personal recovery is a subjective experience" (Slade et al., 2014 , p. 12), in that "the service users/survivors emphasise epistemological privileging: they are the experts because of their experience of mental distress, oppression and discrimination and recovery" ( Timander et al., 2015, p, 331).

One of the most cited definitions of personal recovery was given by Anthony (1993):

Recovery is a deeply personal, unique process of changing one's attitudes, values, feelings, goals, skills, and/or roles. It is a way of living a satisfying, hopeful, and contributing life even within the limitations caused by illness. Recovery involves the development of new meaning and purpose in one's life as one grows beyond the catastrophic effects of mental illness.

(p. 15)

Slade (2009) built on this definition: "Recovery involves living as well as possible" (p. 38).

Farkas (2007) formulated a further aspect of personal recovery, which also highlights the social consequences of a psychiatric diagnosis: the individual is "also recovering from the effects of having been diagnosed with mental illness (e.g. discrimination, disempowerment, negative side effects of unemployment, crushed dreams) as much as from the effects of the illness itself" (p. 69).

Personal recovery is associated with "being in recovery" rather than "to recover from." Davidson et al. (2005) defined it as follows: "Being in recovery instead involves being engaged in an active process of making sense of the trauma and incorporating it into one's life in such a way that its destructive impact decreases over time" (p. 10).

Personal recovery is usually defined not as a state, but as a unique and complex process. Central and recurring aspects of the literature on personal recovery are hope, human meaning, identity (Leamy et al., 2011; Slade, 2009 ), well-being, valued roles, empowerment (Farkas, 2007), healing and connection (Jacobson & Greenley, 2001; Leamy et al., 2011), and optimism about the future (Leamy et al., 2011; Slade et al., 2012).

Common to the definitions of personal recovery is that it does not "necessarily imply becoming symptom-free" (Tew, 2013, p. 361). Slade (2009) stressed the subjective aspect of personal recovery: "Personal recovery is not always about symptoms, although it is almost always about the relationship with the symptoms" (p. 43).

Often, personal recovery is about going beyond many of the experiences that are linked to serious mental health problems, such as "losing one's familiar sense of who one is and how one relates to one's social world" (Tew, 2013, p. 362). It focuses on stigma, oppression, exclusion, marginalization, helplessness, and being out of control (Tew, 2013). In this context Tew raised the recapture of power in relation to the symptoms, and the individual as a social actor, as key aspects of the recovery process. Even Davidson et al. (2005) emphasized the dimension of control. Regarding symptoms, they wrote that "the person goes from being controlled by them to bringing them under some degree of personal control" (p. 10).

Personal recovery is often described as "more of an attitude, a way of life, a feeling, a vision or an experience than a return to health or any other kind of clinical outcome per se" (Davidson et al., 2005, p. 15; see also Leamy et al., 2011). Therefore, according to some proponents of personal recovery it represents a paradigm shift in relation to notions of mental illnesses as chronic conditions or of recovery as a clinical question (Slade et al., 2014).

Criticism

Taking into account the individual's own assessment of his or her condition and situation has been described as an important step for the understanding and management of SMI (Topor et al., 2011). However, such an assessment entails a number of problems.

One such problem is the difficulty of operationalizing such a personal assessment in a research context (Slade, 2009). Assessment of personal recovery may be important for the individual's empowerment. It can also bring important knowledge about SMI, factors that may contribute to or counteract a recovery process, and the normative aspects of the recovery field, through dense descriptions of the experience of people in their social context. However, personal recovery is more difficult to manage in quantitative research, as different people can assess different conditions and situations equally and similar conditions and situations can be judged differently by different people. Frese, Knight, and Saks's (2009) conclusion of a review of the user's perceptions of their own recovery was that there is "a rather wide divergence concerning their perspectives on their recoveries and on recovery in general. . . . There is no monolithic 'consumer perspective' on many aspects of recovery" (p. 377). The need to operationalize the definition of personal recovery so that it could be studied beyond individual cases has been formulated even among the supporters of this type of definition ( Davidson et al., 2005).

Another problem is that personal recovery, paradoxically, can be construed as an acceptance of the existence of an objective and even chronic sickness. Several definitions of personal recovery are based on the person's adaptation to his or her limitations, caused by the illness. Personal recovery means, wherever possible, to live as good a life as possible within this framework. It is thus not about how illness can be influenced by a better life and possibly "cured," but about how people change their view of themselves and their lives within the frame of an ongoing illness. If the journey has no end, has the journey not become a chronic condition ( Topor, 2001)? An answer has been that the journey should be considered as life, and "life is not an 'outcome'" (Davidson, Tondora, & Ridgway, 2010 ). If the emphasis on the individual's central role in their own recovery process has been analyzed as an important contribution to our understanding of mental health problems and their development, the recovery process has also been analyzed as a social process (Borg & Davidson, 2007; Mezzina et al., 2006 ; Schön et al., 2009; Tew et al., 2012; Topor et al., 2011). The emphasis on the individual's solitary struggle that occurs in some definitions of recovery has also been criticized by people with personal experience (Rose, 2014 ).

The Relationship Between Clinical and Personal Recovery

In practice, clinical recovery could be achieved without personal recovery, and personal recovery without clinical recovery. However, they can also be intertwined ( Slade et al., 2014).

The most obvious difference between clinical and personal recovery concerns who performs the assessment of it; in one case, the professional, in the other, the individual. Often the professional assessment is presented as scientifically based and objective, and the person's as subjective. For a long time, work has been under way to develop objective criteria and scientific assessment tools for both clinical and personal recovery, but so far they have proved largely to reflect contemporary social norms and beliefs. They have not been able to prevent the individual evaluators' subjectivity from influencing the assessment, even in the case of clinical recovery.

Another common way to distinguish between personal and clinical recovery has been to describe the latter as a state and the former as a process, where concepts like "journey" have been used (Deegan, 1996; Jacobson & Greenley, 2001). It might also be possible to argue that the various definitions of recovery are about different phenomena or different aspects of the same phenomenon. Both explanations might be considered as important for understanding an individual's situation and also the phenomena of mental health problems and recovery.

Roe, Rudnick, and Gill (2007) defined recovery as a process linked to recovery as an outcome, including the illness: "Being in recovery implies, by definition, that there is something from which the person is recovering from" (p. 173). On the contrary, Davidson, Tondora, and Ridgway (2010) proposed "at least temporarily" (p. 2) to define recovery from and recovery in as two different and unconnected phenomena. Ramon et al. (2007) tried to formulate a compromise and stressed recovery as "a complex and multifaceted concept, both a process and an outcome" (p. 119).

However, both opposite viewpoints share a common ground in considering a division between an illness and the life and experiences of the person in their societal and interpersonal context. Symptoms of the illness and the well-being of the person are unclearly or not at all connected to each other.

Discussion

Different, contradictory, and complementary definitions of recovery have been formulated, and proponents of each one argue that their definition is the right one. As each definition reflects different approaches, it is improbable that a consensus will be reached. Nevertheless, different definitions will influence the studies' results, regarding both the possibility of recovery and the processes that comprise it.

Looking at the definitions of recovery and their criteria, it is clear that both main options, clinical or personal recovery, have their pros and cons. Both definitions are based on subjective and normative criteria. Both are dependent on aspects of the context outside the person's state and influence.

Different factors have been selected, defined, and assessed by users and professionals. Their respective sociocultural backgrounds naturally bias these factors, and there is no possibility of basing these assessments on any objective instruments (Cooke, 2014; Frances, 2013). Often, the factors are not about the person's state but reflect the resources at his or her disposal—the person's recovery capital (Tew, 2013).

The necessary clarification of the concepts used in research should also take into account knowledge about the impact, in terms of stigma and self-stigma, of a psychiatric diagnosis on the social relationships between the person who receives the diagnosis and his or her social context (Davidson & Strauss, 1992; Goffman, 1969; Scheff, 1984).

Consequences for Future Research

Research is based on choices. Choices have consequences for the results of the studies. The search for clear-cut categories within the psychiatric field has been going on since this field emerged, and will continue. In this context, the search for the objective definitions of core concepts in the mental health field risks making us forget that it is a social field, without reference to an objective reality somewhere out there waiting to be discovered and measured.

In this context, it would be possible to assume a reflexive position (Bourdieu, 2004; Wacquant, 2007), like the one we have tried to apply in this article:

It entails . . . the systematic exploration of the "unthought categories of thought which delimit the thinkable and predetermine the thought." . . . What has to be constantly scrutinized and neutralized, is the very act of construction of the object, the collective scientific unconscious embedded in theories, problems and . . . categories of scholarly judgment.

(Wacquant, 2007, p. 40; emphasis in source)

In the current state of knowledge, it seems important for future research not to fall for rhetorical claims about objectivity versus subjectivity and science versus experience, and instead to rank the different forms of recovery. Reflexivity, subjecting one's main concepts and choices to critical scrutiny, should not be seen as an obstacle but as an obvious part of scientific research, as such concepts and choices will influence the results.