A criticism of the use of ideal types in studies on institutional logics

Aline of research directed towards studying how the preferences, identities, values, and assumptions of individuals and organizations are embedded in institutional logics has been established within the institutional area in recent years. One of the alternatives proposed to study the logic of organizational fields comprises using “ideal types”. In this article, I seek to reflect on the use of ideal types by some authors on institutionalism in studies on the logics present in the contexts studied. I argue that there is a misreading of what is advocated by Weber with respect to ideal-type construction, ignoring the issue of axiological neutrality and the impossibility of ideal types being found empirically. In this respect, I argue that the way in which ideal types are used in studies on institutional logics needs to be reconsidered, taking these issues into consideration.


A criticism of the use of ideal types in studies on institutional logics
M ore than 30 years have passed since the emergence of classical studies which became known as "new" institutionalism. Focused on more cognitive and cultural explanations of organizational forms, this area brought legitimacy as a central aspect to individuals' and organizations' actions in seminal studies. It was proposed that actions, practices and formal structures did not only reflect technical demands and dependence on resources proposed by rational and adaptive theories, but were also influenced by institutional forces. Organizations altered their formal structures in order to gain legitimacy in their context, adapting their formal structure according to myths institutionalised in the environment, thereby avoiding the risk of being labelled negligent, irrational or even obsolete (MEYER; ROWAN, 1977). On the other hand, organizations became increasingly similar at field level. Behind this homogeneity was the action of memetic, coercive and normative isomorphic mechanisms (DIMAGGIO; POWELL, 1983).
The focus given on how organizations must conform to and be legitimated by their context led new institutionalism to be understood more as a theory of stability than a theory of change (GREENWOOD; HININGS, 1996). Where a low priority is awarded to individual actions, particularly ones concerned with changing what is institutionalized. The environment was placed as something external to organization and not influenced by this, instead of being seen as something with which the organization constantly interacts and is, therefore, also constructed by it.
Different efforts have been made in the institutional area in order to introduce dynamics to institutional studies in recent years. One example is the line of research interested in studying how the preferences, identities, values and assumptions of individuals and organizations are embedded in institutional logics. That is, the material practices and symbolic constructions through which individuals organise space and time, producing and reproducing its material substance and giving meaning to its social reality (FRIEDLAND; ALFORD, 1991). One of the alternatives proposed for the study of logics of organizational fields comprises the use of "ideal types" (e.g. OCASIO, 1999;GOODRICK;REAY, 2011), which is largely influenced by Doty and Glick's (1994) typological approach.
I seek to explore this particular development in the line of institutional studies in this article. I intend to not only revise the basic assumptions behind the idea of institutional logics, its classical authors and how this responds to the problems raised by contemporary institutional authors, but reflect on one of the main methodological tools used in the study of institutional logics: ideal types. This concept naturally refers to the ideal-type construction proposed by Weber, used to bring the researcher closer to reality. More than a methodological tool, the ideal type also carries a particular vision of the world, and a positioning about the possibilities and limits of social sciences. Its import into studies of an institutional area prompts some questioning: in what way does the ideal type in the study of institutional logics come close to Weber's construction? In what way are these concepts distanced from them? How does the use of this concept reflect in the way institutionalists position themselves on reality? What type of social science does it reflect? Can the use of typologies, as currently presented, contribute to institutional thinking?
These are some of the questions which guide this article. They highlight a longstanding debate in social sciences of the way Weber's concepts are interpreted by different authors and imported to studies in some theoretical areas. Lazarte (1996, p.27), for example, criticises the way the various readings of Weber reduced the polyphony of his work. He quotes the Weber represented by Talcott Parsons, as almost a "socio-psychologist", of Adorno and Horkheimer, as a positivist defender of the status quo, the phenomenological Weber of Schultz and, lastly, the main target of his criticism, the "domesticated" Weber portrayed by Cohn in Crítica e resignação: fundamentos da sociologia de Max Weber, "a rationalist resigned to his own incapacity of putting science to the service of substantiating universal values".
The discussion on how his theoretical precepts are imported into contemporary studies is no less important. Many of the theoretical areas currently established suffer from the influence of Weber's studies. Sociological institutionalism itself, in its seminal Guillermo Cruz studies, was inspired by Weber's work, whether to highlight how the institutions and their resultant myths influence the bureaucratisation process (MEYER; ROWAN, 1977), or even to argue that rationalisation and bureaucratisation have changed in society, resulting from processes where organizations became more similar, without necessarily becoming more efficient (DIMAGGIO;POWELL, 1983).
I would like to consider another line of studies influenced by Weberian thought here, where the concept of ideal types arises within studies on institutional logics. Initially I present Weber's proposals on social sciences and ideal-type construction. Then I revise studies on institutional logics to finally make some critical reflections on the way that ideal types are used in the study of institutional logics, where I return to some of the questioning given above.

***
The majority of Weberian formulations on ideal-type construction arise from his concerns with the conditions and limits of social sciences. They can be grouped together into some central questions: What is the purpose or even the focus of studies on cultural science? Are there objectively valid truths in the sciences which are concerned with cultural life? What is the validity of value-judgments in the scientist's activities? To Weber (1949) social sciences should be understood as those subjects that, on the whole, deal with events in human life, according to the cultural meaning that they have. It is an empirical science of a concrete reality (Wirklichkeitswissenschaft), which seeks to understand the particularities of reality, the relations and meanings involved in individual events in their expressions and the way they elapsed historically in one specific way and not another. In this aspect, it ends up also being a selective science: the researcher can only grasp a limited fragment of infinite reality as the object of scientific understanding. It will be him who determines what may be considered "relevant" or even "worthy of study," based on the meaning that the cultural phenomenon studied presents to him.
A "value orientation" is therefore assumed by the researcher, with the empirical research being intrinsically linked to the "world of values". Whether tacit or explicit, conscious or unconscious, social phenomena are selected, organized and analyzed according to individual assumptions. As Weber (1949) emphasizes, it is not possible to analyze the reality "without assumptions", since any criteria in the things themselves that allows an investigator to select a cause or aspect that deserves attention in an isolated manner does not exist. It is precisely this assignment of a "general cultural meaning" for the infinite phenomena which we are in contact with -making them interesting and meaningful to us -allowing this approach of reality to be something other than a "chaos of existential judgments". This is also one of the aspects which makes the study of phenomena in social science something particularly distinct from the naturalist tendency to establish "laws" of general validity, which allow reality to be "deducted". From the point of view of naturalism, it is possible to arrive at a knowledge of reality that is "purely objective" and absolutely rational, i.e. knowledge without any relation to values and free of any individual particularities. In contrast, when "objectivity" is questioned in social sciences, it has to be intrinsically connected to this individuality and unilaterality ignored by naturalist thinking. In other words: The objective validity of all empirical knowledge rests exclusively upon the ordering of the given reality according to categories which are subjective in a specific sense, namely, in that they present the presuppositions of our knowledge and are based on the presupposition of the value of those truths which empirical knowledge alone is able to give us [...]. The "objectivity" of the social sciences depends rather on the fact that the empirical data are always related to those evaluative ideas which alone make them worth knowing and the significance of the empirical data is derived from these evaluative ideas (WEBER, 1949, pp.110 and 111 -emphasis in the original text).
Therefore, at the time in which the individuality of a phenomenon is taken into consideration, the causal explanation of this is not so much a question of determining "general laws", but rather studying concrete causal relations. Knowledge of causal www.revistaoes.ufba.br

A criticism of the use of ideal types in studies on institutional logics
laws would not constitute the end to social sciences studies, but rather a means in the effort used in investigation to understand the empirical social reality (WEBER, 1949). With this in mind, it should not be assumed that the reality taken as a whole may have an internal order and general laws that may be able to impose a "loyalty" to it on any researcher. In social sciences, the object of knowledge is not imposed on the analysis as something already given but is constituted, through the researcher's methodological procedures (COHN, 2003, p.22). Having said that, although scientific activity is always limited and conditioned by individual values and assumptions, the authority of science cannot be called upon to impose any type of value on individuals. As Jaspers (2007) highlights, Weber is opposed to any attempt to make something that is accessible to knowledge as completely and definitively true, in order to prescribe what is correct and what should be done. Only dogmatically compelled sects would be "able to confer on the content of cultural values the status of unconditionally valid ethical imperatives" (WEBER, 1949, p.57 -emphasis in the original text). Thus, carrying out a "scientific duty" of seeing the factual truth should be separated from the "practical duty" which leads the individual to fight for his own ideals. Empirical knowledge of evaluative judgments should be distinguished, precisely, so as to avoid this type of "preaching" or even "moral inculcation" of ethical imperatives in research activities, making it explicit that "the arguments are addressed to the analytical understanding and where to the sentiments " (WEBER, 1949, p.60).
On discussing practical evaluations in academic teaching for example, Weber (1949) criticises the propagation of the professor's personal ideas in his classroom teaching activities. To him, the student should only receive something which stimulates his capacity for understanding and reasoning in the classroom; the capacity to recognise facts in reality -and suppress the unnecessary impulse of showing emotional states or personal feelings -even how to execute a given task. He stresses that it is important to oppose all types of "professorial prophecy" of placing personal judgments, particularly when there is no possibility to discuss issues from different angles, thereby preventing the student from being excessively influenced by the professor's evaluative suggestions to the point of no longer being able to resolve his own problems, according to his own principles. This is no different in research activities. To Weber, the scientist should refrain from any evaluative judgments in his analysis in social sciences since, as a scientist, it is not his responsibility to impose an imperative nature on his conclusions (COHN, 2003). Weber (1949) also highlights the position of historians in general, who have difficulty in accepting giving up evaluative judgments on political, cultural, ethical or aesthetic issues. He recalls that the empirical-historical causal chain is automatically "broken" in the majority of historical work at the moment in which the historian starts to make evaluative judgments. Such as when he tries to explain something that has been a source of ideals for the subjects studied, as the consequence of a "failure" -which is not the researcher's responsibility to judge.
The issue of "ethical neutrality" is something that should be highlighted, not only in teaching practice or research activities, but also in professional life. As Weber (1949, p.5) reinforces, on executing his responsibilities, the individual "should confine himself to it alone and should exclude whatever is not strictly proper to it -particularly his own loves and hates". However, the separation between the "empirical reality" and "essence of being" does not imply sacrificing the critical natural of scientific activity in any way. As Jaspers (2007, p.117) reminds us: The scientific obligation of seeing the truth of the facts and practical obligation of defending one's own ideals are different duties. This does not mean that the satisfaction of one is possible without satisfying the other. Weber only turns against the junction of the two things; it is only through their separation that the pure realisation of both becomes possible. There is no relationship between scientific objectivity and amorality [...]. Only the distance in relation to the object and to itself allows serene questioning of reality. But determined valuations, which are contained in the act of knowing are in turn essential conditions of the knowledge because they educate for sensitivity, in the face of all possible valuations.

Guillermo Cruz
Having dealt with these issues, the time comes for us to turn our attention to one of the central points of this article: ideal-type construction. As stated previously, an individual encounters the infinity of reality, of which only a limited percentage can be learnt. In order to come closer to this reality and capture part of this multiplicity of meanings, concepts need to be constructed -prepared in the most conscious way possible with regards to their meaning -which serve as instruments to approach this reality and see which aspects it corresponds to or is different from. Taking this into consideration, ideal types are obtained through unilateral accentuation of one or more individual points of view, in order to form a homogenous framework of thought (WEBER, 1949), where specific traces of social reality are emphasised in order to have the purest expression of this (COHN, 2003). It results from the researcher's consideration of how a specific social and historic phenomenon could have developed, if it had elapsed in a totally "logical" way, lacking contradictions. For illustrative purposes, Weber (1949) presents the example of conducting a war. If we were to try and understand how a war is conducted, we would need to imagine that there is an ideal commander in charge on both sides of the conflict, with full knowledge of everyone`s resources and all of the resultant possibilities, acting in an entirely error-free and "logically perfect" way".
An important aspect to be highlighted in an ideal-type construction is its "separation" from reality. An ideal-type cannot be "ascertained" in reality, nor can it be constructed, with the aim of ascertaining in reality how "precise" its description of the phenomena is. It is not presented as an "accurate" image of reality, through which reality may be classified, but constitutes a set of mental constructs through which the researcher comes close to or even diverges from reality. And it is precisely this distance from reality which gives the ideal types their quality as constructs, since this distancing may contribute so that the researcher may clarify aspects of the reality that he is approaching or even reveal new elements. As Weber highlights (1978, p.21), "the more sharply and precisely the ideal type has been constructed, thus the more abstract and unrealistic in this sense it is, the better it is able to perform its functions in formulating terminology, classifications, and hypotheses".
Therefore, care needs to be taken with the meaning that the word "ideal" has here. The frameworks of thought prepared by the researcher need to be taken as "ideal" in the purely logical sense and never in the sense of "exemplary" (WEBER, 1949). The ideal type is a framework of thought prepared by the researcher in his attempt to analyse "unique historical configurations" and their individual components through generic concepts, which do not represent historical reality in any way, nor are they a scheme through which an action or situation may be classified as an example. Weber (1949, p.97-98) draws attention to constructions of utterances made in the sense of revealing the "essence" of a phenomenon of social reality (e.g. the "essence of Christianity"). These attempts try to present a historical "portrait" of empirically existing facts, which makes them have a relative and problematic validity, in terms of ideal types, also introduced as ideal types in the practical sense. That is, they contain judgments of what that phenomenon should be to the scientist, that is "essential" to him, that directly or indirectly, consciously or unconsciously expresses specific personal ideals. Presented in this way as "exemplary types". This returns once more to the question of neutrality in scientific activity and the importance of the researcher being careful not to transform his research, his effort to understand reality, into an act of preaching. As Weber (1949, p.98 and 99) reminds us: [...] the elementary duty of scientific self-control and the only way to avoid serious and foolish blunders requires a sharp, precise distinction between the logically comparative analysis of reality by ideal-types in the logical sense and the value-judgment of reality on the basis of ideals. An "ideal type" in our sense, to repeat once more, has no connection at all with value-judgments, and it has nothing to do with any type of perfection other than a purely logical one..

***
Now the basic principles on Weberian ideal-type construction have been presented, as well as its basic assumptions, attention can now be given to another central priority of this article: studies on institutional logics. They originate from www.revistaoes.ufba.br Friedland and Alford's (1991) seminal article that suggested the existence of a central logic in each of the institutional orders existing in contemporary Western societies, such as the church, family and market. This central logic or, simply institutional logics, represents a set of material practices and symbolic constructions that form the basic organizational principles for each of these institutional orders. The institutional logic of capitalism, for example, would be the accumulation and commercialisation of human activity. In turn, state logic involves the rationalisation and regulation of individuals' practices through legal and bureaucratic hierarchies (FRIEDLAND; ALFORD, 1991).

A criticism of the use of ideal types in studies on institutional logics
For Friedland and Alford (1991) there was a need to reconceptualise the meaning of "institution" in institutional thinking, in order to consider the relationship between the symbolic and practical aspects of organizational life. Up until that time, the understanding of institutions was principally directed towards material aspects: institutions were mainly seen as a set of norms or standards existing at the supra-organizational level. The focus of these studies was directed precisely on the "homogenising nature" of institutions, with analysis directed towards subjects such as legitimacy, structuring and homogenising the fields and mechanisms to maintain order both within and outside organizations (e.g. MEYER; ROWAN, 1977;DIMAGGIO;POWELL, 1983). However, various gaps have emerged from this focus on stability in institutional studies, such as the issues of change and conflict, which could not be explained by this bias. Amidst this context, Friedland and Alford (1991, p.243) proposed the idea of institutional logics, based on the concept of the institution as being "simultaneously material and ideal", representing both the activity patterns through which individuals produce and reproduce their material substance and the symbolic systems which contribute in order to make it a significant experience. Thus, while institutions limited individuals' actions -serving as a base for structuring organizational fields -, they also provided opportunities for changes to the institutional context: through the contradictions inherent to the multiple institutional logics which the actors come across, supplying them with a range of "cultural resources" to govern their actions (THORNTON; OCASIO, 2008).
The idea of institutional logics continued to be explored by various authors following Friedland and Alford's (1991) introductory contributions. The majority of the subsequent work continued to recognise the existence of multiple logics in the reality in which the actors that they studied were immersed. Curiously, however, it was not uncommon to observe a tendency among the authors to put forward a logic as predominant. A rapid analysis of the studies along the lines of institutional logics 1 shows two forms of approach which were commonly adopted: on one hand there are authors who consider the way that a logic predominates in a specific field, being a "dominant" logic in this and serving as a guide for social actors` behaviour (e.g. OCASIO, 1999;HININGS, 2005); and on the other, there are those who accept the co-existence of multiple logics for extensive periods of time, without there necessarily being a predominance of one over the others (e.g. DUNN;JONES, 2010;GOODRICK;REAY, 2011). The studies which are accommodated in the first case tend to explore the way how a logic, dominant in a specific period, was "replaced" by another -disappearing or otherwise. Reay and Hinings (2005), for example, focus on the re-composition of the organizational field of health in the city of Alberta, Canada in which two institutional logics compete: the logic of professionalism and corporate logic. There was a change in the structure of the field and its dominant logic over time, with corporate logic prevailing. The change is strongly influenced by government actions to restructure the field, by inserting a new actor -regional health authorities -, which reflected the government's desire to have a new dominant logic in the field, based on efficiency, effectiveness, customer service and corporate processes. Although a new logic emerged in the field following government actions, the old, dominant logic -the logic of professionalism, based on valuing the doctor-patient relationship and a quality medical service -is not eliminated, only subjugated, continuing to be important in the field.

Guillermo Cruz
The second line of studies on institutional logics also turns its attention to different periods in a specific field, but visualising how elements from different institutional logics co-existed in a specific period, having a higher or lower preponderance in its social and historical context at different times. The study carried out by Dunn and Jones (2010) is along this line, where two logics are presented -one of science and the other of care -which co-exist in the field of the American medical schools studied by the authors. The science logic focused on knowledge of the diseases which is constructed through research and innovative treatment, while the care logic gives greater attention to the clinical skills used by the professional to treat patients and improve health levels in the community. Making use of a historical narrative of the main events which have taken place since the first decade of the last century and a quantitative analysis where vocabularies, keywords and practices that underwent alterations over the years are explored, Dunn and Jones (2010) suggested that both of the logics are present at the different times, and although there is a slight predominance of elements of some of these, they are interlaced for most of the time: the topic of manage care focuses on the cost and efficiency of medical coverage; that of public health promotes community health based on a statistical analysis of clinical results and medical schools and, in the process, simultaneously emphasise the care that is centred on the patient and based on scientific research.
The authors used different research strategies to study logics, the majority combining qualitative and quantitative analyses, utilising ethnographies, case studies, content analyses and regressions, etc. However, attention is drawn to the use of "ideal types" by some authors. For example, Thornton and Ocasio (1999) analyse the relation between institutional logics and power structures in the higher education publishing industry, where they observe a shift in the dominant institutional logic in an industry from the logic of professions to the logic of markets. In editorial logics, the publishers' means and ends were influenced by the editor-author relationship and power structures were determined by the size of the organization and its structure. In market logics, the means and ends are based on competition for resources and acquisitions, while the power structures came to be influenced by competition on the product market and the search for corporate control. In their approach to the reality studied, Thornton and Ocasio (1999) initially sought to identify the logics prevailing in the industry and specify how these change over time. In order to make this identification, they conducted interviews with actors in the field (editors, investors and association managers), and historic research and, from this information, prepared their "model", involving the two logics cited. This model is centred on a set of nine characteristics: the type of capitalism, organizational identity, legitimacy, authority structures, the mission, focus, strategy used, investment principles and rules for succession.
In order to prepare the two "ideal types" of logics, Thornton and Ocasio (1999) follow the assumptions developed by Doty and Glick (1994) on the use of typologies to construct and model theories. In their article, Doty and Glick (1994), highlight that constructing typologies has been criticised for developing classificatory systems and not theories. In response to these criticisms, they made a distinction between typologies and classificatory schemes. While the schemes categorise the phenomena in mutually exclusive sets, through a series of decision-making rules, the typologies dispense with any set of decision-making rules to categorise aspects of the reality observed. They identify ideal types, representing "a unique combination of the organizational attributes that are believed to determine the relevant outcome(s)" (DOTY; GLICK, 1994, p.232). They quote the five types of organizational structures proposed by Mintzberg (1979;1983) to attain maximum organizational efficiency and the three ideal types of strategies identified by Porter (1980Porter ( , 1985 to maximise competitive advantage as examples. When talking about the question of the similarity of the ideal type with reality, Doty and Glick (1994) confirm that examples of organizations which correspond to the ideal type are very rare, so that the organizations may be more or less similar to the type, but normally cannot be specifically associated to one of the ideal types.
In their study of the pharmaceutical profession, Goodrick and Reay (2011) also employ the use of ideal types, following a methodological line similar to that adopted www.revistaoes.ufba.br

A criticism of the use of ideal types in studies on institutional logics
by Thornton and Ocasio (1999). They evaluated literature on institutional logics and the study of professions in order to identify each of the four ideal types of logics (state, market, corporate and professional). They carried out a review of historical data in the last 150 years (consulting books and journals in the area and a record of actors' impressions in the field) and they compared them with the ideal types (developed with the assistance of a theoretical framework) where nine categories are identified for each of the types. Having "closed" the ideal types for each logic proposed, they set off to evaluate the "strength" of each of the characteristics in the different "historical eras", attributing values from 1 to 5, according to the "proximity" of the element developed in the type, with the historical records of the context analysed in that specific period. Making a simple average between the categories, Goodrick and Reay (2011) reached a value which represented the "strength" of the ideal type developed to explain each of the time bands considered. Considering the variations in the values of each ideal type over time, and the alternation between each of the types in the ranking considered for each era, the authors concluded that no logic was predominant, but there was a co-existence of multiple logics, in a "constellation of logics", with one or more logics being manifested in a more pronounced way in each era.

***
Having arrived at this point, and given a brief demonstration of some of the logics studies in institutional analysis, I now return to the questioning posed above, in the introduction to this article: do the "ideal types" developed by the institutional authors come close to or move away from the Weberian ideal type? In which way can the use of typologies contribute or otherwise to the development of this theoretical branch? Generally speaking, when considering the way that some authors from institutionalism make use of the ideal-type construction, a dangerous inclination towards verificationism and prescritivism can be implicitly or explicitly observed. On one hand, it is assumed that the ideal type may correspond with reality and also that it could be "verified" in some way how much of that ideal type corresponds with it in reality. On the other, it tends to place the ideal type as a kind of "example", as a reference point to be reached by social actors. Thornton and Ocasio (1999) construct their ideal types from reality, at variance with the path taken by Weber (1949) and never contrast the type developed with reality in another way that is not through the perfect correspondence between the ideal type and reality. No effort is made by the authors to think about in what way the type constructed also moves away from what is observed in reality, nor about how this may contribute to reconsidering the understanding about that social and historical context and the ideal type of institutional logics used. The ideal type ceases to be an instrument with which the researcher comes close to reality, in order to capture the multiplicity of meanings which exist in it, to become a type of summary of the authors' impressions, placed in a representation of the "essence" of the institutional context.
In turn, Goodrick and Reay (2011) try to do something slightly different, at least with regards to constructing their ideal types. They do it from the theoretical framework that they had previously consulted, to then face the reality or records of it. Although there is a slightly greater effort by these authors to try and make use of the ideal-type construct as an instrument to approach the researched object, up to what point comparing the type constructed with records can be a successful way to learn about the material practices and symbolic constructions of that reality (basic idea of "institutional logics") is questionable. The analysis is mostly being directed to interpreters than protagonists of that socio-historical context. So, are the authors capturing the "logic" of the reality existing in the past or impressions and valuejudgments by individuals who were in contact with it? Even if we ignore this question, other aspects remain. Like Thornton and Ocasio (1999), the authors do not explain in which aspects the constructed types move away from reality. Nor do they present the "original ideal type" -originating from a "pre-analysis" of data -, and how it was altered or otherwise after this approximation with the object. Or also what were the changes in the authors' understanding after being in contact with the data.

Guillermo Cruz
However, of all the questioning and criticism that could be directed to the way that ideal types are constructed and used by Goodrick and Reay (2011), none capture our attention more than the way the authors try to evaluate the "strength" of every construct developed. There is an effort directed much more towards ascertaining the similarity of their proposals with the reality than understanding the true multiplicity of what they were studying. Again, we have a very distorted use of the ideal-type construction proposed by Weber. Strangely, all of the authors of institutional logics that I criticise above quote Weber's work in their articles. But it appears that they were not concerned with looking for texts which discuss the methodological question in more depth.
The problems are no smaller in Doty and Glick's (1994) proposals on the use of typologies, which serve as a basis for analyses carried out by scholars of institutional logics. The typologies which they propose served to construct hypotheses of "relationships between the similarity of an actual organization to an ideal type and the dependent variable(s)" (DOTY; GLICK, 1994, p.234) and in some cases -which are rare, but possible, as the authors remind us -can be observed in reality, in individuals' and organizations' behaviour. This tendency to believe that the typology developed may correspond with reality may partly explain the attitude of the institutionalists criticised above. The way how the typologies proposed by Doty and Glick (1994) may take on the condition of "models" or "examples" to be followed to obtain something that may be classified by the researcher as "appropriate" or "positive" should also be highlighted -as noted in the suggestion by the authors of Porter and Mintzberg's work as examples of typical-ideal forms. Curiously, Doty and Glick (1994) make use of Weber's (1949) precepts to construct their definition of the ideal type, remembering that they are a unilateral accentuation of one or more points of view which may be a guide to formulate hypotheses. However, they appear to make an erroneous reading of that defended by Weber, based on what the most advantageous solution is for them, ignoring the question of axiological neutrality and the types` non-attachment with reality.
This reflection on the Weberian ideal-type construction and the way that the authors have made use of ideal types in studies on logics, although brief, makes us question the routes taken in institutional thinking. Without a doubt, the idea of institutional logics -particularly in the way proposed by Friedland and Alford (1991) may contribute to institutionalists' intentions to bring dynamics back to their studies, making it possible to consider the conflict presented inside institutional fields and allowing a change to what is institutionalised. However, the way that recent studies along this line have made use of ideal types needs to be reconsidered. The time has come for institutionalists to reflect on the tools that they are proposing to study reality: how are they making use of their references? When are they getting close to the authors that they quote? And, particularly, when are they moving away from them? Clearly, there seems to be little or no attention by some authors in the institutional logic vein of studies to distinguish constructs that use the "ideal types" proposed by Weber. They need to do the same as Weber when he introduced the idea of the ideal type as a proposal for social science to perform their empirical social reality studies and that is to make clear what they do and do not want to say by "ideal". At this point I think that could be as important as it was for institutionalists to question the excessive focus on stability and also question the methods which accompany their new proposals.
And if it is to follow what is proposed in the Weberian ideal type, then at least we expect that this could be done exactly as Weber (1949) suggested, with the appropriate care given to questions such as value-judgments. The use of ideal types in their studies could be a major contribution to authors using the institutional approach and for other theoretical lines, particularly if we consider it as a tool which is prepared to face reality, both for its similarities and differences in relation to it. Accepting the impossibility of learning the infinity of aspects of the real world and understanding its tools as distinct from reality may help the authors in this field to perhaps reconsider what they think about institutional logics. Is there effectively a logic (or more) behind that reality, which may explain its operation in some way? Is there a "logic" inside the institutions, pervading the material and symbolic aspects which are socially shared and constructed by individuals and organizations? Would another term be more appropriate? These are just a few of the questions which remain to be answered in this field but are central to thinking about the course that it will take.