Abstract
In this paper, the author introduces a post-Meadian concept useful for contemporary ethnography. Emergent ideal types capture the imaginary (that which is not bound to space and time coordinates) in the doing ethnography and making empirical claims (the empirical is considered that which is limited to space and time coordinates). She uses a heuristic device, grounded narrative fictions, as one way of depicting the imaginary for ethnographic audiences. The exploration draws on the ethnography of a kindergarten/Grade 1 class. The author does not use many scholarly references. Instead, she tries to develop two innovations to ethnographic analysis (emergent ideal types and grounded narrative fictions) from within an ethnographic example. She tries in this thought piece to explicate the linguistic mechanisms through which ethnographers draw on the imaginary to link empirical patterns with interpretive analyses.
There is a leap of faith in social science (Kierkegaard, 1992) that involves a non-empirical imaginary: that which in any instance is not accessible through space and time coordinates alone. Let us say that Schutz (1966) has a cube in front of him (the empirical) and that he is imagining a variety of other possible cubes (the imaginary). He could imagine an infinite number of varied cubes. However, as he imagines these various cubes, he is not directly thinking of the characteristics common to all imaginable cubes, such as rectangularity. Schutz referred to this “set of characteristics, unchanged among all the imagined transformations of the concrete thing perceived—the kernel, so to speak, of all possibly imaginable cubes” as the eidos of the cube (p. 114). Schutz was capturing the imagined possibilities in the empirical. Social science is constituted of such leaps from the cube before us in space and time to the eidos of the cube and back again to the cube. This is not peculiar to social science (I recently started reading a history of algebra the title of which includes “the imaginary and real”), but in this paper I will refer to the context of social science exclusively.
In this paper, the imagined and the empirical will both be thought of as intersubjectively structured (Carspecken, 1996). Following Schutz (1966) and others, the empirical will refer to claims that are dependent on space and time coordinates and are, in principle, accessible to multiple witnesses (Habermas, 1981). I will refer to these empirical claims as the discourse of the real. The imagined will refer to paradigmatic structures, mental configurations, concepts, inferences, and so on that are not constructed through a one-to-one dependence on space and time coordinates, although the imaginary might make reference to space and time. In other words, the imaginary is not a mere reflection of that we presume to be real. It is something altogether different yet related. If both the leap of the imaginary and its contingent, empirical claims are intersubjectively structured, examining the intersubjective structures that are involved in both would help us better articulate and explain social scientific leaps back and forth. This interesting thought project has practical potential for qualitative methods.
In this paper, I make a twofold contribution to qualitative research methodology. First, I specify a substantive concept (emergent ideal type) that links intersubjective structures of the empirical descriptions with intersubjective structures involved in explaining the relation between the empirical descriptions and the imagined. Second, I demonstrate the usefulness of the concept for ethnography by applying it to a specific empirical example. Here we find the imagined in the discourse of the real.
In agreement with Habermas (1981, 1987), it is expected that research investigations into actors' lived experiences reconstruct in a radical way what is already internal to participants' lived experiences. Thus, it is expected that the imagined in ethnography reconstructs from the imagined in everyday life. Theoretical concepts that articulate the imagined would therefore be substantive and not merely methodological because they would simultaneously refer to both a substantive aspect of understanding as it ensues in everyday life and a methodological tool to aid researchers in putting the implicit into discourse.
This paper is largely creative, in that it exemplifies the creative aspect of doing social science. It is also intuitive, in that I have sought to locate substantive concepts through ethnographic exploration, radically turning inward to understand something that might be relevant beyond that context. There are four major sections: (a) The Discourse of the Real, (b) The Imagined in Social Science Methodology, (c) Emergent Ideal Types: A Post-Meadian Concept, and (d) Capturing the Imagined in the Discourse of the Real: An Ethnographic Example. Although the last section of the paper is devoted to exemplifying the points through a particular ethnography, I use examples from that same ethnography in earlier sections of the paper.
The discourse of the real
I am using the phrase “the discourse of the real,” which demands some definition and explanation. I introduced this phrase to refer to the empirical: that which is contingent on space and time coordinates and is usually accessed through the senses. This point calls for some clarification. Postmetaphysical ideas of empiricism involved a discrete subject-object distinction. Objects were thought to exist outside and separate from subjects. Truth was the closest representation in discourse to the natural existence of the object prior to and outside of the subject who was investigating it. That which was assumed to exist external to the subject and was available for study by multiple subjects was considered “real,” but social reality was not as easy to describe because it was difficult to make a definitive line between subject with social reality as object. In this section of the paper, I review three social science appropriations of “reality.” These appropriations reflect contemporary debates on the topic of reality. First, I review Bhaskar's (1986) critical realism. Second, I review Harré's (1998) pragmatic realism. Last, I review Habermas's (1981, 1987) presupposed realism.
Bhaskar (1991) claimed that “reality can unequivocably (and no longer anthropocentrically or epistemologically) be accorded to things” (p. 27) themselves. According to Kivinen and Piiroinen (2005), Bhaskar's critical realism assumes that the objects of study exist independent of the observer's claims about them. “Only by means of presuming that there are intransitive objects, scientific knowledge can be seen as progressing toward ever more accurate descriptions of the human-independent world, as becoming closer to reality as it is” (p. 232, italics added). Critical realism makes strong ontological claims. The study of social life, for Bhaskar (1986), involves the study of social structures that are real entities, although they are indirectly accessible to social science through causal effects. In other words, structures exist in time and space through their causal effects. Nevertheless, the idea of social structure is meaningful only if it refers to some significant “feature of reality” (Bhaskar, 2001, in Kivinen & Piiroinen, 2005, p. 233), the significance of which is empirically assessed through causal relations. Bhaskar's claims about reality directly inform his criticalism: If we are to act on reality with critical intent, we must be able to identify what is real (Bhaskar, 2002). He might ask, for example, if we cannot say that poverty is real given its effects, then on what grounds would we commit ourselves to changing the conditions of poverty? If we cannot say that gender is real, then on what grounds can we describe things as unequal based on gender? For Bhaskar, there is a singular reality that exceeds and precedes its observers. Reality is not determined by various perspectives; rather, various perspectives can increase overall understanding of the real world.
Bhaskar's (2001, in Kivinen & Piiroinen, 2005) notions of reality depend on the idea that for a single phenomenon or object of interest there would correspond a reality. Putnam (1990) pointed out that the causal criterion used by Bhaskar for locating social structures in the empirical world requires acts of choice on the researcher's part. Those acts of choice cannot be explained or reduced to the system of objects themselves. To fill in, Kivinen and Piiroinen argued for a more pragmatic rendition of reality, namely that “causes can only be seen from an actor's point of view, in respect of this or that purpose” (p. 233), and therefore reality must also be mediated by perspective and intentionality.
For pragmatists, actors meaningfully use social conditions, structures, and so on both intentionally and unintentionally. Reality is explained in terms of “use” for theorists like Harré (1998) and Habermas (1981, 1987). Harré, who was described by Kivinen and Piiroinen (2005) as a pragmatic realist and by Krausz (2000) as a constructive realist, claimed that reality exists so long as people believe in it. In other words, Harré contended that there is a world of stuff independent of interpretive activity (schools, for example) but that this world is known to humans only through the way in which it is ascribed. Furthermore, all ascribable properties are interpretation dependent. Thus, there would be no necessarily singular ontological reality independent of human need. Harré (Harré & Bhaskar, 2001) proposed that our commitments to realities were practical-moral ones. Kivinen and Piiroinen called this realism with a small r. Ontologies would be justified according to the extent to which they help understand phenomena in a useful way. From this perspective, multiple realities (generated through use and need) are the most plausible conceptualization for reality because different needs and uses will concomitantly generate multiple realities (notice how the theory itself resorts to a single definition of reality).
Habermas's (1981) pragmatic view of reality would also be considered realism with a small r. For Habermas, reality is presupposed and indicated through ontological claims embedded in truth claims. What we end up with is a more tamed version of multiple realities, because Habermas posited three specific forms of truth claims, each of which presumes a particular kind of ontology. The presumed or presupposed reality embedded in the claim becomes part of the way in which conversing humans determine whether the claim holds up. Each of the three types is linked to one of three ontological presuppositions. For objectivity, actors presuppose an external, objective reality (much like Bhaskar's [2001, in Kivinen & Piiroinen, 2005] reality). We make reference to this reality when we talk. We also make reference to a subjective world or reality, and we make reference to a social world. Multiple realities, to this way of thinking, can be organized according to three formal systems of reference or presupposition. We must assume reality to talk, exchange, improve life, and so on. Each of these three ontologically presupposed realities is implied in establishing the validity of the claim and in understanding the claims themselves. I interpret this to mean that we have ways of discussing reality though which a variety of experiences and perspectives are dialogically situated with a co-arising possibility for multiple realities, all of which are reciprocally interdependent (not free floating) on the things said about them.
One important point that all three of these theorists raised is this: Social reality itself is not something of direct experience. It is, rather, socially mediated. For this reason, I use the phrase “discourse of the real” with the idea that “discourse” indicates the social mediation necessary to talk, think, and act on what is real. In line with Habermas (1981), however, I am trying to suggest not that reality is an ontology-first concept but, rather, that reality is pragmatic supposition to any truth claims. There can be multiple such suppositions. Empirical claims presuppose the very particular reality of objects and phenomena in space and time.
The imagined in social science methodology
The imagined is paradigmatically and intersubjectively linked to actions that can be described with space and time coordinates, or what I will also refer to as the discourse of the real (reality here is a presupposition to the claims) (Habermas, 1981; McCarthy, 1994; Tugendhat, 1986). Schutz's (1966) ideas help us see this.
Schutz (1966) distinguished between two processes of imagining: eidetic universalization and empirical universalization. According to Schutz, empirical universalization is an inductive process “not only contingent in the sense that [its] formation starts from a particular given contingently in factual experience, but also in the sense that the conceptualization proceeds on the ground of comparison with likewise contingently given similarities” (p. 80). In statistical analyses, measures of central tendency are examples of empirical universalizations. Eidetic universalization, on the other hand, grasps the “invariant content, the inner structure, without which the object could not be the kind of object it is” (p. 81). Continuing on, Schutz claimed that “this invariant element prescribes the limits to all possible variations of the same prototype; it is that element without which an objectivity of this kind can neither be thought nor intuitively phantasied” (p. 81). New statistical methods involving factor analysis are more capable of mapping out eidetic universalizations. Schutz's ideas were not explained through a theory of intersubjectivity; rather, his work emphasized “experience” as the ground for empirical perception.
I started this paper by calling what Schutz (1966) was holding “a cube,” which referred to something that he might really have been holding. The use of the name to identify the object requires a leap. Wittgenstein (1969) argued that words and names are not unambiguous, even when defined through a set of discrete attributes or essence of the thing named. He proposed instead that meaning ensues from the way in which words are used. Use indicates a series of similarities, or “family resemblances.” These family resemblances work off the idea that the link between empirical phenomena and meaningful paradigmatic interpretations is linguistic and is based on a primary dialectic of similarity and difference.
Wittgenstein (1969) also argued that these mental, linguistic associations were the product not of a private, idiosyncratic process but, rather, of a social one (Tugendhat, 1986), rendering them always, at least in principle, understandable and defensible to others. Accordingly, family resemblances, concepts, and other ways of talking about various theories about the imaginary, are assumed here to be intersubjetively structured, even when private. This intersubjective quality makes Wittgenstein's family resemblances accessible to the interpretive social sciences. We can read Wittgenstein to imply that intersubjectivity and pragmatics form the ground for understanding what we can observe rather than experience (as Schutz [1969] had posited). This is important to keep in mind as we assess the following ideas.
Types, metaphors, models, and concepts are some of the many different means by which social scientists have leapt to something beyond the empirical. We do this in everyday life as well. For example, if I report the simple claim that there are six chairs in the room, this claim implicates, among other things, the concepts of zero, chair, and in. Moreover, the idea of “chairness,” for example, does not exist in space and time but is nevertheless imagined, or is part of something that is imagined, when I claim that there are six chairs in the room. The claim itself also implies how it would be tested empirically. In this way, the imaginary can assume a central position in making sense of qualitative data and in the construction of ethnographic texts. It is also tacitly assumed in the analysis and report of quantitative data (Bhaskar, 1998; Torgeson, 1958). In this section of the paper, I examine a few of the varied social science conceptualizations of this imaginary. The purpose is not to explore all such possibilities but to note a few of the ways in which the imaginary has been conceptualized as intersubjective and social so as to mark a path toward the concept that I will be developing in this paper: emergent ideal types. Notice that I am now talking about concepts of concepts.
Below I describe four examples through which I seek to articulate the imaginary in social science, namely Weber's (1949) ideal types, Dexter and LaMagdeleine's (2002) use of metaphors, Lakoff and Johnson's (1980, 1999) ideal prototypes, and Habermas's (1981, 1987) concept of the ideal speech situation. Of the four, Weber's ideal types and Magdeleine and Dexter's metaphors are most specifically about doing social science. Lakoff and Johnson's ideal prototype postulates a conceptual connection between ideal and typical concepts as metaphors within how folk make sense of their everyday experiences, although this is not a distinctly social science concept. For each of these three, the imagined is produced as a type of concept which differs from some real, observable manifestation. This contrast of the imagined to the real is fundamentally different from Habermas's use of the term ideal.
Weber's ideal types in the conduct of social science
Weber's (1949) ideal type is a formulation reconstructing implicit ideals in empirical phenomena from which all specific empirical phenomena will deviate to various extents. An ideal type, therefore, will not exist empirically in the sense that one would be able to find empirical “examples” that conform to and thereby duplicate the ideal type. For Weber, ideal types were a methodological tool useful in the analysis of data, especially historical data.
According to Berger (1976), Weber described four different kinds of ideal types, all of which worked off of a couple of key principles. First, they were derived inductively from across multiple examples of similar cultural phenomena. Ideal types do not manifest as aggregates of characteristics; instead they include characteristics that might have manifested regularly across the types if not for some disturbing influences; that is, empirical phenomena were thought to have these characteristics in common under ideal situations. These ideal situations did not exist because of the presence of actual disturbing influences. Those disturbing influences, whatever they were, disallowed the genuine pureness of a phenomenon to manifest empirically. Berger wrote, “The source of the dissimilarity [between the concept and its empirical counterpart] is to be found in the fact that the situational conditions vary and that the inner states (treated as motivational) are complex” (p. 125). This idealization represents not what the phenomena have in common but, rather, what they would have in common when liberated from other influences.
Weber (1949) proposed that social science researchers could develop these ideal types as a method of capturing, in an exaggerated and one-sided way, the essence of some cultural phenomena. Once constructed in discourse as an ideal type, the researchers could bring these ideal types to bear on the empirical world. Through this way of thinking, the empirical world, of course, would not measure up because of the presence of distorting features within the culture. For Weber, the essence of ideal types was radically imaginary and external to the real.
An example of one of Weber's (1949) famous ideal types is the bureaucracy, where no single empirical example of a perfect bureaucracy exists in the empirical world, but, rather, each approximates a perfect ideal bureaucracy. Weber did not connect this heuristic social scientific tool with the interpretive processes in which people intuitively engaged while making sense of everyday social phenomena, for example in their everyday interactions with bureaucracies.
Dexter and LaMagdeleine's metaphors in qualitative research
Dexter and LaMagdeleine (2002) explored the use of metaphors in qualitative research. They argued that a researcher's attempts to both understand the essence of cultural phenomena and bring out that essence for others involves the use of metaphors. According to Dexter and LaMagdeleine, metaphors can be both generative and reflexive. Metaphors exaggerate or foreground sets of characteristics or components for particular purposes. Thus, researchers can talk about points at which a metaphor fails to work in contrast with the more successful aspects of a metaphor-choice. Multiple metaphors can be applied to the same set of data. Those multiple metaphors indicate varying positions through which particular interpretations can be rendered. Such metaphors are developed through a mapping process that involves (a) locating best fit, (b) heightening awareness of saliency and distribution, and (c) grasping form and content relations. This process results in the articulation of an image or picture that is meant to convey a whole set of internally meaningful relations by developing an external construct that would simultaneously deliver the essence of those relations. Accordingly, metaphors are derived by researchers in order to deliver complicated relations of findings. The metaphor is an articulation of the imagined.
What the metaphor articulates is the imagined essence of cultural phenomena, which is neither what it is empirically nor what it is metaphorically. The imagined exists in the space between these. In this way, metaphors can both generate understandings of data and reflect understandings of data. They summon meaning for the reader.
It is interesting to compare Dexter and LaMagdeleine's (2002) use of metaphors with Weber's (1949) ideal types with respect to ethnography. The primary similarity is that metaphors and ideal types in these specific contexts are developed inductively, suggesting that the relationship between the imagined and the empirical is inductive. Furthermore, both metaphors and ideal types are heuristic for social science. There are important differences across these concepts. The metaphor does not attain the status of measuring post; it merely meaningfully delivers aspects of the imagined. Although metaphors fit various empirical phenomena to certain extents, they fit on the level of invoking an imagined set of internal relations similar to those discovered empirically; whereas ideal types are articulations of ideal, imagined versions of the cultural phenomena and can be used directly to locate where, in what way, and to what extent empirical phenomena vary. Metaphors, on the other hand, help to hand over the essence of the phenomenon in its complexity without emphasizing the variability of empirical phenomena. Weber's ideal types are imagined, but metaphors (as talked about by Dexter and LaMagdeleine, 2002) are expressions of the imagined. For Dexter and LaMagdeleine, metaphors are explicitly described as intersubjectively structured, and this is why they can work as expressions of the imagined.
Let us compare the two using Schutz's (1966) distinction between eidetic and empirical universalizations. Barber (1988) interpreted the difference between eidetic and empirical universalizations as one of degrees. This interpretation would say we move toward the eidetic universalization by first inducing empirical universalizations. Both Weber (1949) and Dexter and LaMagdeleine (2002) would conform to this interpretation of Schutz: The use of induction leads eventually to grasping the essence, or eidos, of a phenomenon, which can be expressed through a metaphor (in Dexter and LaMagdeleine's case) or can produce an ideal type (in Weber's case).
Lakoff and Johnson's metaphors and prototypes
Dexter and LaMagdeleine (2002) wrote about metaphors as heuristic tools for qualitative researchers, but Lakoff and Johnson (1980, 1999) talked about metaphors as necessary for the ways in which people make sense of their own experiences. These metaphors are more organic than the ones Dexter and LaMagdeleine described. Actually, Lakoff and Johnson developed a strong theory of metaphors, which would include the kind that Dexter and LaMagdeleine wrote about, with more organic metaphors as primary. An example of an organic metaphor is force, whereby the body experiences force physically and comes to conceptualize aspects of the nonphysical world metaphorically in terms of force (Lakoff & Johnson, 1980). Lakoff and Johnson called these embodied metaphors. The imagined is the form the physical experience takes as something else. These organic metaphors form crucial systems of meaning that are invoked through actors' everyday social intercourse. They are intersubjective with an experiential basis.
Lakoff and Johnson (1999) also wrote about prototypes, contrasting ideal prototypes with typical prototypes by using the examples of the typical husband (a typical prototype) and the ideal husband (an ideal prototype). Lakoff and Johnson asked readers to conjure up the difference in imagery between these two types.
Neither prototype is meant to be taken as an actual empirical possibility (although for different reasons). The typical prototype (Lakoff & Johnson, 1999) is a generalization of aggregates across empirical examples. This typical prototype is similar to Schutz's empirical universalization, whereas the ideal prototype is more like Weber's (1949) ideal types (what the phenomena would be like without the interference of distorting features). Imagining the ideal prototype involves thinking of this as a flawless or perfect version of the prototype, whereas imagining the typical prototype involves imagining characteristics of the empirical types deduced with their distorting influences.
The ideal in Habermas's ideal speech situation
Thus far, we have seen the word ideal connote something like perfect, pure, or undisturbed, which misses the interpretive generativity that we find in Habermas's (1981, 1987) use of the word ideal in reference to the ideal speech situation. Habermas provided a more directly intersubjective, presuppositional use of the term ideal. For Habermas, the ideal speech situation is both presupposed and counterfactual in all attempts toward reaching understanding. The ideal speech situation is the description of conditions under which communicatively achieved understanding is pursued, even fallibly; namely, conditions of reciprocity and symmetry: that all participants or participants' perspectives are given equal opportunity for expression and challenge in any communicative effort aimed at understanding. The ideal speech situation is a condition presupposed in understanding and in reaching consensus.
Often enough Habermas's (1981, 1987) ideal speech situation is interpreted as if it is an idealistic belief in what might be possible within actual speech situations. This misinterpretation results in people thinking that Habermas's theory is very unrealistic and renders the ideal speech situation as something akin to either (a) Schutz's (1966) empirical universalization as an empirical universalization of speech situations or (b) Weber's (1949) ideal types, where the ideal speech situation is the situation that would obtain in the empirical world if all distorting elements were eliminated. This last interpretation is inaccurate because it is insufficient. Habermas has argued that the ideal speech situation is usually counterfactual (McCarthy 1994). This conception, when taken alone, leads people to limit their understanding of a Habermasian critique. Through this limitation, Habermas's speech situation is no more than Weber's ideal type applied to communicative contexts. The misinterpretation and limitation are the effect of a similar theoretical mistake in understanding how the ideal speech situation is naturalistically and philosophically connected to the empirical world. Given the more common use of the term ideal, these mistakes are understandable; however, they are also worth clearing up.
According to Habermas (1981, 1987), the ideal speech situation is presupposed by any communicative act precisely because it is the way in which intersubjective sensemaking can be constituted. The ideal speech situation is a force in the holistic impetus to even act meaningfully (Carspecken, 1999). Habermas has not only described a utopia or a pure form of communication; he has reconstructed that which makes possible any attempt to make sense of oneself and others. The principles of democracy and egalitarianism implicit in the ideal speech situation can unfold as empirical phenomena (for example, through voting), putting the ideal speech situation in the status of limit case, but the ideal speech situation itself is fundamentally a constitutive force presupposed in any communicative act, even nondemocratic, non-egalitarian ones. Therefore, the ideal speech situation (a) is carried through acts, not merely referenced by them; (b) works through intersubjective (position-taking) structures; and (c) cannot be limited to empirical claims, although it will always be connected to them.
Habermas (1987) developed these points by drawing on Mead's (1934) theories about the generalized other, intersubjectivity, and the self. The most basic critical potential associated with the ideal speech situation lies originally in the juxtaposition of communication's meaningful dependency on the ideal speech situation and its deferment of the ideal speech situation rather than on its secondary status as a limit case through which empirical acts might be measured. Therefore, the ideal of an ideal speech situation refers to presupposed principles through which communicative activity is actually possible but which is not likely to be fully empirically embodied in that same communicative activity. This dual character of Habermas's ideal speaks both to its generativity and to its pragmatic interpretability in everyday communication, and it is thereby also distinguished from both Schutz's (1966) eidetic and the empirical universalizations in the domain of the imaginary.
The imaginary, in this sense, is a more radically internal concept than those that we have explored thus far. It deepens our understanding of internal interpretability without stepping outside of it (Habermas, 1981). The term presupposed is used to indicate the leap to the imagined. In this case, the presupposed is that which can be reflectively put into discourse and used to describe empirical phenomena. In addition, there is another important difference here: The empirical world is conceptualized not solely as a world of existing things but as a world to which interactants can refer when communicating and achieving understanding.
Emergent ideal types: An introduction
Emergent ideal types, as I have conceived of them, are social attitudes regarding identities, roles, dispositions, and other self-related structures that are organized into a coherent system of normative expectations. They function as a limit case presupposed when social actors interpret themselves and others as particular types of people. The types emerge, in the first place, from the everyday interactions of social actors meaningfully engaged with one another. Their value for social science is linked to this. Emergent ideal types are ideal for self-other relations in the same double-edged way that the ideal speech situation is ideal for reaching understanding: They are generative and interpretable. Emergent ideal types are the imaginings through which one's interactions give one a sense of identity, self, and otherness. They are not ideal representations of a perfect image of something but, rather, of dialogically imagined possibilities [hence the generativity] for subject-specific interpretations given a shared set of normative relations (hence the interpretability). They are types because they have some substantive stability, that is, a boundedness that renders them recognizable as such-and-such. In other words, interpretations coalesce in way that can be referred to with some unity.
Emergent ideal types: A post-Meadian concept
Having established some sense of the social science context for thinking about the imaginary in relation to the real, in this next section of the paper I develop the concept emergent ideal types as part of the imaginary domain with leaps to the empirical: the discourse of the real. Emergent ideal types take us into a narrower discussion of the imaginary because we begin focusing specifically on the imaginary as it relates to ethnographic discourse about real subjects. The points will be exemplified using ethnographic data and findings from a previous study (Korth, 2001).
Mead's generalized other and emergent ideal types
Mead (1934) developed a dialogic theory of the self capable of explaining how individual selves and group norms are inherently mutually constitutive. Mead's concept of generalized other is crucial to this theory. According to Mead,
We get an “other” which is an organization of the attitudes of those involved in the same social group. The organized community or social group which gives to the individual his unity of self may be called “the generalized other.” The attitude of the generalized other is the attitude of the whole community. (p. 154)
Mead used the example of a game to illustrate this. A player must grasp what each possible play means from the perspective of all the players' positions; the organization of these positions is what he refers to as the generalized other. In other words, it is the unitary organization of these perspectives that constitutes the generalized other for particular social groups; it is not the mere accumulation of different positions toward one's actions but, rather, the qualitative unity through which all particular positions might grasp their relationships to one another for the entirety of the event. To keep with Mead's game example, players might similarly grasp the game as a whole and the possible plays as parts of that whole. This involves grasping the norms implicit in each of the possible plays as well as the system of norms that serves to unify the various possibilities.
The unitary aspect can be spotted in how one's actions would be evaluated. According to Carspecken (2002), this evaluation involves taking a third-person position on one's own actions by position taking with the generalized other. I might therefore “evaluate” my play as batter using this unified grasp of the system of possibilities and perspectives that comprise the generalized other of the “baseball game”: the set of expectations regarding what batters are supposed to do in relation to the rest of the players and the goals of the game. According to Mead's theory, all interpretations involve this kind of position taking with a relevant generalized other.
With respect to the interpretive process, Carspecken (2002) wrote,
We do not simply take the position of another person, we do so by simultaneously taking the position of a third person who could comment on the interactions of all participants and we assume that the other participants in the interaction can also occupy this position of the generalized other. (p. 17, italics in the original)
Therefore, the third-person position is one that could comment on the expectations for play regarding any one of the player's positions: one unified third person position through which all relevant positions are understood.
If the generalized other for a particular interaction is made up of the shared set of norms implicit in acting appropriately, then the attitudes that coalesce (let us say as “player positions”) assume the form of what I am calling emergent ideal types. These emergent ideal types are often taken for granted within a system of normative relations constituting the generalized other. To use more of Mead's (1934) formulations, we could say that emergent ideal types are nodes of social attitudes that are organized into a whole as the generalized other. They are the nodes of attitudes that one would recognize as relevant to understanding one's self from the generalized other. Here we can see both the generative and interpretive aspects. Emergent ideal types are generative, in that they are the attitudes that constitute the possibility of acting in particular sorts of ways and in that they open up the possibility for socially interpretive relations and recognition.
Returning to the game example, remember that the generalized other is the system of rules and norms that unite all possible plays from all possible positions in the game. The positions of pitcher, catcher, first baseman, and batter reflect coherent nodes of social attitudes and indicate possible emergent ideal types. Emergent ideal types are ideal imaginings in the sense that (a) they are presupposed in interactions both structurally (in terms of normative relations implicit to the generalized other) and substantively (in terms of the content boundaries), (b) they do not exist in time and space, and (c) they are both generative and interpretive. This proposal is post-Meadian because it extends the sketch that Mead laid out to include the dialectic of difference in the constitution of unity. So far, I have introduced the concept of emergent ideal types from within the domain of the imagined.
Theoretically speaking, a person constructs social identities (or “mes,” to use Mead's concept of the self) by position taking with the generalized other (Carspecken, 2002). Whatever me is constructed by position taking with the generalized other is done in part by identifying myself within the normative structure constitutive of the generalized other (Carpsecken, 2002). Emergent ideal types work off the principle of difference. Much of what is written about the generalized other captures the sense of singularity in terms of position taking with the generalized other, but few authors have tried to develop the dialectic with difference presupposed by any particular generalized other. Difference is both presupposed and produced the minute one position-takes with a generalized other. It is this position taking and the part-whole dialectic between the generalized other and emergent ideal types that seem to generate narratives that actors use to display the sense of unity of self that Mead considered an effect of the generalized other. This whole process (position taking with a generalized other, grasping the part/whole relations presupposed by the generalized other, comprehending specific activities through the generalized other, recognizing particular identities in reference to the generalized other, and producing narratives which display the unity of self) involves differentiating emergent ideal types.
It would also be useful to locate emergent ideal types in relation to social positioning theory, a contrasting postmodern theory of identity. Davies and Harré (1999) advanced Foucault's work to argue that people identify themselves by positing an identity that holds a relation to social discourse. When people interact meaningfully, they do so by establishing positions with those in the interaction, histories of meaning, and relevant social contexts. These positions co-create identities. “In making utterances, individuals seek to establish positions for themselves, both in an immediate relation with an ‘other,’ but also in relation to the many other exchanges that go to make up patterns of discursive exchange” (Winslade, 2005, p. 353). Moreover, according to Harré (1998), humans take up two identity projects through their activities. They seek to establish their honorability within a social group, and they also seek to establish their uniqueness as one very particular, individual being. These projects are motivational, and they require that the person be positioned socially. There is no singular positionality, as one's projections can be multiple. They are most certainly situational and ongoing. One does not accomplish the projects of being socially honored and individually appreciated once and for all. Neither does one accomplish these projects in the absence of social interaction. With Harré's (1998) conceptualization we see the unitary aspect of social identity involved in the project of being honored by one's social group and the differentiating aspect of identity involved in the project of seeking individual recognition as unique being.
Harré and van Langenhove (1991) argued that social positioning theory provides details necessary to explain how discourse is involved in the production of social relationships and the claiming of personal identities. “Indeed, it is possible to advance a notion of identity as a product of the clustering of repeated identity positions accepted and taken up in a multitude of conversations, rather than resorting to essentialist explanations of personality” (Winslade, 2005, p. 355). Dominant social discourses set the parameters for what is appropriate to talk about, what seems most important, what constitutes a successful argument, and so on. Social positioning theory offers an explanation analogous to Mead's (1934): one positions one's self (analogous to the ME in Mead's self) in relation to dominant discourses (analogous to Mead's generalized other) and thereby simultaneously positions others. As positions are co-produced, they create relational attributes, for example who among the interactants have the right to talk or respond in particular ways. The positions are real to the extent they are visable and viable through the discourse such that they are believable to others. The positions, in effect, become the manifestations of particular realities. The two identity projects (establishing one's self as honorable through the view of particular social groups and establishing one's self as individually unique) both require social positioning. The interpretive milieu is implicit in these projects and in the social positioning theory itself. Metaphysical descriptions of identity, with their links to humanism, are not credible for Harré (1997), yet the expressivism that emerges as an effect of these two identity projects is interesting to him.
Following this line of thinking, our identities are real because they are real for the actors, and discourses are real for the same reason, but reality is not a unitary thing existing distinct from the subject, and the imaginary does not stand in contrast to this reality. For Davies and Harré (1999), social positioning is immanent and is reproduced moment by moment in linguistic action, not from an abstraction. Abstractions are part of creating theories (Aronson, Harré, & Way 1995). Davies and Harré would likely interpret emergent ideal types as a form of discourse, real in their effects and not properly explained through as a presupposition in acting. According to Davies and Harré, the extent to which these ideal types manifest in action, they are real. They would use a real history of experiences and thoughts as well as manifest a contemporary, situated identity interdependent on the real forces of the conversations to explain similar phenomena. The project, from their perspective, would not be about capturing the imaginary in the real, because in the context of emergent ideal types there would only be the real. This problematizes my uses of the terms ideal and imaginary, and I will return this problem in the conclusion.
Capturing the imaginary in the empirical: An example
In this section of the paper, the empirical articulation of emergent ideal types is illustrated. There are two subsections and a conclusion, all of which are intended to illustrate how one captures and then displays this imaginary.
Methodology and context
I am using an ethnography of a U.S. kindergarten-Grade 1 (K-1) class to provide a concrete example of the philosophical points. (Korth, 2002). The class I studied had high level of student activity; a colorful and stimulating environment; class pets; student movement and talk; learning centers; computers; huge collections of puzzles, games, books, blocks, and toys; and the noticeable absence of desks (including a teacher's desk). In general, the classroom was a very rich, safe, orderly, well-planned environment within which children were able to pursue their learning interests. The social as well as material and academic milieu was intended by the teachers to facilitate group work and interaction.
This ethnography began when the school opened in January 1997. I observed the school's only class at the time (a K-1 class of 18 students with 2 teachers). I spent 8 hours a week from January through May observing the class. I became a regular part of the classroom life. Consequently, I was able to talk frequently and informally with the students as part of the everyday life of their classroom experiences. In addition, two other field researchers were making weekly observations. I conducted hour-long interviews with parents and several hour-long interviews with teachers and the principal. Classroom observations were audiotaped and videotaped. Interviews were audiotaped. All data were transcribed to create thick records. These were reviewed by the three of us involved in data collection. This produced hundreds of pages of data. These data were analyzed by me using peers and members as debriefers. These validity techniques were coupled with negative strip analysis and consistency checks (Carspecken, 1996).
To understand the classroom life, as ethnographer I tried to learn the “insider” perspectives, in the language of Mead (1934). This meant coming to understand the generalized others, which were most shared for participants of the classroom. As someone who was a relative outsider, this generalized other was always only partially shared by me, because although I learned to appreciate myself through position taking with these various generalized others, I was always also position taking through a generalized other that involved my ethnographic attitude more directly.
I spent a great deal of time trying to understand and reconstruct the activities in the classroom across time, persons, and interactions so that I could begin to internalize the generalized others that were presupposed by teachers and students. Doing this resulted in the location of tacit, commonly referenced, unified sets of norms. The norms presupposed certain ways of acting in the class that were referenced as a type of person. When I understood the interactions, it was partly because I understood the pragmatically structured ways in which class members recognized appropriate ways of acting and what that meant from various perspectives. Thus, there was a reciprocal and dialogic relationship between my intuiting the types and my reconstructive analysis of classroom activities. These types were not an accumulation of traits or a deduction of qualities. Rather, the word type refers to recognizing and participating in a particular bounded way of understanding self and others. The emergent ideal types reflect parts in dialectic with the generalized other as whole. These emergent ideal types can be differentiated as possible cultural limit cases in actual interactions.
I will discuss two of many emergent ideal types regarding classroom success: the Emergent Ideal Type: Nice Girl and the Emergent Ideal Type: Smart, Capable One[BOY]. To illustrate the methodological process, I describe some of the ways in which classroom actors pointed me toward an understanding of success in the culture so that I can describe the emergent nature of the analysis and pose a hermeneutic definition of success, through which I came to recognize two relevant emergent ideal types. Following that, I present the two specific emergent ideal types through a strategy that I call grounded narrative fictions. All of the names used in this paper are pseudonyms.
Internal references to success and emergent ideal types: Capturing the imaginary in the real
There were many references to success in the classroom; sometimes specific children were referenced as successful, and sometimes democratically established rules marked ideas of success, for example. Sometimes the references were constructed through language, sometimes through nonverbal acts, and sometimes through logical relations. By paying attention to these, I learned to recognize what it meant to be thought of as successful in this particular K-1 class. This is not the same thing as saying that I learned to recognize the general characteristics of success (although this follows), nor is it the same as saying that I captured the eidos of success. What I learned was the way in which the interpretation “success” was generated and bounded through a unified system of norms and a differentiated set of types.
One morning I was in the classroom sitting at a table with four children who were working on a puzzle together. A couple of boys, including Mark, were sitting on the classroom floor nearby, also collaboratively putting together the pieces of a puzzle. The children sitting with me were talking about their puzzle. Then one of them turned to me, pointed to one of the boys sitting on the classroom floor nearby, and said, “You see how many pieces he can work? That's because when he talks nobody understands him.” The children were pointing out to me a particular student, Mark, whom they took to be successful, and they were also articulating some of the ways in which they noted success: in this case, verbal vocabulary and the complexity of the puzzle he was able to solve (Korth, 2002). The student telling me about Mark and his friend was indicating how, through a generalized other, he might position-take in relation to Mark and boys like him. By talking to me about this, he was pointing toward a way of making sense of himself and others, a way that he expected was shared with those he was talking about and others in the class. The student was not merely providing me with a set of characteristics to list out. He was helping me interpret himself and others by sharing with me this particular way of “getting” Mark.
During center time one day, a group of children were shaping play dough. Allyssa pressed her shirt to the play dough and produced textures in interesting patterns. Another child approached the table and told the children that they were supposed to be making dinosaurs. Janie countered, “The other teacher said we didn't.” Then Janie went on to say, “We copy from whatever Alyssa does.” Kun adds, “Yeah Allyssa knows what to make.” Another child adds, “She always helps us,” then Janie concludes the interaction by saying, “We do what Allyssa does.” This example, like the earlier one, illustrates how children referred to a particular student while simultaneously foregrounding characteristics that they associated with success in their classroom environment.
Once during a math lesson, the class was playing Guess My Number. To play this game, the student leader sets boundaries within which the secret number resides, and fellow classmates attempt to efficiently figure out what the secret number is using the boundaries and others' guesses. It was Boyd's turn to lead the game because he had correctly guessed the teacher's number. One of the students said, “If ever Boyd is playing a game, I'm not playing.” Other students chimed in and said the same thing. Their actions within the group referred me to a particular child who might be thought of by students as less successful. At this point, though, I had little go on in terms of understanding why Boyd would be considered less successful. I could begin to understand who students thought of as leaders (and who they did not want as leaders), but I could not just equate leadership with success.
Students also referenced their peers through the use of eye contact, peer selection, and so on. For example, I observed that Jennifer frequently exchanged smiles with other children, often patted other children on the back, and regularly spoke positively to other children. Girls liked to play with Jennifer, and many students approached her for a variety of reasons. She interacted with others a lot.
In contrast, there were some children who were not given attention by their peers. Moreover, once in a while teachers would make reports to researchers like this, “I am concerned about Dave because he doesn't interact with the other children.” Here is a sad but typical interaction for Dave. Dave got up from his chair during center time and began walking. He walked (and this does not look purposeful) by the center where Mark was sitting.
John walks passed Dave.
Dave frowns. He leaves the play dough center and begins walking, somewhat aimlessly, around the room.
The teachers encouraged students to help each other in these kinds of ways, but Dave had a hard time securing such help and interacting with his peers. He spent much of his time in the class alone. These kinds of observations and comments provided an understanding of what and/or who would not be interpreted as successful by class actors.
Through all of these observations/interactions over time, I grasped a presupposed set of relations that were involved in how class actors evaluated and interpreted themselves with respect to success. These various attitudes regarding success were organized into a generalized other whose unitary nature had to do with locating one's self in terms of success. This success was largely visible through the popularity of the children among other children, but it was explained in broader terms: “That's because when he talks, no one can understand him.” Popularity was an internal attribute to success in this class. The empirical scene made sense through the cultural structures that were implicit in how classroom actors were interpreting each other. Further work using reconstructive horizon analysis facilitated the articulation of the notion of successfulness within the class. The analytic process and the data collection process took on a dialogic relation to one another, as did the various levels of inference involved in understanding how success was both achieved and pointed toward, including getting at the boundaries of success and its silenced oppositions. This was not a solely inductive process and involved the imaginary as interpretation rather than as the accumulation of a set of characteristics. In other words, the eidos of success– the sense already presupposed through the empirical characteristics but not equivalent to them — was captured. 1
Emergent ideal types were (re)constructed in the form of what I call grounded narrative fictions. These grounded narratives were the best display for core structures that presupposed the emergent ideal types. Grounded narrative fictions were a product of my social science, but they drew off of an impetus to narrate identity.
Grounded narrative fictions: Writing the imaginary into the discourse of the real
I played with several forms, including metaphors, in an effort to display emergent ideal types as they were presupposed through position taking with a shared generalized other. Grounded narrative fictions were the form that seemed most intuitively connected to emergent ideal types: They were capable of manifesting the interpretive fluidity and generativity associated with emergent ideal types and their generalized other. Grounded narrative fictions comprised actual data that were reconfigured (not decontextualized but relocated). These narratives facilitated the contextualized, implicit understanding of the subtleties involved in grasping emergent ideal types as presupposed by a generalized other. The narratives also brought out the interpretive boundaries of those types. On the risky side, grounded narrative fictions read as if the emergent ideal types were embodied by particular students rather than that they were the radically imaginary presuppositions organized through a unified generalized other.
Methodologically, (a) the emergent ideal types were thoroughly reconstructed, then (b) the data were reexamined in light of these reconstructions. This involved specifically looking for examples and counterexamples of interpretive, implicit references to the proposed types. Returning to the data provided both an opportunity for a validity check and an opportunity to begin (c) telling the stories that condensed the intuitive experience into a single narrative. The point of the stories was to set the context through which hermeneutic interpretations might emerge for readers, allowing them to grasp the generalized other and emergent ideal types presupposed through the hermeneutics of classroom life.
“I will help”: A nice girl
Alison (a fictional student who is constructed of actual data from across time, space, and person to capture the presuppositions of success which characterize the Emergent Ideal Type: Nice Girl) entered the classroom smiling. She looked around the room and seemed aware of various things going on. Regardless of whether it was center time or group time, she frequently acknowledged others in the class. For example, when she walked by children working in centers, she was likely to smile at them. When Alison joined a group of students already busily engaged in their activities, she did so by matching what they were doing: She smoothly integrated herself into the ongoing flow. Today, like many other days, she walked past a center, and one of the children said to her, “We saved you a spot over here, Alison. Remember we made patterns in the clay yesterday.” She helped others either when asked or by noticing that someone needed help and then initiating an offer. For example, while the students were writing in their journals, Alison noticed that one of her classmates was having trouble figuring out how to write a particular word. Alison gently tapped her friend on the shoulder and pointed to a spot in the environment where the problem word was written. Alison helped others clean up, make things, read, and so on. She seemed highly aware of the goings-on of her fellow classmates.
Alison was rarely in conflicts. She asserted herself but seldom ended up in a confrontational dispute. When she wanted a turn at something but had thus far been denied, she said, “I want a turn, please,” and usually she got one. Here is an example illustrating how she handled potential disputes without the events escalating into a conflict over access to markers. Alison was dealing with the problem at the same time that Susan, a less successful student, was trying to deal with the same problem.
There was a small, rectangular table with one long side pushed up against a wall. There were four chairs around the table, one on each end and two on the long open side of the table facing the wall. Margaret was sitting at one end of the table with the bucket of markers on the table directly in front of her. Cathy was sitting next to Margaret. Both of them were using the markers to work on their journals. Susan was seated in the chair next to Cathy. Alison joined the three of them by sitting in the chair at the end of the table opposite Margaret. Neither Susan nor Alison could reach the markers while sitting in their chairs.
This was what happened next.
Alison was not vocal in her resolution of this problem. Rather, she was interpersonally smooth, refraining from insisting that Margaret do something about the marker problem, and instead, silently and with a positive approach (smiling), Alison just got a handful of markers and returned to her seat to quietly complete her work. She left the group having stirred no negative feelings. She found a way to accomplish her goal while simultaneously maintaining positive interpersonal relationships with everyone at the table. Alison's success here did not facilitate the resolution of conflict for Susan.
Alison talked favorably about her work one-on-one. As she participated in the learning centers, she said things to her nearby classmates like this: “Look at this. See what I did.” She made similar positive comments about her classmates' work, “That's nice.” Alison was a good reader, and she was recognized as such by her classmates.
In summary, Alison tended to display happiness, personal assertiveness in one-to-one situations, good reading skills, engagement with classroom activities, support to fellow students, and positive conflict resolution. She was like a magnet to whom conversation, eye contact, and references keep returning.
The Emergent Ideal Type; Nice Girl was a non-empirical imaginary involved in how class members made sense of one another. References to the Nice Girl invoked an interpretation pattern that emphasized female gender and morality.
NICE GIRLS DO THINGS WHICH ARE NICE FOR OTHERS
NICE GIRLS HELP OTHERS
NICE GIRLS WORK HARD AND PRODUCE GOOD RESULTS
NICE GIRLS ACT FRIENDLY AND GET ALONG WITH OTHERS
NICE GIRLS ACT CONGENIALLY
NICE GIRLS DO NOT ACT COMBATIVELY OR ARGUMENTATIVELY
Therefore, when classmates were position taking with a shared generalized other, there was an Emergent Ideal Type: Nice Girl, which was one of the presupposed attitudes organized into the set of norms and values invoked through this generalized other. When classmates would position-take with this generalized other, they would recognize themselves and evaluate themselves by grasping the unity of the generalized other as well as in relation to the parts, like this Emergent Ideal Type: Nice Girl.
“Look at me. My idea is best”—The smart, capable one [BOY].
Craig, the name for the fictional character of this emergent ideal type, walked confidently into the room as if he had a purpose. He faced “the world” with his head up. His steps were quick, and he got right into an activity where there were other boys. He seemed focused on his activities, even when they last for only short periods. He was highly verbal and able to articulate theories, ideas, and opinions as well as engage in logical-verbal disputes with apparent ease. He talked more than other children in the room, and he used a noticeably more advanced vocabulary. He was considered “smart” by other children in the class, who used indicators such as the complexity of puzzles and the “way” he talked to distinguish him as “smart.”
Craig was frequently in disputes with fellow classmates. These disputes did not seem to produce any kind of unhappiness for Craig. In other words, disputes did not seem to be counted as any kind of lack of success on his part, as anything to avoid (whereas for girls, a dispute is interpreted as a lack of success, a kind of failure). He seemed, on the other hand, to feel proud of his disputes and proud of winning them, as often as he did. Usually these disputes were verbal but once in a while became physical. In addition to disputes, Craig was able to offer counterpronouncements that won the assent of many in the class without requiring much persuasive effort.
Craig acted with confidence and self-assurance. He made frequent bids for leadership during center and group activities. It was common for him to interject some interesting piece of information or some interesting theory or explanation into the ongoing group time discussions. For example, one day the teacher was reading a book to the class. She had just read a paragraph that contained the word glacier. Craig raised his hand, and this is what happened:
Right then, he went on to tell his classmates about glaciers. His frequent bids for leadership during group time resulted in his receiving a lot of attention from the teachers and peers. He spoke and answered and posed questions, spontaneously moving to the front of the circle to share his ideas with the group. He commanded a lot of attention throughout the course of the day and spoke very positively about his own work and ideas. He did this both through small groups and for the whole class. He was verbally assertive, but in a way that was much more public than Alison's mode of verbal assertiveness.
Craig interacted most often with other “successful” boys, although his interactions were not limited to this group. He seemed confident and happy as he undertook classroom activities. He frequently sought and received class attention and was considered smart by his peers. Unlike Alison, it was rare for Craig to offer help to his classmates or to compliment others on their work.
The emergent ideal type of successful boy was named The Smart, Capable One[BOY].
SMART BOYS EXPRESS GOOD IDEAS
SMART BOYS PERFORM WELL
SMART BOYS EXPRESS THEMSELVES WELL
SMART BOYS PERSUADE OTHERS
SMART BOYS ACT RATIONALLY AND REASONABLY
SMART BOYS EXPLORE THEIR CURIOUSITIES
SMART BOYS DO SMART THINGS
The reason I called this emergent ideal type Smart Capable One with the [BOY] as subscript makes sense because these characteristics of success were the ones articulated by the teachers and parents as what would be considered successful in the classroom in a presumably ungendered way, but in practice this form of success was really only actively interpreted as such for boys (Korth, 2002). I wrote the “boy” as subscript to indicate how gender was hidden in the discourse within the claim that this form of success was not gendered.
The imaginary in the discourse of the real: A summary
In the above descriptions, I wove together empirical examples to illustrate emergent ideal types in narrative form. These activities did actually occur but not as the actions of any one particular child. The point of using a narrative like this is to grasp the discursive nature of the emergent ideal type and to illustrate how such types would be recognized implicitly through roles and other pragmatic structures.
Now it is time to return to the original promise: demonstrating how emergent ideal types capture the imaginary in the empirical. Remember that the empirical has been referred to as claims that are dependent on space and time coordinates, which are produced in a form accessible by multiple folks. Empirical claims manifest one kind of interpretive position from which to act meaningfully. It is not that the empirical is non-interpretive; rather, empirical interpretations specifically report on activities that can be seen, heard, and so on by multiple others (at least in principle). All claims immediately depend on a set of intersubjectively shared structures (Brandom, 1994) and content, such as the concept of success, which might be taken for granted in the process of understanding. If it is thought of in this way, then the empirical seems to be a particular kind of differentiated interpretation from the more holistic imaginary.
The imaginary is part of the intersubjective web of possibilities that might be both impending and referenced. Capturing the imaginary involves putting intersubjective structures and content into discourse. Emergent ideal types are presupposed attitudes comprising a shared generalized other. The main point of the paper was to introduce the concept emergent ideal type, which was theorized as part of the process of capturing the imaginary in the empirical. This can be turned around by acknowledging the dialectic between these.
Conclusion
I began this paper with the idea that our ethnographic reports necessarily involve something more than the empirical. The point is fairly well accepted in the qualitative social sciences, but it is not well conceptualized itself. In this article, I have contributed to the qualitative methodological theory by defining the imaginary as the intersubjective structures that could be inferred from the empirical and on which the empirical was likewise dependent for its own saliency. This set up the argument that this imaginary is not something peculiar to the conduct of social science. Ethnographic participants, in making sense of their lives, will employ the same leaps of faith hermeneutically that ethnographers employ methodologically.
I introduced the imaginary concept emergent ideal type, and then I illustrated its relevance to ethnographic data analysis. The emergent ideal type is a substantive concept, in that it captures something that actors refer to in the process of coming to understand one another. The empirical and imaginary seem, in the end, to constitute a horizon relation from which actors reciprocally engage in coming to understand one another. This involves differentiating the empirical from the imaginary through time- and space-dependent claims. It is through this iteration in time and space that the imaginary is constituted, and yet it is also only through the generativity of the imaginary that the empirical can become differentiated. The underpinning explanation for this dialogic reciprocity is a theory of intersubjectivity (Carspecken, 2002).
Ethnography adds a layer of complexity because now the empirical is make up of the ethnographer's empirical claims about the lifeworld (actors' empirical claims plus the imaginary through which their empirical claims are sensible), plus the ethnographer's own imaginary. To put forward empirical claims about the lifeworld, one needs to intuit the lifeworld actors' imaginary. I labored over this as a methodological problem and decided that articulating the emergent ideal types that one captures when capturing the imaginary in the empirical is one way to do this.
Emergent ideal types might be put into discourse in several ways. I illustrated one of these: through the composition of grounded narrative fictions. I embedded the use of cultural reconstructions and the reconstruction of identity claims into the grounded narrative fictions because the narratives recreated an intuitive, naturalistic opportunity for readers to appreciate the hermeneutic grasp of the emergent ideal types. Such an intuitive display will implicitly involve the more basic paradigmatic structures of validity claims (Carspecken, 1996). Moreover, each of these ways of bringing the emergent ideal types into discourse mirrors ways these emergent ideal types might be brought into discourse in the lifeworld.
The concept of emergent ideal types was linked primarily to Mead's (1934) generalized other and secondarily to Harré and van Langenhove's (1991) idea of social positioning. Linking the imaginary (emergent ideal types) to the imaginary (generalized other, social positioning) requires philosophy, but it is relevant to empirical work. I have argued that our qualitative methods rely both on empirical observations and imaginary structures. Moreover, what we produce as research, this discourse of “the real,” is never purely observation. Qualitative methods will benefit from more exploration into the dialectic between the “imaginary” and the “real” in both our understanding of social phenomena and in our understanding of methodology.
At last, it is possible to problematize the imaginary as I have used it to describe emergent ideal types with Harré's (1998) pragmatic constructivist realism. Social positioning theory provided another plausible way of describing what I refer to as emergent ideal types. Contrasting my post-Meadian account with social positioning theory promises interesting insights. However, because this paper has methodological interests at heart, it is the challenge Harré poses related to the use of a category “the imaginary” that must be addressed. For Harré (1998), Davies and Harré (1999), Harré and Langenhove (1991), the discourse of gender identity would be real but indirectly accessed through metaphors, for example. This indirect access suggests for me the leap to which I have been referring. For Harré (Harré & Bhaskar, 2001) and others, indirect access is more of a procedural than a substantive feature of realities. The idea of multiple realities owing to the various situated aspects of identity and the various discourses prevailed on when securing one's identity are compatible with how emergent ideal types are proposed here, but the ideal and presuppositional description of those types is antithetical to the explanation that would be offered by Harré. Such divergent explanations in social science are worth dialogue, and it is my purpose in this paper to continue (not resolve) such conversations.
Footnotes
1.
These interpretations were checked out through interviews, peer debriefing, and negative strip analysis (Carspecken, 1996).