* Thanks to David Heise and Bernice Pescosolido for their review of, comments on and suggestions for this paper.

John O'Brien JOBRIEN@ucs.indiana.edu
Post Doctoral Fellow Program in the Measurement of Affect and Affective Processes INDIANA UNIVERSITY, Bloomington, Indiana USA 47405
Submitted: Mon, 6 Jun 94 18:09:24 EST

Abstract: Research suggests that certain connections between substance abuse, decision-making, emotion and identity/role change are best explained through construct realist theory. Six central theoretical assumptions are presented. First and second, links between knowledge and reality occur through abstract cultural mechanisms. Third, universal causal connecting structures are not identifiable through traditional definitions of knowledge and reality. Knowledge and reality are partly organized holographically; and, the heuristic of a constrained hologram accounts for regular differences in subcultural meanings, especially those that influence choices and behaviors. Fourth, some seemingly disparate grammar and symbol systems are not isolated configurations. Instead, they are bounded holographic systems. Fifth, social reality is constructed in limited domains. Sixth, there are meaningful correlations between the cross-culturally universal evaluation, potency and activity factors that can be derived from semantic differential measures. Those patterns embody distributed minimum units of cultural \ structure. A qualitative model based on these assumptions is presented. The approach produces a working synthesis between ethnomethodological, affect control, macro action and production system strategies. The model was developed by using ETHNO: computerized, logic driven, qualitative modeling software. The structure that was developed applies equally well to micro level substance abuse and macro level international military intervention scenarios. Since a single model that uses a constrained holographic heuristic accounts for both micro and macro event-sequences, it is concluded that a form of linkage occurs during choices and subsequent behaviors, and the link has properties of a cultural junction.

This work presents my formulation of construct realism as a synthesis of traditional and post-modern strategies. It provides what I believe is a novel micro macro link that occurs through the abstract media of cultural structure a and meaning. The connection is illustrated through a qualitative model that integrates ethnomethodological, affect control and production systems concepts. That theme is organized so that six core theoretical assumptions emerge from the presentation. A single qualitative model is applied to two diverse topics: the micro construction of behaviors and identities associated with substance abuse, and the structure of macro level socioeconomic and sociopolitical forecasts. I have not grouped the postulates under a single subheading. Instead, they are exhibited when logically called for. The first four can be found in the Introduction, and the fifth emerges in the following section titled A Developmental Model. The sixth assumption is submitted in the next segment, Macro Considerations It is central to the discussion of how and why the model portrays a micro macro link. In the conclusion my intent is threefold: to review the theory's broad implications, to show what specifics are inferred about preventing substance abuse and to suggest potentials for short and long term macro event forecasting.

Introduction

Post-modernism and traditional positivism represent those two perspectives, but neither offers what I feel is an adequate resolution for the problem of agency. Regardless, a solution can be developed by theoretically integrating positivist and post-modern concepts. That theory is construct realism. It is based on a unique set of assumptions about reality and culture, and about how those assumptions are realized through connotative meanings.

Pure post-modernism, interpretive approaches and radical constructionism tend to assume that there is no necessary connection between knowledge and objective reality (Von Glaserfeld, 1984). Positivism asserts an absolutely necessary connection. The counter proposal of construct realism is that there are definite, necessary and forceful relationships between knowledge and all but incomprehensible objective reality. Whatever that actuality is, it is not what positivists or post-modernists call social reality. Ultimately, however, it decides what is understood as real (O'Brien, 1992b). This assertion allows the use of quantum analogies to deal with the issue of culture and social determinacy versus indeterminacy. Is culture constructed through micro interactions, or is it an apriori determining force?

The second cardinal assumption of construct realism is that culture is a vital part of the connection between knowledge and reality. This suggestion was born in the statement that culture is "a set of control mechanisms - plans, recipes, rules, constructions, what computer engineers call programs for the governing of behavior" (Geertz, 1973). Geertz teaches that thick description and interpretations are professional techniques for understanding those control mechanisms. Descriptions of incidents and events represent interpretive understandings of what may or may not occur in reality. If the abstractions used in a particular culture are learned, unconscious biases will not be imposed during analysis of its day-to-day details.

Construct realism holds that everyone, in every culture, routinely makes thick descriptions and interpretations. Concrete experiences are regularly abstracted, and different individuals share the same abstractions. The resulting thick understandings control the details of day-to-day life, since we all unwittingly manage our own and others' experiences through those abstract understandings. Ergo, social situations map to interpretable culture; and, the cultural management of inherent tendencies enhances or inhibits social processes (Spiro, 1984: 328:330-335).

The relationship between culture and social structure is depicted in the institutional approach to cultural management. This is Fararo and Skvoretz' production system strategy, which evolved from Parsonian functionalism. The production system is seen as "the organization of social action in hierarchies forming institutionalized normative culture that provides the structure of social action and social systems" (1984: 178). Their method suggests two universal subdivisions of culture. There is content: the socio-cultural objects and events that are at once material, yet indeterminate and fabricated. Circumstance and process allude to that category. Second, there is the structure that organizes the objects and events into hierarchical orders and levels of meaning.

The suggestion that structured and integrated cultural systems exist implies that relative social reality can be accepted, yet theoretical accretion is still possible. Cumulative theory building in the social sciences usually denotes a positivist framework. Post-modern approaches substitute intertextuality and linguistic relativity for causal explanation. That substitution virtually denies the existence of actual universal connecting structures. At best, post-modernist speculate that it is impractical to identify and use such determining structures if they exist at all.

The third major assumption of construct realism parallels that post-modernist speculation. Actual connecting structures and cause cannot be identified through positivists' definitions of knowledge and reality. (SEE: NOTE 4) This is a presumption that aspects of knowledge and reality are organized holographically. A holographic system is a separate and identifiable thing: an object, an event, a circumstance, a process or a single order of organization. Every component in this kind of system contains some minimum unit of information out of which the entire system is constructed. For example genetic processes are similar to holographic processes, except that they are constrained by time and space. Beginning with the genes in a single fertilized egg, virtually every cell that subsequently develops contains the same genetic structure. These later cells also specialize as neurons, epidermal cells, liver cells and so forth; but, they still contain the same genetic codes. Normally, if a genetic change occurs in a cell then every time that cell or its offspring divides the new cells contain the same change. If the bio-technology were sufficiently developed, that principle would make it possible to clone an entire human being from only a single cell.

The correct term for a constrained holographic system is holonomic. This comes from the Greek holo, meaning complete, and nomo, meaning law (Bradley, 1987; Morgan, 1986; Heise and O'Brien, in press). Holographic systems must be constrained and bounded to array in different hierarchical levels; but, time and space are not the only possible constraints. The Heisenberg uncertainty principle allows for another set of dimensions, energy and momentum. Energy corresponds to the concept of intensity in the dimensions of time and space, and momentum can be conceptualized as activity or change.

Nevertheless, the organizational structure of all holonomic systems is the same. There is a minimum unit embedded in every component. That unit conveys the particular system within its constraining dimen- sions, and the components effectively interlock and process in parallel. Within a holonomic system there are no causal relationships between the system's unitary components. Everything in the system relates to everything else in the system. If the constraining dimensions are conceived of as energy and momentum, then the system theoretically spans more than one point in time. In that case, temporal relationships between those things identified as system components do not suggest cause. Instead, it is the relationships between orders of organization and different levels within those orders that imply structural regulation; and, structural regulation implies cause in energy-momentum systems, not temporal contingency. Therefore, when investigating a social situation we can look at relationships between different non holonomic systems, separate holonomic systems or holonomic systems versus non holonomic systems. Also, the interactions of elements within a particular non holonomic system can be analyzed, as can the relationships of components within a specific holonomic system. In each instance, however, there may well be different criteria for causation.

Why is a holonomic metaphor practical? There are many reasons. People are said to organize themselves holonomically, but that organization is said to be embedded within other social structures. Little is known about how these groups operate, or what influence they have on other social structures. The metaphor provides a theoretically coherent way to investigate that issue (Morgan, 1986: 95-105; Pribram, 1986). Moreover, the metaphor provides a joining mechanism between the individual and the group that other concepts do not. It links local interaction and global phenomena. For example ethnomethodology shows the existence of mutually set orders of social organization, including social structures and rules for negotiating social structures. Ethnomethodologists (sociologists concerned with the methods by which people create, maintain and heed implicit cultural rules) have illustrated emotional commitments to common understandings in social relationships. This has been done by having a group participant purposely violate some mundane, shared understanding (Garfinkel, 1967: 47-48). The reactions of all the parties involved in the situation illustrate a distributed set of rules for behavior embedded in and affecting people's affective responses during social situations.

A holonomic metaphor is a clear theoretical tool that allows understandable modeling of the relationships between identities, socially constructed emotions, implicit cultural rules and institutional social structures. The metaphor also provides a practical explanatory mechanism for the social psychological assertion that cognition and emotions are a single system (Franks, 1989: 97; Solomon, 1988: 83). It has equivalent explanatory power for other unifying assertions, such as the singleness of cultural grammar and symbol systems, e.g., denotation and connotation (Raynolds, 1970; 1988). Holonomics has the capacity to integrate the psychological and social structural aspects of culture into one system, and it abrogates the agent versus agency problem (O'Brien, 1992a). As a heuristic it provides anyone the opportunity for replicable qualitative modeling. Whether the analyst develops inferential or predictive capacities in the model depends on the approach (Bradley, 1987; O'Brien, in press). Since the scheme is proto theoretical, reductionist or purely rationalist understandings do not apply to it. The proto order is conceptualized as a dynamic shell within which macro action, affect control, production system, self, identity, role, value, attitude and other transformations are embedded. Such a uniting structure can be understood and modeled by using a holonomic metaphor.

For example, everyone generates thick understandings when they manipulate and organize their own abstractions to make events fit perceptions. (SEE: THE DISCUSSION OF THICK DESCRIPTION ON PAGE 2) That assures the reality of their own subjective beliefs in causality by producing the necessary and sufficient conditions for those beliefs, especially during the creation and maintenance of social truth. Positivists argue that truth is absolute; post-modernists deem it relative. The construct realist position is that the components of holonomic systems are relative. Truth is not, since there are universal regularities by which social truth is constructed (Bradley and Swartz, 1979: 2-3).

Ethnomethodology, macro action and production systems strategies illustrate some of those processes. Through breaching experiments, ethnomethodology demonstrates mutually embedded orders of social structure. Macro action refers to event chains that link micro and macro phenomena, where the connection is a causal action. The link is constrained in time and space, and the social system becomes an instrument with a purpose. It influences situations and social processes alike through contingency (Heise and Durig, 1992b). (SEE: NOTE 6) A similar idea is outlined in the concept of production systems, i.e., the production unit. However, where macro action begins at the level of micro interaction and builds upward, production systems begin at the macro level and decompose toward micro behaviors.

Both macro action and the production unit are valid fundamental units of analysis, but they are not technically minimum units of any holonomic system. Proposals for minimum units require tangible specifica- tions for what is universally distributed within a particular system. Such proposals also need to specify what processes allow for identical information to be transformed in such a way as to produce dissimilar system components (Slawski, 1986). Macro action and production systems approaches do neither. They only characterize different grammar and symbol systems as general unifying mechanisms.

The fourth central assumption of construct realism is that some grammar and symbol systems are unified by proto units, commonly embedded in each subsystem. These minimum units decide fundamental patterns for cultural routines, cultural rules, production units and macro action. That does not imply that ethnomethodology or the unitary suggestions of macro action and the production unit are fallacious. They are credible within their domains, but there are many units that could be devised for many domains. There is a more fundamental realm than those. It can be found in the concept of culture as an organized unifying system, managing other elements through proto structures.

A Developmental Model

Ten interconnecting levels of event chains can be included. All operate simultaneously and in parallel. That capacity is ideally suited to analyzing four aspects of an event-sequence. First, it can be shown how a distributed minimum unit of information joins the perceptually different components of a holonomic system. ETHNO can represent the effects of distributed evaluation, potency and activity relationships on the comp- lementary components within a multi level holonomic system. That shows the proto theoretical unity behind socially constructed emotions and situational decisions based on pattern recognition (Heise and O'Brien, in press). Second, it can model relationships between the holonomic and non holonomic systems that influence a social situation. Third, the software can identify reciprocally constitutive interactions between agency and social structure. Fourth, it can be used to infer abstract connections between separate holonomic systems, or within a single system.

A caution is appropriate. ETHNO was not designed to specify a minimum unit for a holonomic system. If one is used as a model chain component, then that unit must be predefined. Furthermore, the logic principles ETHNO uses to generate a framework highlight the temporality of event structures. The internal relationships of holonomic systems constrained in energy-momentum dimensions presume a-temporal and simultaneous processing. That disparity demands a developmental framework if a holonomic metaphor is used, not a causal one.

Consider substance abuse. It is usually dealt with through medical or legal models. Medical models are biased in favor of biology, psychophysiology and psychology. They imply that people have no choice about being ill, and that substance abuse behaviors are symptoms of an illness. The medical model tends to see the causes of abuse as genetic predispositions, progressive behavioral developments due to repetitive substance exposure and so forth. Those factors do affect the behaviors related to substance abuse. Nevertheless, they do not explain why some people with a genetic predisposition, or who are repeatedly exposed to certain substances, fail to become abusers. They also fail to reveal why some people without those proclivities do abuse substances. Overall, deterministic models do not provide realistic prognoses for long term abstinence, nor produce consistently accurate predictions about permanent cures. There are also legal models for substance abuse. These are based on the assumption that people freely choose what they do. To one degree or another legal models also consistently fail to predict who will abandon substance abuse.

The principles of construct realism allow for antecedent influences, as does the medical model. The same principles also allow for subjective cultural perceptions and preferences. These may be used to predict choices and behaviors. Thus, a model built on construct realist principles appraises both deterministic effects and indeterminate individual choices.

Consider the following summary of an actual case study. It took place in the urban North East, during the mid 1980's. After finishing a full-time day job, a thirty-eight year man got into his run down pickup truck and drove to a restaurant/lounge. It was a Saturday; and, he had not been drinking nor taking drugs all day. It was also Valentine's Day. He was moonlighting at a wedding reception. His night job was as a guitarist, and he was going play with his band at the wedding. When he had finished playing, he left almost immediately. He was then stopped by the police, arrested and taken to a State police barracks for a breathilizer test. It turned out to be borderline, one point zero one (1.01) approximately an hour after he left the reception. The musician only had one drink; yet, he was eventually convicted for driving under the influence.

There are many events that could have taken place in that type of situation. Between sets or after the reception he could have had something other than alcohol to drink. He also could have drunk beer, wine or liquor. He might have either had too much to drink, or stopped drinking before he exceeded the blood- alcohol content limit for driving. In most States, the public's common knowledge is that the limit is reached after two drinks. The musician might have had no problem in driving to or from the reception, but there was always the possibility of having an accident or being stopped by the police. If he was pulled over, they could have warned him or arrested him.

Modeling this holonomically requires several intertwined event-sequence chains, operating on four different levels. The first depicts concrete events, i.e., potentially observable things. (SEE: level 000 in Figure 1, page 8) The final element of the model chain is identity. If we want to know how identities and roles change, e.g., from a good person to a drunk, then we must know what takes place to invoke one identi- ty/role or another. By using ETHNO, events can be examined at the concrete level in relationship to theoreti- cally proposed abstract event-sequences that occur simultaneously. Inferences can then be made about what intertwined events logically take place to produce a given identity/role. We think of things as following each other from past to present, so this discussion is sequenced to conform to those perceptions. The character of events develops out of actors' reflexive post dictations (interpretations) of their own behaviors. Such interpretations are efforts to select event paths that will keep their lives meaningful.

The model begins when an agent leaves work, or any other place. The agent decides to drive to a bar or lounge, and pulls out onto the highway. When the agent arrives at the lounge, the car is parked, and he or she goes inside for a drink. We might believe that a choice to drink has already taken place. The medical model might argue that the subject has no choice but to drink. Those perspectives are not exactly correct. There is one further consideration.Not until that drink is in the agent's hand can any real choice be made that can alter event-sequences. Assume that the agent elects to have a drink, and then downs it. There are only two possibilities from here on, a non abuse chain and an abuse chain. Before the situation can evolve along either path, there is a critical moment of choice. That is usually not considered in medical models, but it is the core assumption of legal models.

In a construct realist model of substance abuse, event-sequences may change after free variation or choice points. Choices can be conscious or unconscious, and they always precede behavior. The socially desirable decision is represented as a non abuse event-sequence chain. This parallels behaviors in case studies, where an actor has no more than one drink. There is also an abuse event-sequence chain. In this chain, the actor may have more than one drink, and go to the rest room and telephone a friend. Arguments are also possible, and the actor can order still another drink afterwards. The model's order-choice-drink-order sequence repeats until the actor decides to stop drinking and leave. In both the abuse and non-abuse chains, electing to drive reinitializes a common choice-driving-choice-stopping series.

The construct realist model takes into consideration that until other people in the context of a situation come to regard actors' behaviors as socially unacceptable, they have unrestricted agency. As constructionist argue, the cultural rules of social interaction can be freely negotiated. If the police catch an actor driving after drinking, that seriously limits that actor's agency. As determinists argue, by the acts of representatives of a formal institution, agency is given to social structure. In the construct realist model, agency transformation is signified when an actor is stopped by the police. Since the incident constrains the actor's freedom, situationally applicable rules and their outcomes cannot be freely negotiated. This would be so, even if an actor disagreed with the police about taking a drunk test.

Figure 1 - Common Micro Macro Structure for an Ethno Generated Qualitative Event Sequence Model

Moreover, abusers are labeled as either medical or criminal problems. If the law is not involved, society disfigures their identities by deciding that they are physically or mentally ill and treats them accordingly (Hewitt, 1976: 190-226). The legal system labels them criminals, if their actions are formally noted. If not, abusing alcohol or drugs can result in a label of mental illness by significant others. Either option limits social freedom, responsibility and agency; and, labeling not only denotes identity/role change it also means alterations in status and power.

Interpretive Considerations

What distinguishes a construct realist model is the proto theoretical approach. It joins all those factors into a single system. This effectively augments ethnomethodological considerations, as well as affect control and production system strategies. How can those factors be combined? Theory suggests that a TEST- OPERATE-TEST control logic can be used to deal with subjective data (Miller, Galanter and Pribram, 1960). Cultural minimum units are quantifiable as affective values derived from semantic differential measures. The units function to integrate the environment/context of behavior, the actor as a purely qualitative symbolic information processor, macro level production plans and micro level affect control interaction modeling. Affect control and production systems speculations use similar control logic. Both propose common starting elements: situation spaces and object-types.

In the model's abuse event-sequence, situation space refers to the context/environment in which an actor is stopped by the police. Everyone involved enters this space with their own fundamental sentiments: relevant evaluation, potency and activity profiles. Those define each person's contextual self image, possible relevant behaviors and the objects and others involved in the situation. Object-types are the law (a cultural object) and things such as people, place, time, etc. (physical objects). Object types also include the roles and identities that each person or thing is endowed with, e.g., , , , the social status of the actor , the social meaning of the time and date and other social objects. Another object type is the objective or goal that everyone in an event has.

In the case study, the police could have been required to meet a quota or to follow special procedures on Fridays, Saturdays and Sundays. They might have believed that they only wanted to prevent an accident. That would have reaffirmed their identities as good people. The musician may have just wanted to get home. He may also have hoped to avoid arrest, or to have reaffirmed his self image. (SEE: THE SUMMARY OF AN ACTUAL CASE STUDY, page 6)

In the model, fundamental sentiments and context specific transient values (i.e., affective evaluation, potency and activity values) are portrayed as holonomically distributed cluster values. These are derived from relevant correlations between evaluation, potency and activity factors. (SEE: Figure 1, page 8) Cluster values are part of the cognitive maps that we all carry into any situation. They are components of a preset plan for social action, a set of production rules inherent to the proto unit. The suggestion that cultural minimum units derive from correlations between evaluation, potency and activity factors embodies corresponding hypotheses about unspecified control transfers and relevant institutional production rules. These stem from object-types and tests of realized situation-types versus proposed situation-types. The units also embody mechanisms by which priorities influence outcomes.

The fifth essential assumption of construct realism is that social reality is constructed, but only in limited domains. Recall that the fourth postulate states that the minimum units contained in holonomic cultural systems are junctions between rules, institutions and actions. The systems they embed in organize as hierarchies. Those hierarchies then form distinct domains of rule systems, institutionally normative cultures and the micro processes involved in day-to-day interaction. The effect of the relationships between fundamental sentiments and institutionalized expressive actions dynamically characterizes that link.

Consider the construct realist model in relationship to the proposal that evaluation maps to status, and potency maps to power. That directly ties evaluation, potency and activity profiles to individual stress reduction. The actor and the officers would all be predisposed to reaffirm their identities. More, everyone would share the production rules that must be followed in the situation.

In the actual case study, the subject reported that there was no on the spot breathilizer test. He was given a discretionary examination that the police were free to interpret subjectively. He stated that he was told that his truck was stopped because one tail light was burned out, but he believed that the police were looking for people because the bars had just closed. He said he had not been drinking at the reception, but did have one shot of whiskey just before he left. The best man tipped each member of the band fifty dollars, and bought them all a drink. Since it was less than fifteen minutes later that the police stopped him, it does not appear that the musician drank enough alcohol to be drunk, legally or otherwise.

To examine the cultural management of that situation from a production system viewpoint, consider the production unit as the actor and the context. The latent production state is defined by the cultural rules, perhaps that acceptable standards of personal appearance and transportation are required, and that citizens defer to representatives of collective authority (Fararo and Skvoretz, 1984: 145).

The action process occurred when the musician truthfully answered questions about his drinking, and then showed irritation when asked to step out of the car and be tested. We can assume that the irritation arose in relationship to his need to confirm his identity. Moreover, the subject has stated that he was aware of situationally appropriate cultural appearance and transportation rules, but has not abided by them as part of his life style. In that context the action process is a TEST-EXECUTE structure: a test of the capacity of all parties to abide by production rules governing normative American culture.

The dispute in the case study became a symbolic space, controlling the acquisition of everyone's status. Ethnomethodologically speaking, the musician failed to abide by deference rules when he breached the normative appearance and transportation rule systems that interpenetrate members of society. That failure in abstract symbolic space (TEST) was followed by an affective response on everyone's part. Everyone's subsequent behaviors were intended to reduce their own stress levels, and reaffirm the validity of the unspoken cultural rules. The police used their formal power to make his appearance, behaviors and irritation meaningful. They insisted on a concrete examination for alcohol impairment at the local state police barracks (EXECUTE). When the musician displayed irritation, it was an attempt to reaffirm his status, role and identity. The unavoidable perception that others held distinctly different opinions about what his appearance and behavior meant (i.e., his identity and status) was enough to interrupt his identity confirma- tion process and cause him stress. He was inexorably being changed from good man to criminal suspect. Because he had chosen not to abide by normative appearance and transportation rules, anything that he did or said was liable to be interpreted as alcohol impairment in the subjective cognitive maps of the police. Even normal characteristics like being bone-tired at 2:30 a.m., or trying to walk a straight line in a tuxedo and dress shoes on Valentine's day, was probably understood as denoting the excessive use of alcohol.

After he was arrested, those subjectively interpreted meanings of his identity and behaviors became part of the public domain. At State police headquarters the meanings were reinforced by what was subjectively interpreted as incontrovertible hard evidence, the results of a breathilizer test. Eventually, the musician was tried and convicted under a plea of no contest.

All of the parties in this situation collectively fabricated the drunken-driving and alcohol abuse event. The police officers and judge interpreted the meaning of his appearance, behaviors and the breathilizer test results as criminal alcohol abuse. The musician did not maintain a subjectively proper image, or sufficiently abide by deference rules. When he was asked why he did not contest the charges he responded that it was far less expensive to pay a fine for a first offense, and to pay for and go to a jail-treatment center at the Holiday Inn for two days, than it would have been to hire an attorney.

This fabrication took place in accord with universal regularities that equally affected all of the systems, events and people involved. Some of those universal regularities can only be explained by construct realist assumptions embedded in the model. There are the holonomic relationships behind choices that link concrete and abstract events. (SEE: Figure 1, page 8) Other regularities can be explained through assumptions about causal affects in other types of systems, for example the effect of genetic predispositions. These are also included in the model as limitations and enhancements, e.g., the predispositions and affects that come from an actor's physical condition and structural predisposition. These are portrayed where concrete events instantiate abstractions at level 010. Personal appearance and biological processes are subsumed under physical condition. Some biological and neurological processes are not holonomic systems, but they directly and indirectly influence holonomically distributed information (Pribram, 1986). In the model, situation specific transient evaluation, potency and activity profiles directly influence subsequent choices and behaviors. Meanings are derived from the profiles (Smith-Lovin and Heise, 1988).

There is more than one abstract stage in the model. Level 020 represents the effect of alcohol and other substances on bodily and neurological processes, along with the influence of evaluation, potency and activity values. Physiological changes occur in conjunction with emotions, but we know that many emotions and emotional display rules are socially constructed. Because evaluation, potency and activity values can be demarcated from those measures, physical conditions instantiate the effect of substances. We may assume that this also influences transient values.

Two abstract tiers are not enough. There is a real possibility that some people are genetically predis- posed to substance abuse. The effects of substances instantiate a greater abstraction at level 030, where genetic makeup is a factor in structural predisposition. In the model, genetic makeup signifies a separate holonomic system constrained in time and space.

The model also depicts control mechanisms. Embedded in every concrete event are transient evaluation, potency and activity value profiles. These are immanently compared to corresponding preexisting standards. The comparisons result in deflections. Behavior choices that immediately follow transient impressions function to minimize deflections. The arguments are that deflection is stress, and that socially constructed emotions are derived from level 010 evaluation, potency and activity profiles (Smith-Lovin and Heise, 1988).

I propose that there are other tiers of event chains that operate in parallel with those subordinate levels. (SEE: Figure 1, page 8) We generally think of these as culture, and they are represented in level 030 of the model. I believe that there are many such cultural tiers, although in the model they are condensed into one level. The reason for that is quite pragmatic. Evaluation, potency and activity factors are universal components of meaning, but the correlations between them are inconsistent across and within cultures (Osgood, May, and Miron, 1975; Schneider, 1991; Schneider and Heise, 1993). That implies that denotative evaluation, potency and activity categories form taxonomic structures whose specifics differ cross-culturally. Therefore, there are a multitude of such intertwined structures at different levels. Signifying them in condensed form illustrates their effect, without undue complication.

For example, the Indo-European spectrum denotes eight root color categories. The human eye is said to perceive a virtually unbroken progression of wave lengths from the ultraviolet to the infrared, but categorizing physical perceptions is a cultural process (Berlin and Kay, 1969). What that suggests is that similar evaluation, potency and activity profiles should distribute around definite cluster means. Also, consider complexes of denotative meanings relating to concepts of authority. These differ from one culture to another in regular patterns. It has been shown that their evaluation, potency and activity profiles do distribute around similar cluster means in different cultures (Schneider, 1991; 1992a; 1992b; Schneider and Heise, 1993).

The model represents those clustering processes at level 020, where physical or material condition affects cluster standards. Logically, alcohol or other substances would modify the results of comparisons between transient profile values (010) and cultural cluster values (020). (SEE: Figure 1, page 8) This is due to the relationship between behavior and transient value profiles. In the case of substance abuse, it amounts to the physical changes that come from using alcohol or drugs.

Tier 030 in the model contains an event (i.e., the cultural analog) that condenses a series of possible higher order abstract or physiological event-sequence chains. The analog signifies information about the evaluation, potency and activity correlations within the cluster standards shown on level 020. This is a cluster of clusters, a categorical standard of standards, a super cluster that can be influenced by a separate holonomic system: genotype. In level 030 it is termed structural predisposition. Super clusters affect evaluation, potency and activity cluster mean standards at level 020; therefore, they also indirectly influence identity/role changes. What accounts for that is the structure of the minimum unit embedded in the cluster mean, the identity/role, the socially constructed emotions, behavior choices and so forth. The concept is consistent with recent assertions in identity theory. The model shows how the self can be entwined in social action, and how it can develop as a dynamic self organizing system (Burke, 1991: 846; Burke and Reitzes, 1991; Schwalbe, 1992: 270:272-282:287:288:290-291, re: notes 1 and 2).

Macro Considerations

Phrased differently, do postulates of holonomic organization in social systems inherently restrict unitary processes to the individual; or, is some form of pan-individual operation implied? That question is beyond the scope of this work, and it is intentionally bypassed. Even so, the holonomic metaphor does stimulate the thought that shared similar experiences are a reasonable bridge between the individual and social structure (Heise, 1990: 59-60). The real issue is what that sharing implies, unity or similarity?

The only assertion in this work is that corresponding structures and processes exist across subjective and objective realms. These are described in both the physical and social sciences (Gleick, 1987; Schwalbe, 1992; Prigogine and Stengers, 1984). Careers have been defined as cultural structures, unfolding in harmony with institutional rule systems (Heise, 1990). Rules of assembly for the discrete masses described by set theory correspond to logic rules (Heise and Durig, 1992a). Thermodynamic processes map to the affective evaluation, potency and activity structure of feeling processes. Speculations that the mind arrays in different psychological schema map to considerations of the social group (Gazzaniga, 1985; 1988). Still, cross-level system structures, events and rules of process map imperfectly.

Information contained in a minimum unit is theoretically identical within and across different compo- nents of its particular holonomic system. So, the structure of the unit must be shared by all the components in the system. Since the structure is common, allowable difference within and between the relationships of the elements in the unit must serve as differentiating attributes of the components as much if not more than differential contents.

The sociology and anthropology of knowledge document differentiated distributions of knowledge in any given population, as well as a person's different total knowledge at different points in time. Since a minimum unit has some characteristics of a form of knowledge, what does a minimum unit consist of? It theoretically contains all of the information necessary to construct a total holonomic system, so how does the information contained in a minimum unit differ from an individual or collective definition of total knowledge?

The best answers condense into a single negative declaration. Knowledge is not defined in a positivist or post-modern framework. Nevertheless, those definitions are a part of the construct realist understanding of knowledge. As it is referred to in the assertion of a necessary connection between knowledge and an all but incomprehensible reality, information has two elements: information content and information structure. The first parallels both positivist and post-modern ideas. This is knowledge that relates to the questions of whom, what, where, when, which, how and why. Since that genre of information is temporal and spatial, it varies from individual to individual and from time to time. The second element is another order of information altogether. It is knowledge about the organized taxonomic production of the questions them- selves, and it may relate to energy and change dimensions.

Minimum units exhibit the same type of bifurcation. They have structure. That structure then organizes variable contents in different components. This effectively states that concrete social phenomena located in time and space can be organized by structures in the dimensions of energy and change. If a system's components are related through complementarity, the common structure may be so transformed from one component to another that it appears different. Common structure can organize various degrees of similar knowledge among different people, if they all share sufficiently similar experiences. This is the case with a cultural or social group. Two people may technically speak the same language, i.e., they share a common grammar and syntax, but their vocabularies may be quite different. Consider the variations between day-to-day English and formal legal English, or American English and British English.

A social group's total knowledge consists of two parts. First is the aggregate of individually differentiated knowledge contents. The second factor is shared and partially negotiated structure for how to know and arrange those contents. The same can be said for individuals; common structure organizes different contents at different times. A fifteen year old high school student is not expected to have the same technical vocabulary as that same student would have ten years later, after obtaining a professional degree in the social sciences. Over time, an individual's total knowledge consists of both differential contents of what is known and a common structure of how to know it. When time is not a consideration, the same structure can organize different bodies of information content. For example, some forms of cognition and emotion are different manifestations of the same thing, e.g., the word joy versus the experience of joy (Franks, 1989: 97; Solomon, 1988: 83). The symbol is not the thing, and the thing is not the symbol. At another level they are.

The phenomenological assumptions underlying affect control theory explanations of the definition of a situation illustrate that point (Smith-Lovin and Heise, 1988). Without elaboration, the basic suggestion is that everyone uses two precise versions of three separate elements during any interaction. The elements are self, possible behaviors and objects/others. Each is fashioned from three subfactors, and every actor's goal is to minimize their personal stress in every situation. The goal and elements function to create meaning in a situation.

One version of the three elements is a standard. Standards are formulated from evaluation, potency and activity factors. Self, behavior and object/other each have their own standard. Those standards are initially carried into every interaction by an actor, and their profile values denote the actor's situational self image, situationally appropriate behavior and expectations about the object/other in a given context. Version two is a transient. Transients are built up from the same subfactors, i.e., values for evaluation, potency and activity. Like the standards, there are distinct transients for context specific images of self, behavior and ob- ject/other. Transients emerge from an event. They are not carried into it. Through an immanent control system, the actor compares each transient to its equivalent standard to discern any difference. Differences are termed deflections. Summing all relevant deflections for self, behavior and object/other produces a measure of stress. The assumption is that a person always chooses the behavior that will minimize their stress in the situation. If stress cannot adequately be reduced through a choice for behavior, then it may be minimized through an alteration of the behavior standard, object/other standard or self standard. Action takes place only after behavior is chosen, since the theory is developed around the belief that choice alone minimizes deflections. A concrete action establishes a new event, and the process begins again for everyone involved.

Affect control theory shows that there are distinct kinds of knowledge. There is knowledge content, where denoted meaning derives from relevant evaluation, potency and activity values. This type of meaning varies as those profiles vary. Affect control theory also suggests a second sort of knowledge necessary for defining a situation, a form of non conscious knowledge. This is information structure. It is assumed to be based on the cross-culturally universal existence of evaluation, potency and activity factors that are organized in specific ways to produce consistent meaning classes. The proto theoretical construction of a minimum unit refers to that second form of information.

That brings us to the sixth assumption of construct realism. Quintessentially the form and duration of correlations between universal evaluation, potency and activity factors are distributed minimum units of cultural information. (SEE: THE DISCUSSION OF UNIVERSAL TAXONOMIC STRUCTURES, COLOR CATEGORIES, AND AUTHORITY CLUSTERS, page 12, 13) Take the taxonomic classification red; it has many modified subclassifications such as light red, dark red, bright red, fire engine red, blood red, pink, mauve, et al. Each subclassification is culturally constructed and denotes knowledge, but positivists would argue that each represents a physical perception of a wave form of real light. As in the linguistic distinction between phonemics and phonetics, cultural organization may remain the same over time even when the specific content of perceived information varies. Moreover, color and emotional synesthesia are phenomena frequently expressed in day-to-day discourse (Osgood, 1975). To say that one is sees red is the American way of implying anger. Affect control theory shows that denoted meaning for the color red may be decomposed into a set of relationships between evaluation, potency and activity values. None of the subclassifications of red have precisely the same profile, but all of the correlations among the three factors for all of the color's subclassifications are similar and clustered.

Affect control principles explain how some aspects of the meaning of an event or situation is constructed, but meaning is not restricted to the specific profile values. Those denote meaning. There is another broad type of meaning, connotation. Connotation maps to the structure of the correlations between evaluation, potency and activity. It is somewhat divorced from particular denotative values. For example, there is a connotation that underlies each denoted meaning in the cluster of American English words within the category red. This can be interpreted as a culturally organized connotative cluster of particular denotations, related to a spectrum of particular perceptions.

There are levels of connotation. There is the cultural cluster mean that designates a specific culturally constructed color category. Theoretically, there should also be super clusters of similar cultural cluster means that group somewhat akin cultural categories into taxonomic structures. Thus, moving down the taxonomic structure, the category red has many subcategories. Moving up the structure, red is part of a broad cultural construct meaning color.

Another form of connotation has to do with the phenomena of color or identity synesthesia. This is metaphor, if you will. It is hypothetically possible that the meanings of specific denotative profiles have symbolic connotations because of similarities between their evaluation, potency and activity factor correlations and the correlations of the cluster means of a different cultural category. That would explain why some socially constructed emotions and identities cross associate to color, and why those cross-associations differ from culture to culture and subculture to subculture, e.g., being a red, feeling blue, seeing red, being black, being yellow, green with envy and so forth.

This view of connotation is supported by research using the visually abstract non verbal projective differential scale. A strong and consistent cross-cultural Japanese to American correlation is described for specific evaluation and potency values obtained from the projective differential (Raynolds, Raynolds, and Sakamoto, 1988). In American society, specifically selected verbal terms such as left versus right, up versus down, good versus bad, strong versus weak, etc. show strong correlations between evaluation, potency and activity values obtained from the projective differential and equivalent values obtained from the semantic differential (Raynolds, 1970). Therefore, given a specific environment/context and a situation to be defined, the full knowledge of meanings represented within a holonomic system can be extracted from the structure of its minimum unit as it organizes specific contents.

Macro Scenarios

The worst case scenario contains every possible model incident. It depicts the use of nuclear weapons, incursions into Ethiopia and Yugoslavia, United Nations economic and military sanctions against the United States and the decline of the United States. Scenario two is moderate, but excerpted from the full event- sequence in scenario one. The only deviation is in those incidents that occur after points of choice. Episodes in scenario two include intervention in Somalia, unilateral expansion of American military force into Yugoslavia and condemnation of the United States by the United Nations. The final event-sequence is the best case scenario. Again, the event chain is excerpted from scenario one, and differences are developed around choices. The United Nations and the United States intervene in Somalia and rapidly succeed. The United States then withdraws, as public opinion turns against the operation. Finally, the United Nations resumes unassisted aid deliveries to Somalia.

No alterations in the basic micro event-sequences, abstract level relationship structures or instantiation relationships are necessary for any of the macro scenarios. Just as in the micro level substance abuse scenario, situationally relevant transients compare to preestablished standards for actor, behavior and object/other. This is a first order control loop between the concrete and abstract in which processing happens in parallel. A second order control loop matches the same transients with equivalent preexisting cluster standards for actor, behavior and other/object. That is represented in level 020. Simultaneously the transients incorporate into an analog standard of standards for acceptable cultural clusters. This is a third order control process at level 030, involving another comparison with an unspecified analog standard. If deflections are not resolved at level 010, the fundamentals for that level are adjusted in layer 020, where cultural categories cluster. Should that fail, another control mechanism is activated in level 030. There is also a level 020 comparison of cluster fundamentals and level 030 analog standards. This process can adjust the level 020 cluster fundamentals, and ultimately those at level 010. (SEE: Figure 1, page 8)

Conclusion

Metaphorically, that means that the proposal of a system as more than the sum of its parts ceases to be a cliche. It also implies that non conscious pattern recognition precedes rational decision-making (Crumley, 1985; O'Brien, 1992a). The metaphor shows that such a system can be theoretically coherent according to Leibniz' principle of the identity of indiscernibles (Bradley and Swartz, 1979: 39; Heise, 1975: 27).

The construct realist model does not use elaborated specific processes, where they are represented at each level. The exact contents of minimum units are not used, nor are any detailed operations of the transformation mechanisms that apply to them. The model is not designed to calculate interactions, as do specific affect control or production system simulations. It only marks the relationship of diverse orders of events as they are hierarchically arrayed in different levels.

Theoretically, the preidentified processes influencing decision-making are subject to a rigorous logic governing the fabrication of situations. That logic and the processes are congruent with sociological speculations about the institution/individual. They are also compatible with assumptions about the function of distributed minimum units at individual and group levels (Giddens, 1987). Therefore, subjecting the metaphor to trial by cross-level modeling does not contradict the suggestion that common cultural structures provide a form of micro-macro union. The model is cross-institutional and cross-cultural, since the same event-sequence structure can be used to scrutinize micro or macro incidents from the perspective of a particular culture.

For instance, structural predisposition in both micro and macro scenarios denotes a distribution of standard and transient evaluation, potency and activity profiles. It also denotes aggregated deflections from control comparisons at different levels. In a model of a democratic society, that symbolizes aggregated or collective public opinion. It is derived from the relatively free interactions of individual agents. Shift agency to the institutions of a totalitarian power, and the final outcome of event-sequences could change. However, the model's structure remains the same.

If a simulation were developed from this qualitative model, preidentification of evaluation, potency and activity cluster and analog correlations could occur in two ways. First, a culture specific data base could be used for analyses. This restricts the simulation to a particular culture, since the relevant evaluation, potency and activity correlations embed in the data base. Multiple data bases would be required to simulate cross-cultural or cross-subcultural interactions. Second, a holonomic metaphor implies that specific correlations for relevant cultural analogs and clusters can be used to adjust a single data base. Figure 2 illustrates two points critical for this approach: (1) how hierarchies of holonomic processes deal with evaluation, potency and activity profiles; and, (2) how a distributed proto theoretical minimum unit coming from the correlation of evaluation, potency and activity factors unifies different components. (SEE: Figure 2, page 21)

There are several implications. First, Raynold's findings suggest that the minimum units acting as proto theoretical containment shells may consist of visuo-spatial or proxemic evaluation, potency and activity factor profiles. Those relate to other systems by way of the correlation patterns of the factors themselves. Second, the model is a practical supplement for the medical and legal approach to substance abuse, one that considers both determining factors and abstracted processes for individual choice. Third the concept of structured proto units suggests that abusive behavior may be predictable, for an individual or for classes of individuals. That indicates that a simple and supplemental treatment program for abusers can be developed. The program would focus on identifying the proto theoretical relationships of the evaluation, potency and activity values of abusers. Retraining and resocialization techniques could then be implemented to transform the proto structure, by reestablishing values that more approximate non abuse patterns. If successful, such a program would have immense social impact.

Fourth, when used for macro analyses the model suggests a means to augment traditional political- economic forecasting with considerations that allow for routine cultural differences. In essence, this standardizes a verifiable interpretive approach. Beyond the theoretical ramifications of that statement, the practical results might include the development of predictive procedures for the kinds of social disorganization and reorganization recently seen in Somalia, and in Eastern and Western Europe. They might also allow for cultural systems to decison-making interface research. These could span diverse areas, including ecologically oriented choices, economic behaviors, ethnic interactions, cross-cultural interactions and so forth. The final results would contribute information vital in policy making.

Figure 2 - A Multi Layer Cultural Control System *

References Cited

Alexander, Jeffrey C., Giesen, Bernhard, Munch, Richard and Smelser, Neil J.
 (1987)  The Micro-Macro Link.  Los Angeles: University of California Press.
 
Archer, Margaret S.  (1985)  "The Myth of Cultural Integration."  The British
 Journal of Sociology XXXVI(3): 333-353.
 
Adler, Franz  (1960)  "A Unit Concept for Sociology."  American Journal for
 Sociology LXV (January):356-64.
 
Berlin, Brent and Paul Kay  (1969).  Basic Color Terms: Their Universality and
 Evolution.  Berkeley: University Of California Press.
 
Bradley, Raymond T.  (1987)  Charisma and Social Structure.  New York: Paragon.
 
----------  (1991)  "Dynamics of Collective Organization: A New General Theory
 of Affect, Power, and Organizational Function."  Paper presented to the Social
 Theory Roundtable of the 1991 American Sociological Association annual 
 meetings.  Cincinnati, OH (August).
 
Bradley, R. and N. Swartz  (1979)  "Relations Between Propositions" in Possible
 Worlds: An Introduction to Logic and Its Philosophy.  Indianapolis: Hackett.
 
Burke, Peter J.  (1991)  "Identity Processes and Social Stress."  American
 Sociological Review, 56(December: 836-849).
 
Burke, Peter J. and Donald C. Reitzes  (1991)  "An Identity Theory Approach to
 Commitment."  Social Psychology Quarterly, 54(3):239-251.
 
Crumley, Carole L.  (1985)  "Pattern Recognition in Social Science."  Social
 Science Newsletter 70(3): 176-179.
 
D'Andrade, Roy  (1981)  "The Cultural Part of Cognition."  Cognitive Science
 5:179-195.
 
----------  (1984)  "Cultural Meaning Systems."  Pp. 88-119 in Richard A.
 Shweder and Robert A. Levine (eds.) Culture  Theory: Essays on Mind Self, 
 and Emotion.  Cambridge: Cambridge University Press.
 
Durkheim, Emile  (1974)  Sociology and Philosophy.  The Free Press: New York.
 
Fararo, T. J. and John Skvoretz  (1984)  "Institutions as Production Systems."
 Journal of Mathematical Sociology  10:117-182.
 
Fielding, Nigel G. and Raymond M. Lee  (eds.)  (1991)   Using Computers in
 Qualitative Research. London, England: Sage.
 
Fiske, Donald W. and Richard A. Shweder (eds.)  (1986)  Metatheory in Social
 Science.  Chicago: University of Chicago Press.
 
Franks, David D.  (1989)  "Alternative to Collins' Use of Emotion in the Theory
 of Ritualistic Chains." Symbolic Interaction 12:97-101.
 
Garfinkel, H.  (1967)  Studies in Ethnomethodology.  Englewood Cliffs, NJ:
 Prentice-Hall.
 
Gazzaniga, Michael S.  (1985)  The Social Brain: Discovering the Networks of the
 Mind.  New York:  Basic Books, Inc.
 
----------   (1988)  Mind Matters.  Boston:  Houghton Mifflin.
 
Geertz, Clifford  (1973)  Interpretation of Cultures.  New York: Basic Books.
 
Giddens, Anthony  (1987)  "Structuralism, Post-structuralism and the Production
 of Culture."  In Anthony Giddens and Jonathan Turner (eds.),  Social Theory 
 Today.  Oxford: Polity Press.
 
Gleick, James   (1987)  Chaos: Making a New Science.  New York: Viking.
 
Griffin, Larry  (in press)  "Narrative, Event Structure Analysis, and Causal
 Interpretation in Historical Sociology."  American Journal of Sociology.
 
Greenfield, Sidney M.  (1964)  "A Unit Concept for Sociological Theory."  Indian
 Journal of Social Research 3(December):252-260.
 
Heise, David R.  (1975)  "Causality and Causal Analysis,"  in Causal Analysis.
 New York: John Wiley and Sons.
 
----------   (1979)  Understanding Events: Affect and the Construction of Social
 Action.  New York: Cambridge University Press.
 
----------   (1986)   "Modeling Symbolic Interaction."  Pp. 291-309, in S.
 Lindenberg, J. S. Coleman, and S.  Nowak (Eds.), Approaches to Social 
 Theory.  New York: Russell Sage Foundation.
 
----------  (1989a)  "Effects of Emotion Displays on Social Identification."
 Social Psychology Quarterly 52: 10- 21.
 
----------   (1989b)  "Modeling Event Structures."  Journal of Mathematical
 Sociology  14(2-3): 139-169.
 
----------   (1990)  "Careers, Career Trajectories, and the Self."  Pp. 59-84 in
 J. Rodin, C. Schooler, and
 K. W. Schaie (Eds.), Self-Directedness: Cause and Effects Throughout the Life
 Course.  New York: Lawrence Earlbaum.
 
Heise, David R. and Alex Durig  (1992a)  "Qualitative Models."  Pp. 1582-1586 in
 E. Borgotta and M.  Borgotta (Eds.).  Encyclopedia of Sociology.  New York: 
 Macmillan.
 
----------   (1992b)  "Material and Subjective Conditions for Ordered Social
 Production."  Paper presented to the American Sociological Association. 
 Pittsburgh, PA  (August).
 
Heise, David R. and John D. O'Brien  (in press)  "Emotion Expression in Groups."
 Chapter 34 in Michael A. Lewis (ed.), Handbook of Emotions.  New York: 
 Guilford.
 
Hewitt, John P.  (1976)  Self and Society.  Boston: Allyn and Bacon.
 
Kay, Paul.  (1969)  "Comments on Colby."  Pp. 78-90  in Stephen A. Tyler (ed.),
 Cognitive Anthropology.  New York: Holt, Rinehart, and Winston.
 
Latour, Bruno   (1987)  Science in Action.  Cambridge, Massachusetts: Harvard
 University Press.
 
 
Miller, G. A., Galanter, E., and Karl A. Pribram  (1960).  Plans and the
 Structure of Behavior.  New York: Holt.
 
Morgan, Gareth.  (1986)  Images of Organization.  London, England: Sage.
 
O'Brien, John D.  (1991)  "Units, Elements and Values: The Issue of Social
 Collectivity."  Paper presented to the 1991 North Central Sociological 
 Association meetings, Dearborn, Michigan (April).
 
----------  (in press)  Culture, Cause and Order.  Lewiston, New York: Edwin
 Mellen.
 
----------   (1992a)  "The Reality of Cultural Integration."  Theory, Culture,
 and Society. (submitted ms.).
 
----------   (1992b)  "Construct Realism: An Examination of Uncontested Facts."
 (unpublished seminar paper, Indiana University).
 
Osgood, Charles E., May, William H., and Murray S. Miron  (1975).
 Cross-cultural Universals of Affective Meaning.  Urbana, Illinois: 
 University of Illinois Press.
 
Popper, Sir Karl R.  (1982)  Quantum Physics and the Schism in Physics.  Totowa,
 New Jersey:  Rowman and Littlefield.
 
Pribram, Karl H.  (1986)  "The Cognitive Revolution and Mind/Brain Issues."
 American Psychologist 41(5) May: 507-20.
 
Prigogine, I. and I. Stengers  Order Out or Chaos.  New York: Bantam.
 
Raynolds, Peter A.  (1970)   The Projective-Differential: A General Purpose
 Inkblot Test for Studying Denotable Objects.  (Doctoral Dissertation, 
 University of California, Los Angeles 1969)  Dissertation Abstracts 
 International, 31, 512A.
 
Raynolds, Peter A, Raynolds, Gennie H., and Shiroi Sakamoto  (1988)  "Consistent
 Projective Differential Responses by American and Japanese College Students." 
 Perceptual and Motor Skills  66:395-402.
 
Schneider, Andreas   (1991)  "The Clustering Approach in Cross-Cultural
 Comparison."  (unpublished seminar paper, sociology of emotions, Indiana 
 University Program in the Measurement of Affect and Affective Processes).
 
----------   (1992a)  "Clusters of Meaning: A Cross-Cultural Study of the
 Authority Concept."  (unpublished seminar paper, social psychology, 
 Indiana University).
 
----------   (1992b)  "Development of a Historical Theoretical Framework to
 Explain Given Cross-Cultural Differences in Authoritative Roles."  
 (unpublished seminar paper, comparative international development, Indiana 
 University).
 
Schneider, Andreas and David R. Heise  (1993)  "The Affective Basis of
 Authoritative Roles." (unpublished seminar paper, sociology of emotions, 
 Indiana University).
 
Schwalbe, Michael L.  (1992)  "The Autogenesis of the Self."  Journal for the
 Theory of Social Behavior 21(3):269-295.
 
Smith-Lovin, Lynn and David R. Heise (eds.)  (1988)  Analyzing Social
 Interaction.  New York: Gordon Breach Science Publishers.
 
Solomon, Robert C.  (1988)  "On Emotions as Judgements."  American Philosophical
 Quarterly  25:183-191.
 
Stephenson, William  (1986a)  "William James, Niels Bohr, and Complementarity: I
 - Concepts."  The Psychological Record 36:519-27.
 
----------   (1986b)  "William James, Niels Bohr, and Complementarity:  II -
 Pragmatics of a Thought."  The Psychological Record 36:529-43.
 
Stinchcombe, Arthur L.  (1968)  Constructing Social Theories.  Chicago:
 University of Chicago Press.
 
Von Glaserfeld, Ernest  (1984)  "An Introduction to Radical Constructionism."
 Pp. 17-40 in P.  Watslawick (ed.), The Invented Reality: How Do We Know What 
 We Believe We Know?  Contributions to Constructivism.  New York: Cambridge 
 University Press.