The recursive inquiry

Material Information

The recursive inquiry a conceptualization of an information-processing sequence that derives new meanings from highly equivocal, uncertain messages and meanings
Barnett, Craig Scott
Publication Date:
Physical Description:
xi, 229 leaves : ; 29 cm.


Subjects / Keywords:
Human information processing ( lcsh )
Human information processing ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Thesis (M.A.)--University of Colorado at Denver, 1994.
Includes bibliographical references (leaves 225-229).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Arts, Communication and Theatre.
Statement of Responsibility:
by Craig Scott Barnett.

Record Information

Source Institution:
University of Colorado Denver
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
32480743 ( OCLC )


This item is only available as the following downloads:

Full Text


THE RECURSIVE INQUIRY: A CONCEPTUALIZATION OF AN INFORMATION-PROCESSING SEQUENCE THAT DERIVES NEW MEANINGS FROM HIGHLY EQUIVOCAL, UNCERTAIN MESSAGES AND MEANINGS by _Craig Scott Barnett B.A., University of Colorado, 1973 A thesis submitted to the Faculty of the Graduate School of the University of Colorado at Denver in partial fulfillment of the requirements for the degree of Master of Arts Communication and Theatre 1994


(S) 1995 by Craig Scott Barnett All rights reserved.


This thesis for the Master of Arts degree by Craig Scott Barnett has been approved for the Graduate School Wm. Michael /1-11-1'1 Date


DEDICATION The preparation and completion of this thesis began in 1987. It has come to completion mostly because my wife, Amy, and my children, Mitchell and Ashley, selflessly and often reluctantly gave of their time and interests to watch me go to the "cave" and. write. I dedicate this work to these three most precious people. vii


Barnett, Craig Scott (M.A., Communication and Theatre) The Recursive Inquiry: A Conceptualization of an Information-Processing Sequence That Derives New Meanings from Highly Equivocal, Uncertain Messages and Meanings Thesis directed by Professor Samuel A. Betty ABSTRACT The recursive inquiry is introduced as a conceptualized information-processing meani by which an individual or a group of individuals derive useful from highly equivocal, ambiguous, uncertain, complex or paradoxical messages. Examples of a recursive inquiry are provided from organizational, intrapersonal, and interpersonal contexts. Work by Claude Shannon and Warren Weaver and Karl Weick are used to show that the recursive inquiry consists in a series of interlocked processes that seek highly diverse ranges of messages from the external environment, remove the equivocation in those messages and meanings, and create new meanings based upon the relationships between and among information and meanings. iv


The relationship between these derived meanings and the resuitant behavioral responses is mapped onto Gregory Bateson's Logical Categories of Learning. After a discussion .of each of the five Learning Levels, the recursive inquiry is seen to be a highly advanced, Level III Learning activity. Blaine Goss's Information/Noise Ratio is rised to evolve an expression that interrelates meaning with information and equivocation. Discussed within this relationship is that the amount of meaning is determined by the perception of the amount of present and the removal of that equivocation. The removal of equivocation is seen to produce information and what are called relational meanings. These rela iorial meanings are subsequently tied to responses. Also, behavioral responses are identified as patterns, and through a discussion of Moire Pattern Phenomena, the recursive inquiry is discussed as a proess that significantly Ghanges responses and-patterns of responses. In this manner, a change in response terns is seen to be the result of a change in relational meanings. Peter Senge's Vision/Current Reality Gap is discussed as a metaphor distinguishing the separation v


between an envisioned pattern of behavior from the status quo pattern of behavior. The separation is also seen to be filled with equivocation, and an Equivocation Gap is introduced and discussed. In the final chapter, heuristic suggestions are made for further investigation in the areas of Construe-tionism, Social Cognition, Transformative Leadership, Intra-and Interpersonal Communication, Organizational Learning, and Metaphysics. This abstract accurately represents the content of the candidate's thesis. I recommend its publication. Signed Samuel A. vi


CONTENTS PREFACE . 1 CHAPTER 1. PRUDENS QUAESTIO DIMIDIUM SCIENTAE 9 Introduction 9 The Question 9 The Recursive Inquiry Thesis Purpose and Overview 21 2. LITERATURE REVIEW AND DEFINITIONS OF KEY TERMS Introductory Remarks . 28 The Communication Transaction: The Mathematical Theory of Communication 29 The Presence of Equivocation and Ambiguity in Three Forms of Miscommunication 47 Equivocation and Ambiguity from a Receiver's'Perspective 55 Nonstraightforward Uncertainty from a Receiver's 56 Ambiguity and the Equivocation Fallacy 62 The Processes of Establishing Meaning: Karl Weick's Social Psychology of Organizing 70 Removing Equivocation--Weick's Assembly Rules and the Double Interact o Gregory Bateson's Logical Categories of Learning: A Hierarchy for Mapping Change through Response . 107 viii


Level Zero Learning 112 Level I Learning 117 Level II (Deutero-) Learning 119 Level III Learning 124 Level IV Learning 130 Concluding Remarks 131 3. AN EXPOSITION OF THE RECURSIVE INQUIRY AND THE GENERATION OF PATTERNS 135 Introductory Remarks 1 35 Blaine Goss's Information/Noise Ratio and the .Limits of Information and Meaning in Messages 140 Relational Meanings, Responses to Relational Meanings and Patterns 151 Moire Phenomena: The Combinatory Properties and Constituent Parts of Patterns 162 Peter Senge's Vision/Current Reality Gap and the "Equivocation Gap" 169 Collaboration Activityat Learning Level II and as Inquiry at Level III 177 4. CONCLUDING REMARKS AND HEURISTIC DIRECTIONS Five Speculative Footnotes to the Recursive Inquiry 191 Elaborations into Current Communication Theory 199 Constructionism 202 Social Cognition 204 ix


Transformative Leadership 205 Bateson's Logical Categories of Learning and Communication . 206 Elaborations into Intrapersonal and Interpersonal Communication . 208 Intrapersonal Communication 209 Interpersonal Communication 213 Elaboratioris into Pattern arid Mega-patterns Elaborations into Organizational tearning 219 Elaborations into Metaphysics 220 . 223 REFERENCES 225 X


ACKNOWLEDGEMENTS This thesis was made possible because of an incredible amount of support and encouragement from so very many people. Many of these instances of support carne at critical times for me. A special thanks is due my family--especially Loretta and Junior Weeks, and my mother. Also, a special thanks is due Dr. Nancy Pokorny for her assistance and sympathetic understanding. Finally, Bill and Lizzie Carder deserve a most special thanks for all their unselfish interest in me over these many years. Also, I wish to express my to my committee members for their patient and persevering work on an admittedly massive final piece. Through my association with each of them, I was duly rewarded with critical ideas and thoughts that are interwoven into this thesis. Without these valuable contributions this work could never have been accomplished. In particular, I wish to thank Professor Sam Betty for directing this work and his unabashed enthusiasm for continuation of it through its darkest hours. Also, Professor Mike Monsour deserves a most hearty thanks for shaping this work into an intelligible, readable, and cogent piece. He is an unselfish and patient champion of quality. xi


PREFACE This thesis has grown out of the following series of events that occurred in 1987. As the supervisor of our telemarketing operations, part of my duties included participation in the preparation of the annual budget. Like many businesses, ours set its expected growth goals in terms of a "bottom line" total, and our department's specific goal that year was an increase of 10% in "bottom line" contributions over the previous year. Having established this sales department goal, the manager of the department and spent many days behind closed doors tinkering with specific line items, considering what training would be most beneficial and would produce the greatest sales dividends, tightening our bad debt collections procedures, reducing our customer service expenditures, and establishing individual employee goals to achieve these planning targets. In short, our strategic planning was based upon rational consideration of "hard" data in the form of dollar totals, manipulation of the data into a "desired" form, and projection of "safe," "easily obtainable," "modestly bold" sales goal 1


increases to make up the difference or defray any contingency surprises. This was how we implemented Harold Geneen's "Three Sentence Course on Business Management." According to Geneen, "You read a book from beginning to end. You run a business the opposite way. You start with the end, and then you do everything you must to reach it" (Geneen & Moscow, 1984, p. 35). After all, if such an approach had built the ITT conglomerate, it was good enough for our small sales department. To support our belief, we had never experienced a sales shortfall. In truth, we had modestly exceeded our annual goals. My dissatisfaction with this approach to management began with the relatively secondary nature of sales and customer service in the scheme of the growth and development of our Rather than devoting days of thinking to improving sales and customer servi6e, we had spent days devoted to thinking of ways to hit the sales and customer service budget targets. While these two activities may seem similar, they are not at all similar. Once, I asked myself the question, "What do these saies figures actually mean?11 A large number of surprising answers emerged. A couple of these answers both transformed my thinking on management and caused me to 2


think about ideas that would coalesce into this thesis. The two most startling revelations were: 1) These sales totals represented precisely what our customers considered to be the perceived value of our products to them. 2) These sales totals represented what our customers to be the quality 6f service we provided to them. As I considered these responses in greater depth, I not only recognized that they were intertwined but I also suspected that if we continued tinkering with the results (the "hard" numerical data), we would continue to blur the answer to one of the most critical of business qtiestions: were we actrially condUcting better, more effective businegs as the_sales data indicated or were we merely getting better at manipulating the sales data? To my way of thinking, we had been conceptually c6nducting business backwards. we had the horse pushing the cart. bur sales results were dictating in advance what our day-to-day efforts were to be rather than measuring what our day-to-day efforts were accomplishing. As I considered this matter further, I saw that we were trying to squeeze everything we could otit of 3


the results and assuming that this activity meant we were conducting better business. As we were to discover later, better results do not automatically signify better business. What we did discover was that better business practices lead to better business, and the effectiveness of these practices is what is by reviewing the results. If we instituted better practices, the results necessarily had to refledt this improvement. If we were instituting ineffective practices, the results ought to show no change or decline. While all this may seem like a review of elementary business principles, the appiication of this "discovery" had a profound impact on our company. In our case, the question, "How do we improve our bottom line by 10%?" had caused us to focus on results. To refocus on business practices, we needed to ask ourse1ves the question, "How do we improve the quality of our service to customers by 10%?" This new question caused many unforeseen and startling changes to occur. It not only put our thinking about results into the right perspective, but it also put sales and customer service into the right position in our business. Moreover, a dramatic shift in both our strategic organiza-4


tional thinking and development occurred. Also, a leap in departmental performance occurred. Among these shifts, the following ones were most evident: 1) "Sales" expertise gave way to a more desirable expertise: "learning" skill. Our salespersons became "students" of quality and about what constitutes service and the delivery of quality service. 2) Our employees learned that quality service skills were life skills, and service skills learned on the job had direct bearings on improved interactions at home with family members and in other social situations. Conversely, these outside-the-job interactions honed and improved service skills that could be used and shared in the job place. All employees felt a broadening sense of purpose that more completely integrated the job with off-the-job activity. 3) A heightened sense of camaraderie with others within the department developed along with the related senses of greater esprit de corps, trust, respect, support, and, curiously enough, a developing faith in the new process. 5


This latter issue of 11faith in the process11 was exceedingly difficult to foster. Belief that truly qualitative activity will eventually result in increased revenues proved to be a long-range investment. For some of our employees, this seemed to be too speculative a gamble. While all employees rationally acknowledged that improved service must necessarily cause improved sales totals, quite a few hearts grew faint waiting for the raw data to confirm or deny whether we were headed in the right direction. In one notable example, one of our telemarketers was actually heard to talk a customer OUT OF purchasing a $200 item. In the employee's estimation, the customer had mistaken the contents of the product. The employee insisted that the customer would not be making a good purchasing decision. To our initial horror, the employee a competitor's product to the customer. Four months later, the customer called back, requested the employee by name, and ordered additional products for his branch managers that exceeded 10 times the original purchase. He noted that the advice he had been given previously had been entirely correct. Finally, as a result of this favorable experience, the customer provided us with many 11referral11 customers. 6


As more employees experienced these kinds of transactions, confidence and faith in the process took hold. Our people began to discover that an improved understanding of quality led to improved service levels and that these improved .service levels were entirely measurable. More importantly, they learned that sales results will inescapably take care of themselves when the individual's effort is entirely focussed on improving skills in those activities which the results are measuring. In our case, these activities were defined as "qual ity transactions" with customers. To our department members, the term "transaction" later carne to mean "qual ity exchange of information" rather than "exchange of money for product." Viewed in this way, a "failed" sales call could resulted in money and product exchange, but the "failure" would have consisted in the weakness or ineffectiveness of the interpers6nal transaction. As the department wrestled with the quality question, new concerns and ambiguities began to emerge. For instance, what does "improving quality by 10%" mean anyway? The original question seemed to be opening itself up into subtler and broader matters. We learned that improved quality service entails improving interpersonal communication skills. Did this mean that in order to 7


improve quality service by 10%, we would have to improve interpersonal communication skills by 10% also? or would a service increase of 10% necessitate improving interpersonal communication skills by 15%? And what is meant by "improving interpersonal communication by 10%11? Frustrating and challenging as these new questions were to us, we aiso that we were becoming a different bettei organization. We had become a vibrant, learning department, and we were making a difference to our customers. bottom line contributions in 1987 were 21% over goal. The following year, they were 26% over goal. Clearly, something unusual had happened. question about quality had pr6mpted some extraordinary occurrences. The attributes associated with this brought to my mind. other questions that seemed to have no clearly definitive answersand also seemed to open up into subtler and broader directions . I steered my Master's studies toward understanding more about the general natures and attributes of this class or "genre" of questions. 8


CHAPTER 1 PRUDENS QUAESTIO DIMIDIUM SCIENTAE ("To know what to ask is already to know half") --Aristotle Introduction The purpose of the next two sections is to fill in the conceptual gap between the events chronicled in the Preface and the evolution of the notion of the recursive inquiry, which is the subject of this thesis. This discussion will introduce two fundamental contexts in which this thesis is framed. These are: 1) The relationships between recursiveness and epistem-ology. 2) A communications on pro-cessing highly equivocal information inputs. The final section in this Chapter will be devoted to stating the purpose and outlining remaining three Chapters of this thesis. The Recursive Question Questions are rhetorical means of gathering informa-tion. In commonplace communication, questions help to specify what kind and what amount of information is 9


required of a source by a receiver. Some commonplace questions are phrased to request short, very specific responses. Other commonplace questions are phrased to request longer, more general responses such as descriptions or opinions followed by support or examples. Traditionally, the former questions are called "closed" questions, and the latter are called "open-ended" questions. Normally, these kinds of questions convey a sense of "self-containment." That is, they are posed to solicit some finite amount of information. When enough information is believed to have been gathered, then a sense of closure is reached. The inquisitor has an acceptably complete With this sense of closure, the exchange is complete and a .new exchange is possible. This selfcontainment does not imply that the information cannot be amended or replaced altogether with new information at some later date. Our "quality" question did not fit into either the "closed" and. "open-ended" question scheme. First of all, the quality question lacked any sense of "self-containment." Once we felt that we had a newer and deeper understanding of "quality," some new nuance or subtlety emerged, and we felt that we were back at square one. With the emergence of these kinds of broader and subtler implica10


tions, our deeper understanding of "quality" was balanced against our growing recognition of how much more remained for us to learn about it. In other words, while new learning was a direct outcome of our activity, one outcome of this new learning was a deeper knowledge of what we did not know. Secondly, the quality question seemed to act like a sort of catalyst--unifying other crucial, organizational learning endeavors rather than existing as a separate, discrete activity. As illustrated in the Preface, we discerned early on that any quality customer transaction required high-quality communication skills. Employees initially resistant to developing and improving communication skills realized that communication adeptness was pivotal to our quality goal and was not merely a matter of secondary peripheral fluff that was nice but not essential to the interaction. As such; our quality question seemed to "tie together" organizational efforts that had been treated in the past as discrete and separate activities. Finally, the quality question possessed what Hofstadter (1979) and Pearce, Cronen, and Harris (1982) have referred to as a "strange loop": 1 1


[This] phenomenon occurs whenever, by moving upwards (or downwards) through the levels of some hierarchical system, we unexpectedly find ourselves back where we started (Hofstadter, 1979, p. 10). From the very beginning, the quality question led us into a hierarchically structured process. New layers of learning were being built upon preceding efforts, and the newest layers were laying the foundation for succeeding efforts. While a detailed treatment of this hierarchy of learning is reserved for Chapters 2 and 3, a simpli-fied overview of our subsequent activity will serve to characterize the 11strange loop11 nature of our quality question. Our initial activity was to pool current think-ing and understanding of the quality issue. Next, we asked ourselves the question, 11Knowing what we now know about quality, how do we improve the quality of our ser-vice?" We then commenced gathering information to answer this question. Pooling our was an ongoing process as was new information gathering. Soon, we found ourselves processing this information--trying to figure out what all this new information meant. Inevitably, we found ourselves right back at the original question: "[Taking into consideration what we now know] how do we improve the quality of our service?" While this return 1 2


to the question happened over and over again, we were not covering the same informational ground again. In fact, because we were enlarging our pool of relevant new information, we were broadening the boundaries of our next information gathering efforts. In effect, we were opening up and discovering "newer'' informational boundaries and sources. For those of us involved in the information gathering and processing, new learning connections and were being made with each new "strange loop" cycle. Both Hofstadter and Pearce, Cronen, and Harris relate this "strange loop" phenomenon with-recursiveness. For this reason, the term recursive question was coined to name the class of questions exemplified by our quality question and to distinguish it from "closed" and "openended" classes of questions. A "recursive" question was defined as any question that resulted in a body, or corpus, of information and meanings that were subsequently used as a foundation to build a more comprehensive corpus of information and meanings through a repetitively employed system of cycles. The "size" of this body of information and meanings will always be dependent upon the number of cycles the recursive question undergoes. As more cycles are pro-1 3


cessed, more information is integrated, and more meanings are distilled. By contrast, much smaller-sized, more self-contained packages of .information result from either 11closed11 or "open-ended" questions. While the development of these ideas about recursive questions formed a problem surfaced that redirected attenti6n urther development of thinking about these types of questions. In an exploratory study (Barnett, Harrington, & O'Neill, two demographically sets of students from two different colleges were polled in an effort to if connections existed between recursive questions and the general in-class and out-of-class learning activities of students .. It was expected.that recursive questions could be demonstrated to have an impact on how well integrated and purposive student learning would be. Those students who could identify important and specific recursive questions as guiding their active learning ought to be able to integrate learning activities better than those students who could not identify a recursive question directing their learning. While existence of recursive questions in normal activity was demonstrated among a measurable segment of the studehts, few other valuable conclusions could be drawn from the study. 1 4


The major problem with studying recursive questions was seen to be that "recursiveness" was not necessarily apparent in the syntax of the question. The true test for recursiveness was discovered to be found in the manner with which the question was interpreted and subsequently responded to by the individual. The presence (or absence) of these interpretations and responses was hot found to be easily demonstrated by observers. For instance, the investigators interpreted the question "Why am I here?" and other such ontological questions as recursive by definition. Conversely, the question "Why did. I choose to go back to college?" was interpreted to be The reasoning for this latter interpretation was the assumption that the question would prompt a distinctly limited, self-contained corpus of information. However, further consideration suggested that the former question could have been posed by a student who had intended to mean the latter question. In short, the ontology or recursiveness of any question was found to be determined solely by the respondent's contextual interpretation of the question. No question is necessarily recursive, but any question can be made to be recursive. Recursiveness is determined by an individual's perceptions of the question's intent, purpose, and contextual semantics. 1 5


These ambiguities surrounding intent, purpose, and contextual semantics further study of recursive questions of uncertain value. Moreover, while recursive questions by definition generated recursive responses, it was recognized that recursive responses were not limited exclusively to the realm of questions. Recursive responses can be generated from other contexts as well. For example, "strange loop" recursiveness could be found in an individual's response to paradox arid contradiction. The Zen koan proved to be a striking illustration of this. As a new focus was brought to bear on these responses that are also recursive, a vaster, more generalizable and promising thesis emerged. The Recursive Inquiry The purpose of any Zen koan isto provide a learning exercise for a Zen initiate. Paradoxes form the lifeblood of most koans, and the Zen initiate is expected to resolve a koan's paradox before "graduating" to the next koan. One of the better known Zen koans is "Describe the sound of one hand clapping." A stylized description of an initiate's response to the "one hand clapping" koan will serve to illustrate what is meant by a "strange loop,11 or recursive response. 1 6


First of all, the Zen initiate must discern the context in which the exercise is being posed. In a general sense, two context options are reasonable. One option is an entirely literal interpretation of the Zen master's statement. The second option is to interpret the message as a purposefully posed paradox--a contradiction. In response to the first option, the initiate would probably say, "Nothing. One hand cannot clap." More likely, the initiate would interpret the message in the second, or what we may call an "ambiguous" context. This perspective would be entirely consistent with the initiate's familiarity with the koan format. Recognition of this context rules out the possibility of a interpretation and a literal response. From this context, initiate has discerned that his/her task will be to unravel the paradox. Assume that after a time of contemplation, the initiate returns to the Zen master and says, "the sound of one hand clapping is the sound of half an approval." The Zen master responds, "Good. Now, describe the sound of one hand From the master's the initiate has now learned: 1) that the ambiguous context was a correct assumption. 17


2) that the response was acceptable within that ambiguous context. 3) that at least one more response is necessary and that this response must take into account the correctness of the initial response. It should be noted that the initiate has learned some valuable information by resolving some of the ambiguity that was inherent in the original message. Put another way, the initiate has reduced the koan's initial amount of ambiguity by increasing the amount of known information and meaning. While armed with this new information and meaning, the initiate has nevertheless-returned to the starting point. The initiate has completed-one "strange loop." In the mind of the initiate, the koan now reads, "[Knowing what I know from this 'new learning'] describe the sound of one hand clapping." Assume that the initiate returns still later to say that "the sound of one hand clapping is the sound of purposeless motion." Also, assume that the Zen master's response once again is, "Good. Now, describe the sound of one hand clapping." At the completion of this second "strange loop," reaffirmation of the initiate's original learning has 18


occurred. In addition to this, the initiate has learned: 4) that there exists at least another correct response. After another repetition of the transaction or after some future "strange loop" cycle, the initiate is expected to learn: 5) that there exists a response which encompasses the set of correct responses previously given. That is, there is a response which generalizes the previous responses and which when correctly applied could also generate new, predictably correct responses. This description of a response-generator is the goal of the koan and is, in truth, a type of algorithm. Once the initiate articulates this generalized rule or condition, the exercise is concluded. The "one hand clapping" koan causes a recursive response the moment that a contradictory or paradoxical condition is perceived by the receiver (our initiate) to be present in the message. Even if the initiate had been foolish enough to have assumed the Master's koan was ded for literal interpretation, the response "Nothing. One hand cannot clap" would have been greeted with "Now describe the sound of one hand clapping." Inevitably, the ambiguous context would have to be recognized by the 1 9


receiver, and the paradoxical condition would have to be perceived and responded to. This condition of paradox arises independently of the syntax used to introduce the condition. Consider the following messages and the syntax used to convey them: 1) "Describe the sound of one hand clapping." 2) "Tell me the sound of one hand clapping!" 3) "What is sound of one hand clapping'?" This latter message, of course, is in a syntax referred to previously as a recursive question. Three important conclusions were drawn from consideration of the Zen koan and paradox in general: 1) The degree of recursiveness required of succeeding responses is dependent upon the amount of ambiguity uncovered by the previous responses. (Ambiguity, uncertainty, and equivocation will be discussed in greater depth in Chapter 2.) 2) One outcome of any recursive activity is the conversion of ambiguity into usable information and meaning. 3) A second outcome of any recursive activity is the development of a learning hierarchy (initial responses and responses that can generate other subsequent responses). 20


My original focus on the recursive question had led me to concentrate on a source-dominated communication transaction and on linguistic issues. By refocussing on recursive responses, I was able to shift my efforts to what became a receiver-dominated communication trans aetion and on more epistemological issues. Moreover, concentration on what later became known as recursive inquiries allowed me to broaden my considerations from ambiguous questions to any ki-nd of generalizable condition which a receiver.might perceive to be ambiguous. In addition to recursive questions, conditions include paradoxes, contradictions, and even 11double binds" (Bateson, 1969). Thesis Purpose and Overview The purpose of this thesis is to define and describe that communication activity which will be referred to hereafter in the thesis as a recursive inquiry, the processes that constitute this activity, and the outcomes of this activity. In the course of this definition and description, answers to the following questions will necessarily be explored: 1) What is the role of the receiver in any highly ambiguous communication transaction? 21


2) What intrapersonal (cognitive) behaviors and interpersonal (social) relationships compose the processes which convert ambiguity (and uncertainty and equivocation) into usable information and meaning? Related to this question is: What intrapersonal behaviors and interpersonal relationships subvert the resolution of ambiguity (and uncertainty and equivocation) into usable information and meaning? 3) What formal relationship exists among the variables of information, meaning, and equivocation? 4) What relationship exists between the recursive inquiry and a) hierarchical learning levels? b) advanced forms of .learning? c) individual and organizational learning? d) creativity and collaboration? e) communication theory in general? Chapter. 1 has been devoted t6 elaborating the evolution of the thesis topic. In particular, the notions of recursiveness, learning, ambiguity, information, and meaning have been introduced as related to one another under a rubric which we have called a recursive inquiry. Moreover, this chapter has established that a discussion of the recursive inquiry is bound up with a view of the 22


receiver as a discoverer of new kinds of information and meanings. Chapter 2 combines the Literatqre Review and Definitions of Key Terms. These were conjoined in an effort to maintain the connections that clusters of terms have with their literature heritages and how the clusters and various heritages dovetail. These c6nnections would have been more difficult to make if the Review and Key Terms were more formally separated. There are roughly three general clusters of terms: 1) Those terms referring broadly to any communication transaction between a source and receiver 2) Those terms referring to the processing of information and the subsequent assigning of meanings 3) Those terms referring to a hierarchical learning structure The nucleus of each cluster of terms contains a central body of theoretical literature. At the heart of the first cluster is the Mathematical Theory of Communication (Shannon & Weaver, 1949). The Social Theory of Organizing (Weick, 1969, 1979) is at the heart of the second cluster. Gregory Bateson's Logical Categories of Learning and Communication (1964) forms the nucleus for the third cluster. Extending from these 23


nuclei are related bodies of theory and research which are cited as they elaborate or clarify terms or as they establish additional connections between terms. For instance, definitions for the terms noise, ambiguity, equivocation, and uncertainty have become muddled through both terminological and conceptual diversity. In the discussion of the first cluster of terms, carefully drawn distinctions are made in this thesis between.the four terms. These distinctions are developed through a discussion of the representative literature. Moreover, once the distinctions are drawn, the term equivocation is expounded further through a rhetorical "equivocation fal lacy11 example (Ferguson, 1.9 91 ) _and through the development of Senge's Vision/Current Gap concept (1990). Not all terms have their roots in the literature. Some are of my own invention. Included among these are 11source-specific11 and "source-nonspecific equivocations." Also, the definition for 11pattern" discussed at the end of Chapter 2 was drawn exclusively for its utility in clarifying Chapter 3 ideas. Unless specifically cited, definitions should be construed as derived by me for the thesis. Chapter 3 consists of five sections. Together, these five sections form the theoretical exploration of the 24


thesis. Chapter 3 applies the concepts grounded in the literature and presented in Chapter 2 to the communication activity known as recursive inquiry and extends these concepts. Chapter 3 presents the discussion intended to explore answers to the questions listed earlier. These five sections may .be summarized as follows: 1) A discussion of Blaine Goss's Information/Noise Ratio (1989) will open the chapter. From this Ratio, a formal relationship will be derived that links the relationships between Information, and Meaning together. A discussion of the implications of this relationship will follow. 2) The term Pattern will be introduced to describe the behavioral change that resuits from a receiver res to a new meaning. In conjunction with this term, a consideratio.n of Moire Phenomena and the attributes of patterns combining with other patterns to form new patterns will be applied to the recursive inquiry. 3) Peter Senge's (1990) Vision/Current Reality Gap Metaphor will be introduced and from it, an Equivocation Gap will be identified to provide a metaphor appropriate to an understanding of how the recursive inquiry may be linked to enactment of the environment. 25


4) Using a context of collaboration activity, Level II and Level III communication activity will be examined in an effort to distinguish differences between these two levels. The recursLve inquiry will be shown to be an example of Level III activity. 5) The discussion of Level III Patterns of Learning will conclude with an examination of the implications of the recursive inquiry to Level III Learning. The recursive inquiry will be shown to be the generalizable activity by which meaning is created within this advanced learning level. The recursive inquiry will be prbposed as a viable activity for creating meaning out of highly ambiguous or equivocal messages and, hence, highly complex matters. Chapter 4 concludes this thesis and explores possible inquiries which for testing the validity of the ideas presented and which seem of value to already existing and to new arenas of investigation. Chapter 4 begins with two observations and three cautions about the recursive inquiry which need to be considered. The observations remind us that the recursive inquiry is an activity that does not require advanced educational training and that the recursive inquiry 26


is extremely sensitive to self-interest and will be undermined by self-interest. The three cautions cover issues of morality and ethics, the propensity to lower expectations when the recursive activity becomes more difficult, and the very real matter of disenfranchised individuals unable to participate in the inquiry. The largest portion of Chapter 4 explores heuristic possibilities resulting from the theme of this thesis. Four areas of communication theory are reviewed: 1) Constructionism, 2) Social Cognition, 3) Transformative Leadership, and 4) the Logical Categories of Learning. Thereafter, questions and thoughts will be expltired within the contexts of Intrapersonal and Interpersonal Communication. The chapter will conclude .with questions and thoughts applying to matters of patterns and mega-patterns, Organizational Learning, and Metaphysics. 27


CHAPTER 2 LITERATURE REVIEW AND DEFINITIONS OF KEY TERMS Introductory Remarks Because of the nature of this thesis, the literature review and definitions of key terms were conjoined and presented together. Difficult as this may make the reading of the chapter, presentationof the material in this manner simplifies the exposition of Chapter. 3. It should also be understood that this is not an entirely comprehensiVe literature review. Many relationships may be drawn between what will be presented and fine theoretical and heuristic traditions which will not be cited. While the reader may find some of these relationships obvious, a more comprehensive review lies outside the boundaries of this present work. Also, the literature review does contain extended thoughts apart from--but inextricably related to--the works and authors cited. These relationships will be drawn formally in Chapter 2 rather than re-cite and then re-develop the relationships in the body of Chapter 3. These digressions are taken to lay the foundation for the concepts that are built in Chapter 3. 28


This chapter is broken down into four sections: 1) A description of the communications exchange in the context of Claude Shannon and Weaver's Mathe-matical Theory of Communication (1949) 2) A description of relevant matters associated with equivocation, ambiguity, information, and meaning 3) A description of Karl.Weick's Theory of Social Organ-ing (1969, 1979) as it relates to the processes of deriving meaning from information and message inputs 4) A description of Gregory Bateson's Categories of Learning and Communication (1964) as relate to how an individual responds to information and levels of meaning Altogether, four sections describe what constitutes a message, what makes the concerns of a recursive inquiry complex and difficult, what processes constitute a recursive inquiry, and what of learning result from a recursive inquiry. The Communication Transaction: The Mathematical Theory of-Communication The presence of ambiguity, equivocation, and uncer-tainty has already been introduced as pivotal to this thesis. Also, examples have been provided of a kind of 29


activity referred to as a recursive inquiry. The amount of recursiveness in an inquiry has been introduced as being dependent upon the amount of ambiguity or equivocation or uncertainty perceived to be in the original message. Claude Shannon and Warren Weaver's Mathematical Theory of Communication (1949) is suited for investigating these matters and others attendant to them. A review of this Mathematical Theory of Communication (the MTC) will serve to act as an introduction to the communication transaction as in this thesis. In The Mathematical Theory of Communication (1949), Claude Shannon and Warren Weaver describe how a message may be transmitted from one site to.another and what may happen to that message during the course of its transmission. While the distance between the two sites has a bearing on how the Model is applied the general Theory is adeptly applicable to sites of any distance from one another. It is important at this point _to distinguish the differences between transmission, transaction, and communication. A transmission is any message that is sent from one site to another site. Implied within this definition is that the second site merely receives the message. The second site does not have to respond to nor even under30


stand the substance of the message. The second site simply has the equipment necef:?sary to "receive" the message. One may say, for instance, that for millenia, the visible stars have been transmitting stellar messages to us. With the invention and application of the spectroscope to these messages, scientists were able to "understand" these messages contained information. That is, the message from one star could be understood to be distiguishable from the message sent by another star by its unique spectrograph. With the spectroscope, scientists may be said to be capable of communication stars in a Shannon and Weaver sense. Communication o6curs when a transmission is recognized as a message and the mean ing is responded to by a second site. This response to a message's meaning is the overt or visible sign by which a simple act of transmission causes a second, responsive transmission. This second transmission is frequently referred to as feedback to the first message. Obviously, interstellar communication is uni-directional. The stars and we "listen." To the best of our knowledge, no matter how much we talk back, the stars are not inclined to respond to us nor alter their transmissions because of us. Human communication, on 31


the other hand, is richly complex and multi-directional. Talkers and listeners change and interchange roles constantly in normal communication. In order to distinguish this multi-directional form of communicating from the uni-directional form, the term transaction is introduced. Communication transactions exchanges of messages between or among sites. These exchanges are characterized by transmissions emitted from different sites and feedback transmissions between and among the different sites. In the MTC, a source is any site from which a message is transmitted. A message is defined as a format by which data or information is first organized and then conveyed to another site. This destination site is referred to as the receiver. The chosen mechanism by which a message is conveyed is known as the transmitter. -According to this definition, a transmitter could be an electronic device or a written letter or a sound emitter. The through which the message is transmitted is_ known as the channel. The transmitter and channel are interrelated entities. _For example, verbal exchanges are best conducted in channels conducive to sound transmissions. Under water, verbal exchanges are ineffective; body gestures are a more conducive transmission when water is the transmission channel. 32


This conceptualization of a source transmitting a message to a receiver over a channel that is amenable to the transmission is what was referred to earlier in this section as 11how11 communication occurs. Beyond this, the MCT also describes 11what happens" during the exchange of any message. This second focus is crucial to understanding more about recursive According to Shannon and Weaver, any information becomes a message precisely at the time when it is organized and duly prepared for trartsmission. This process of organizing and preparing for transmission is called encoding. That is, a code is selected by a source that lends itself to the organizing of the information to be sent, and encoding is the process performed by the source to convert the information into the code so that it may be sent. Obviously, the selection of an appropriate code is not always a cut-and-dried matter. There are a myriad of possible considerations for selecting the best code for conveying any given message. Among these considerations are: 1) What code is best given my degree of fluency? 2) What code is best given the degree of fluency of my receiver? 33


3) 4) 5) What code is What code is I must send? What code is must travel? best given the channel I must use? best given the amount of information best given the distance the information In a vast number of human communication transactions, a variety of constraints limit or prohibit adequate opportunity for appropriate code selection by a message source. Inadequately chosen codes can have disastrous effects on the effectiveness with which communication occurs. For instance, while in high school, I took German classes. My great grandmother, a non-English speaking immigrant from southern Germany, was ecstatic to learn of my study. After the first week of classes, I approached her proudly and said, "Grossmutter, Ich kann Deutsch gesprechen jetzt." With equal pride, she launched into a 5-minute, 75 mile-per-hour German monologue. I understood nothing. My great grandmother had understandably assumed that a coding barrier between us had been crumbled away because of my German studies. She assumed we had a code in common when none of substance really existed. Even the smallest sampling of human communication transactions is littered with unfortunately chosen codes that impede or obstruct 34


rather than facilitate or enhance interpersonal communication. The counterpart to encoding a message is the process by which a receiver converts a message back into information. This process is called decoding. In my case, then, my great grandmother had transmitted an information-laden message in a code I was incapable of decoding. If code selections can enhance or detract from what happens to a message, Shannon and Weaver identify another, even more pervasive root of communication enhancement or detraction. In the MTC, the term uncertainty is introduced to name the relationship exists between the amount of information a source intends to .be encoded into a message and the actual amount of information a receiver decodes from the message. Without describing this relationship in complex detail, Shannon and Weaver note that a direct relationship exists among and between the following three variables: 1) The amount of freedom a source has in selecting the constituent parts of the message 2) The amount of information contained within the message 3) The amount of uncertainty contained within the message 35


One way to interpret these variables is to say that any source may purposefully increase the amount of uncertainty that a message may contain. The Zen master encoded such a highly uncertain message for the initiate in Chapter 1. According to Shannon and Weaver, the amount of purposeful uncertainty corresponds directly to the amount of information the source has embedded within the message. Following this uncertainty/information relationship a little further, it can be seen that as the Zen initiate responds to or interacts with this uncertainty, more and then still more information-is culled from the message. As more information is extracted from the message, the less uncertainty will remain. Moreover, the amount of original uncertainty in the message corresponds directly to the amount of freedom the Zen master had in constructing the message. For instance, the Zen master chose to present the koan paradox from a large variety of possible message formats. He/she might have chosen from a set of more direct, less metaphorical messages. As argued earlier, he/she might have choseri a question or command to convey the message. A narrative might have been chosen and so forth. Ultimately, the freedom involved in the message selection process corres-36


ponds to both the effectiveness and the amount of intended information is conveyed. The genius of poetry reflects this same notion. From a myriad of possible choices, the grand poet carefully selects only the precise word or phrase to convey precisely the intended feeling or image. In the tradition of recognized Classical Poetry, it is arguable from Shannon and Weaver's perspective that the carefully mulled-over, precisely chosen word contains vastly more meaning than a word chosen haphazardly or capriciously because the purposefully chosen word was selected from a commensurately larger range of "possible" choices than the capriciously chosen word. We may say that the a source uses to de-select certain choices gives the final choice a deeper meaning. The de-selected choices and the reasons for their de-selection form a richly complex context in which the actual choice is delivered. Shannon and Weaver this purposeful uncertainty as the "desirable" form. They note this desirable form is what is commonly referred to as the novelty and creative expression so vital to the enhancement of effective communication. They argue that if a message were entirely devoid of desirable uncertainty, then by definition the message would be entirely predictable. In the 37


presence of a perfectly predictable message, the receiver would know in advance the entirety of the message and would receive absolutely nothing new in the way of previously unknown information. This notion of predictability in messages is called redundancy. .In view of the discussion so far, a message's amount of redundancy has an indirect relationship to the amount of information it contains: an infinitely redundant message contains precisely zero information. Up to this point, uncertainty has been discussed as an choices added to a message by the message's source. Shannon and Weaver define a second form of uncertainty. 11Desirable11 uncertainty is counterbalanced by that uncertainty which is 11undesirable.11 This second type arises from the unintended intrusion of uncertainty into the message. These intrusions are considered 11unintended11 because they are largely out of the control or outside the intention of either the source or the receiver. Unintended uncertainties by definition damage the integrity of messages. Inadvertent errors made by the source while encoding information into a message are examples of this type. Damage of this kind usually alters the content of the message and as a consequence, frequently 38


results in incorrectly interpreted meanings. By far, though, the most common sorts of 11undesirable11 uncertainty come as a result of intrusion by external factors during the transmission of the message through the channel. This latter type of uncertainty is commonly known as 11noise." Noise may be defined as any external interference that in some way alters, damages, or obliterates any message and, as a consequence, its information content. Static on a radio is the result of some atmospheric influ ence that alters or damages the normal radio message. The static is the altered message resulting from the impact of the atmospheric effect on the.originally undamaged message. In a theoretical sense, noise has a broader connotation than that which is commonly While external influences may include 11auditory chaos,11 or ncaco phony,11 there are numerous other external influences that are considered noise. For instance, the planet Pluto was discovered as a result of an inexplicable intrusion on the normal and redundant messages being sent from the planet Neptune. The alteration of the normal messages implied gravitational influences. In this astrogeophysical case, gravity from an unseen source was the noise source altering the Neptunian messages. The noise source was 39


hypothesized as of planet size. Pluto was discovered shortly thereafter--confirming the of the noise source. To summarize, Shannon and Weaver state that uncertainty is a normal phenomenon in communication. In one sense, uncertainty is directly to the amount of information contained within the message. This type of uncertainty is a clue to the receiver as to the amount of information that was originally encoded into the message by the source. This kind of purposefully encoded uncertainty is, in fact, necessary for information to be distinguished by the receiver as informatitin. A second type of uncertainty is also normal to communication but is a deterrent to effective communication. This second type may be subdivided into two subtypes: 1) Uncertainty arising from unintended encoding errors 2) Uncertainty arising from external intrusions Shannon and Weaver define this &econd type, or "undesirable" uncertainty, as equivocation. Seen another way, equivocation is the total of_all the differences between the originally coded message by the source and the actual message that was received. The equivocation or discrepancies resulting from the differences between these two 40


messages must be caused by one or both of the subtypes of undesirable uncertainty. To the first type, or "desirable" uncertainty, Shannon and Weaver ascribe the processes of entropy. This presence of entropy corresponds to the relationship described earlier as that which exists between freedom of coding selection and information content. For the purposes of this thesis, though, uncertainty that is purposefully encoded into a message by the source will be defined as ambiguity. It is intended that ambiguity will replace the term desirable uncertainty. Distinguishing equivocation from ambiguity in an actual message is a reasonably simple matter for the source of the message. Since the source will have deliberately nested any ambiguity into the message, any other disparities occurring within the message during its transmission and receipt must necessarily be attributable to equivocation. It is from this framework that the MTC elaborates its many brilliant applications. I refer to the MTC as a source-based Theory for the simple reason that the mes sage's source has the advantage of being able neatly to separate those uncertainties that are ambiguous from those that are equivocal. When describing communication from 41


a source's perspective, ambiguity and equivocation are convenient contrivances. On the other hand, from a receiver's perspective, the distinctions between ambiguity and equivocation are considerably murkier. It is not easy for a receiver to recognize those portions of a message that are intentionally encoded ambiguities from those errors that are results. of equivocation. Whenever a receiver finds him/herself pondering the question: "Did the person really mean to say this?" then the receiver is wrestling with the matter of whether the content was intentional (ambiguous), erroneous (eqbivocal), or worse, partiaLly intended and partially erroneous .(ambiguous and equivocal). For example, consider a married coupie. One spouse is away on a business trip. The at home and receives the following telegram: HONEY DONT HAVE A COW. I JUST BOUGHT A BOAR. How is the spouse to interpret this? First, .let us assume that the spouse at home begins to interpret the message as a reliable, error-free message. Uncertainties of meaning are necessarily the result of ambiguity--intentional coding choices. The message, as it stands, seems internally consistent but seems hopelessly incomplete and high levels of uncertainty remain. Why 42


in the world buy a boar? Who is "honey"? Me (I hope)? Someone else? Since wild animals are illegal in the city, does this mean we are moving to the country? To another country? Does "cow" refer to a four-legged animal (as "boar" suggests) or is it a euphemism? If the spouse assumes the message's confusion is the result of a garbled transmission or errors committed by incompetent telegraph agents, then the uncertainty of the message will be assumed to be equivocation based. But, the spouse asks, where does the correcting begin? Is "boar" a misspelling for "boat"? Are there words missing? Are entire sentences missing? Was message switched with another message intended for someone else who is in for a "boar"? Was this message misdelivered to me by mistake when intended for someone else withmy name? A third option allows the spouse to assume that both ambiguity and equivocation are within the message. Obviously, this option compounds the uncertainty of the message since it opens up both sets of possible options and requires the receiver to pursue both avenues of uncertainty. Returning to the Zen initiate, recall that one of the first tasks was to determine how the master's message 43


ought to be interpreted. Is the message to be taken literally? Is themessage to be taken in all its ambiguous glory? Did the initiate understand the message correctly? An advantage the initiate will use to resolve these questions is the source--the Zen master. As was described, the initiate inescapably comes to grips with the koan as intended uncertainty--ambiguity. But the initiate learns this as a fact only because interaction with the source is allowable within the structure of the koan tradition. Even though it is in an indirect manner, the initiate uses the master to distinguish what is ambiguous from what could be construed as equivocal. But what.occurs when the does not have the opportunity to interact with the source as the telegramreceiving spouse serves to illustrate? In the absence of communication transactions with the source orwhen these transactions are restricted to some degree, the elements of ambiguity in a message are difficult to distinguish or are altogether indistinguishable .from elements in the message that could have been damaged by equivocation. Any receiver will normally deal with uncertainties in different ways depending upon the degree to which the source of the message is available for interaction. 44


Recognizing thes.e differing circumstances, this thesis defines source-specific communication activities as those which permit a receiver to enroll the aid of the messagesource to distinguish ambiguous from equivocal uncertainties. Source-nonspecific communication activities are those in which the receiver must distinguish on his/her own the distinctions between ambiguous and equivocal uncertainties. Normally, this occurs in the absence of the source. From these definitions, the "quality" question, the interstellar communication example, and the "boar" telegram exemplify source-nonspecific communication activities. While no direct questions are actually asked of the Zen master in our scenario, the Zen initiate can still distinguish ambiguity from equivocation indirectly through the master's responses to the initiate's messages. The koan illustrates source-specific communication. married couple and the telegram was presented as a communication act, once the one spouse can somehow communicate with the telegramsender, then any resolution of the ambiguity/equivocation matter would make this example a source-specific one. Before closing this section, source-specific and source-nonspecific communication activities are not to 45


be confused with the proximity of the source in relation to the message receiver. In many face-to-face encounters, there are, of course, times when a source chooses purposefully to blur the distinctions between ambiguity and equivocation (Bateson, 1969; Eisenberg, 1984; Beavin Bavelas, Black, thovil, & 1990). A more detailed treatment of this matter will be reserved for the next section, but it is worth noting here that a message source just a step away from a receiver may in one instance be an active participant in source-specific communication transactions and in the next instance disengage altogether from clarifiing a message's ambiguous elements and conduct source-non-specific communication. The introduction of source-specific and source-nonspecific communication has led, inexorably, to the role that meanings, intended meanings, interpretations, and information processing have in communication. Having completed a survey of the rudiments of Shannon and Weaver's Mathematical Theory of Communication, the next section will examine how meanings and interpretations are derived from information processing in both interpersonal and organizational settings. Thereafter, meanings and interpretations will be tied to the handling of the dual forms of uncertainty: 46


i.e., ambiguity and equivocation. In particular, equivo-cation will be discussed as the fundamental concern in the processing of information and the derivation of meaning from messages. Throughout this next section of Chapter 2, a shift will occur in discussing the activities of communication. This shift will move us away from examining the activites and processes of from the perspective of the source to that of the receiver. The next section will alsd explore the distinctions between ambiguity and equivocation. In particular, the problems that a receiver faces in making these distinc-tions are discussed. Additionally, the relationships between the management of equivocation and the derivation of meaning from information will be explored. The Presence of Equivocation and Ambiguity in Three Forms of Miscommunication Shannon and Weaver's MTC was published within an engineering context. Their concern initially was to derive about the transmissions of all sorts of messages. Because this general concern encompassed a whole range of possible messages from digitalized energy schemes to complex linguistic utterances, the matter of establishing meaning from any message was not a primary 47


focus of attention. It may be said that Shannon and Weaver were more concerned about how a "good message" is formed and delivered than what happens to the message thereafter. On the other hand, a "good message" necessarily required Shannon and Weaver to discuss those critical variables that a source must take into account when composing and encoding a message for receipt at the receiver's site. Some of these responsibilities and considerations were listed earlier. Critical to what constitutes a "good message" is that it is recognizable as a message and that it is interpretable by the receiver. In the former case, a message is recognized as a message when a receiver recognizes that it contains information. In the latter case, a message is interpretable when sense can be made o the infoimation content. We have already said earlier that a communication transaction cannot take place unless a transmission from any source prompts a return transmission from a receiver. This transmission of response was defined as feedback. In addition to the actual content of this feedback transmission, the source may discern additional, indirect information about how the original message was processed by the receiver. Based upon this feedback, a source may broadly determine one of two possibilities: 1) the mes-. 48


sage was received properly and understood within acceptable limits or 2) the message was not received properly and understood within acceptable limits. In this second case where the feedback response is indicative of miscommunication, three possible conditions could be causing the problem. The first of these conditions concerns itself an improperly received message. Assuming that the source sent a clear message and assuming that the receiver intended the message to be understood (that is, the receiver was not intentionally trying to be "vague"), an improperly received message is probably a message that was corrupted during its transmission. It will be recalled that Shannon and Weaver referred to this corruption as equivocation. This corruption may be due to the-intrusion of noise as any classroom teacher would attest to when trying to give an assignment to a class during a noisy activity or during an announcement being given over an intercom. Inattentiveness of the receiver could also result in an improperly received message. The between the transmitted message and the received message is the measure of the amount of damage to the message during its transit. Another way to look 49


at the measure of this damage or corruption or equivocation has on the message is the amount of information the source must retransmit in order to restore or repair the damaged message. For the source, this amount of equivocation is relatively easy to isolate. One common technique for measuring this amount is to ask the receiver to summarize or state back to the source what was just transmitted--that is, confirm the content of the transmitted message. In the second condition of miscommunication, a message may be properly received but may be misunderstood by the receiver. This second condition contains problems associated with interpretations and meanings. For any source, miscommunication due to interpretation and meaning can be arbitrarily isolated with relative ease. If the feedback response outside the range of responses although the source is through feedback that the message was properly received, then the difference between what was intended to be understood and what was actually understood is the amount of confusion that must be removed. It be understood that this condition does not refer to any damage to the message itself but refers to an interpretation or meaning problem. Because this is a matter of interpretation, the MTC states that any source 50


has an array of available code options to encode a particular message. If one code proves inadequate, then a secondary coding scheme may be selected and used. This condition is, of course, commonplace. In fact, it is common enough that we as humans often "rephrase" a transmission or encode an example of a transmission without waiting for a feedback cue from the receiver. We do this under the assumption that if we send the original message along with a replica Of the original message in a different code, we are more likely to minimize any confusions of meaning that could possibly occur with the transmission of the original message by itself. Shannon and Weaver refer to this strategy of retransmitting a message as a form of redundancy. This strategy for precise and exact retransmission of the original message is critical when high levels of equivocation may corrupt an important message in transmission. Transmissions from space probes are particularly susceptible to corruption as they are sent over millions of miles df space. To protect invaluable and irreplaceable information, these transmissions are encoded into coding schemes that are heavy with precise redundancies. Another form of redundancy is to encode a meaning into two distinctly different codes. In this manner, 51


a receiver is given two opportunities to understand a meaning. If the first meaning should not "make any sense," then a second different message with a redundant meaning may sent that does "make sense." Or, if both meanings are somewhat but not completely understood, then a comparison of the two meanings should yield a composite meaning derived from where the two separate meanings overlap. In this strategy, meaning is understood as a product of a sort of 11binocular"comparison (Bateson, 1979). Redundancy should be understood by now as an important means of damage to equivocation or to ambiguity. In fact, in terms of redundancy alone, it couldbe said the amount of redundancy necessary to repair a -damaged message isprecisely equal to the amount of equivocation present in originally received message. Also, the number of redundant encodings necessary to repair a misunderstood message is precisely equal to the amount of ambiguity present in the original message. The third condition in which miscommunication occurs is the transmission of a message that is subject to damage by both equivocation and ambiguity (i.e., corruption during transmission and .misunderstanding of meaning). This third condition is a type of "worse case" scenario in which the first and second conditions are present simultaneously. 52


It will be recognized that this third condition is the more complicated of the three miscommunication conditions. The primary reason for this complexity is that repairing the confusion in the message requires repairing damage first to the message itself and then to the interpretation. It will also be recognized that repairing the damage to a message really means removing the equivocation that has intruded into the message before interpretation of it. The process of removing equivocation essentially unveils the message's information content shrouded previously by the intrusion of .the equivocation. Precise retransmission--or purely_redundant message retransmission--was shown to accomplish this removal of equivocation. Repairing the damage due to meanings requires a different approach. Removing confusion based upon ambiguity requires"encoding tbe message into different coding schemes. For highly ambiguous messages, clear understandings may require multiple encodings--multiple applications of the encoding process. These differing processes of equivocation and ambiguity removal may not be substituted for one another. If at the root of a conflict between spouses is a basic misunderstanding of meaning due to ambiguity, no amount 53


of repetition of each one.'s position will result in clearer meaning. While one or the other spouse may give in out of exhaustion and feign understanding, the result can never be a truly clearer understanding of the conflict's root. This is because equivocation removal processes are being used on what is actually damage caused by ambiguity. It is likely that the spouses will continue to visit the conflict over and over again as the ambiguity at the root of the misunderstood meaning can never removed by processes that remove equivocation. Similarly, unclear messages are not cleared up by multiple descriptions. This error is frequently made by communiGators who follow a message immediately with a newly encoded second transmission. While the source may be trying to minimize misunderstanding due to unclear meanings, the fact is that if the message itself was not clear the first time, it will not necessarily be any clearer in a second message either. Likewise, an attorney presenting a highly complex case with precedent-setting implications must ensure that the presentation strategy of the case includes heavily redundant retransmissions of each crucial step in the evolution of the case. Over the long haul, if a strategy of redundancies is not employed, the attorney's main mes-54


sage to the jury members may never be conveyed. This strategy is of even greater importance to the attorney's presentation if another part of the attorney's strategy is to employ ambiguity that clouds his/her counterpart's message. So far, the discussion of equivocation and ambiguity has purposefully kept intact a clear distinction between the two forms of uncertainty. This distinction is also reasonably clear to the message source who, it is assumed, can to some degree distinguish whether there exists a communication problem because the message was unclear, or because the message was not understood, or because of both. Now the question before us in this thesis does not concern itself with how a source repairs his or her damaged messages. Rather the question before us was positioned earlier as a source-nonspecific question: i.e., independently of the presence or absence of any message's source, how does a receivei go about making sense of highly equivocal or highly ambiguous messages? Equivocation and Ambiguity from Perspective The distinctions between equivocation and ambiguity are not so clear to a receiver as they may be to the mes-55


sage source. This lack of clarity contributes to many of the difficulties a receiver must overcome if information from critical messages is to be processed accurately. While it is tempting for a receiver to ignore the complexities involved in unraveling equivocal and ambiguous message elements, improved levels of information processing and understanding can be achieved by a receiver who can: 1) separate the two forms of uncertainty from one another and perform the appropriate processes on each form that removes it from the message, or 2) perform a more generalized process that incorporates two subprocesses: a) one that isolates and removes equivocation from a.message and b) one that isolates ambiguity from a message. A description of this latter process is the goal of this larger of Chapter 2. Before the actual descr.iption, it will be useful to sample a couple of applications that such a process must address if it is to be useful to a receiver who wishes to remove any variable amounts of equivocation and ambiguity from messages. 56


In a most general sense, the process must address how our spouse ought to go about removing the uncertainty in the "I just bought a boar" message. It must be capable of: 1) resolving whether the message is equivocal--i.e., damaged by human mistake, or corrupted during transmission, or misdelivered, or any of a number of errors and 2) making sense of the strangeness, or unexpectedness, or uniqueness of the message's content. We will call these kinds of equivocations and ambiguities straightforward uncertainties. At the heart of this type of uhcertainty, the receiver assumes that the delivered a message which is not intended to deceive the receiver by masking another or secondary meaning, disguising the truthful content of the message in order -to secure an undisclosed advantage, perjuring or presenting false information as though it were true, or any of a number of other possible deceits or tricks which can be facilitated by communication. Straightforward uncertainty, then, is assumed in ciicum stances where high levels of trust, faith, and credibility are established. 57


Nonstraightforward Uncertainty from a Perspective As early as 1956, the Palo Alto Group began investi-gations into.the types of communication that "didn't make sense." The initial study in this body of work focussed on the communication patterns of schizophrenics (Bateson, Jackson, Haley, & Weakland, 1956) whose messages at times could be exceedingly unpredictable, disorganized, and seemingly confused. The Palo Alto group was able to show that the schizo-phrenic's superficially "off-the-wall" messages made more sense when analyzed at the point in Shannon and Weaver's MTC where encoding occurs. It will be recalled that Shan-non and Weaver portrayed encoding is a process of choices. An individual selected a single code from an array of possible codes. Shannon and Weaver's original discussion considered code selection to be centered around the source choosing the ideally right code that reduces the equivo-cation in a given message to a tolerable level for the receiver. In contrast, the Palo Alto Group discovered that code selection by schizophrenics was frequently cen-tered around the concern for choosing the right code that minimized the undesirable consequences perceived by the 58


source to exist at the time of and due to the content of a message. Of course, the research in this direction has broadened beyond schizophrenic messages and has been generalized over many commonplace communication exchanges. Accordingly, any source may choose a code that intentionally equivocates the content of the message precisely because any less equivocal code is perceLved by the source to result in potentially negative consequences. In short, the purpose of equivocation may be avoidance of feedback (Watzlawick, Beavin, & Jackson, 1967, pp. 72-80). Seen this way, sources make encoding choices based upon certain predispositions, affectations, biases, or other selection criteria other than those choices based upon achieving ideally unequivocal, clear messages. For a receiver, this means that equivocation can come in at least two varieties: 1) One type of message may be unintentionally high in equivocation such as any "garbled" or "error-ridden" message sent by a maladept source, or 2) As the Palo Alto Group suggests, the message may appear highly equivocal because the source intentionally chose to make the message highly equivocal. Because these latter kinds of messages "appear ambiguous, 59


contradictory, tangential, obscure, or even evasive" (Beavin Bavelas, Black, Chovil, & Mullett, 1990, p. 28), they are referred to as nonstraightforward messages. Since the intent of a nonstraightforward message is directly dependent upon the sociological and psychological demands of the environment in which the source prepares or encodes the message, "equivocation is the result of the individual's communicative situation" (Beavin, Black, Chovil, & Mullett, 1990, p. 54). Embedded within this notion is a causal relationship of importance to the receiver: The presence of a nonstraightforward message necessarily reveals the source.'s perception that a less equivocal message cannot be safely sent. There are two benefits to a receiver for recognizing and removing the equivocation in nonstraightforward messages: 1) The straightforward message disguised behind the nonstraightforward message may be discerned. 2) The perceived need for an equivocal code selection becomes clearer. It should be evident that this direction of research lends itself directly to analyses of important sociological and psychological situations based the upon study of the nonstraightforward communication present. 60


Let us now rephrase the goal of this larger chapter section. What must an information processing system consist of that would allow a receiver to: 1) identify and remove the equivocation in a message regardless of whether it is a straightforward or nonstraightforward message and 2) identify and interpret the ambiguity in the resultant message (including the actual "message" lurking behind. the nonstraightforward message) in order to understand the meaning of the message? From a receiver's standpoint, a process for the removal of equivocation and ambiguity in nonstraightforward messages would have a variety of useful applications. For instance, work by the Palo Alto group and their successors (cf.,-weblin, 1962; Sluzki, Beavin, Tarnopolsky, & Veron, 1967; Watzlawick, Beavin, & Jackson, 1967) have already discerned that the equivocation in a nonstraightforward message can be tied intimately to the relative amouQts or degrees of negative consequences perceived by the source to be present in a given interpersonal communication transaction. For all intents, this type of communication transaction bears all the qualities of interpersonal conflict. In conjunction with this instance of conflict, the notion of purposefully chosen equivocation 61


has been found to be particularly amenable to conconceptual analysis within the larger context of Kurt Lewin's Conflict Theory (Lewin, 1938; cf., Bavelas, 1978, Chapter 17). Heuristically, what all this points to is another means by which interpersonal conflict be analyzed through the presence of nonstraightforward messages. More importantly a process for removing this purposefully imposed equivocation could greatly aid a receiver in determining some of the informational and perceptual roots behind the conflict--roots that are left unspoken but that are critical to effective management of conflict. Ambiguity and the Equivocation Fallacy Having extended the dimensions of our consideration of equivocation to include straightforward and nonstraightforward forms, it will be useful to extend the dimensions of our consideration of ambiguity. It will be recalled that ambiguity was defined earlier as desirable uncertainty intentionally encoded into the message by the source. Moreover, ambiguity serves to enhance or deepen a message's meaning when used properly. Ambiguity, then, is conceptually opposite that of equivocation, which serves to obscure or disguise a message's meaning. The Zen koan example illustrates the power of ambiguity. 62


Ambiguity may be characterized as a by which a source communicates to a receiver an idea that contains multiple meanings all or which are to be interpreted simul-taneously. Properly used ambiguity extends meaning while maintaining an economically coded message-package. The oxymoron (i.e., "jumbo shrimp") and pun are familiar exam-ples of ambiguity. Consider the following: Webster says that taut means tight. Well, I guess I was taut quite a bit while in coliege (Shulman, 1955, p. 155). At least three meanings are nested within this pun. The first of these centers around the synonym "taut" = "tight" as in a "taut" rope = a "tight" rope. The second meaning centers around the euphemistic definition that "tight" = "very drunk." The third meaning centers around the homonym relationship "taut" = "taught." These relationships can be.expressed in the form of a syllogism: 1) If "taut" = "tight" 2) And if "tight" = "very drunk" 3) And if "taut" = "taught", = "tight" = "very drunk" 4) Then [simultaneously] "taut" = "taught" Obviously, the cleverness of the pun is lost in the analy-sis. In fact, the pun's cleverness precisely consists in the economical packaging of the multiple meanings. 63


The point here is that the degree of cleverness is exactly related to the degree of the ambiguity in the pun. Ambiguity enhances meaning as long as the receiver recognizes the ambiguity for what it is. So long as the receiver is cued into or assumes the presence of ambiguity, then the receiver looks first to identify and then process the message's multiple meanings. As stated earlier, the focus of this subsection is to extend our.understanding of ambiguity. So with this in mind, consider the following: If you take two important things out of a basketball game, you will always get a What two things are they? This, of course, we recognize as a riddle or puzzle. Solving the puzzle requires the receiver to distinguish between two contexts in which the words may be interpreted. If a receiver tries to interpret the riddlewithin the context of the concepts the wordi convey, the puzzle is impossible to solve. That is, the receiver tries to ima-gine what rule or equipment or personnel changes would be necessary for an indoor game to be played outdoors, a game with baskets to be played with bats, a game with five players and two referees to be played with nine players and four umpires and so forth. 64


On the other hand, if a receiver interprets the riddle within the context of the words as merely words (i.e., arranged patterns of letters), the solution pops right out: you take the "k" and the "t" out of "a basketball game" and what remains isalways "a baseball game." The puzzle is more difficult when it is spoken to the receiver rather than written out. This tends to more subtly disguise the word-as-word context which helps to unmask the solution and tends to stress the conceptual context in the receiver's mind. The riddle is made still more difficult when the source verbally poses the puzzle to the receiver .while both are watching a basketball game. This difficulty arises because the context of "bas ketball game11 as a conceptual event has been established rigorously in the receiver's perceptions through viewing the game in action. In the social cognition literature, this activation of a specific context of interpretation is called "priming" (Fiske & Taylor, 1984). While priming may be employed to enhance the interpretation of a message, it is important to note here that a source can enhance the disguise of the puzzle's actual solution by drawing attention away from a context of correct interpretation by stressing attention to an erroneous context. This illustration exemplifies what appears to be the purposeful 65


misuse of context to disguise or shift a message's meaning. This same trick can be performed at a variety of levels of interpretation. Just as contexts of interpretation may be disguised, ideas and even individual words may undergo a disguised of meaning. For a receiver, these shifts of meaning can be exceedingly troublesome to discern. They can become troublesome when a receiver comes across a message that seems impossible to interpret and when there is no .source who can be relied upon to provide a hint or clue as to how the message is to be interpreted. Although at first these shifts in meaning seem likeexamples of ambiguity in messages, they are not. Rather than being meanings that are intended to be interpreted simultaneously (ambiguity), these are true meaning shifts: two or more .meanings may be at play simultaneously, but at least one of them is disguising the application of the more appropriate meaning. This is more truly an example of equivocation intruding upon the realm of meaning than ambiguity. In argumentation, this application of meaning shift to logic is called an equivocation fallacy and is listed among Aristotle's Enthymemes in the classical literature. 66


Subtle shifts in meaning are the most difficult to detect although any shift strictly invalidates an argumentative case. Obvious shifts may be resolved easily by a receiver and may be considered akin to "catching someone in a contradiction." Extremely subtle shifts are extremely difficult to resolve and probably rest at the heart of paradoxes. We would expect that removing the equivocation from these subtle shifts should cause the paradoxes to dissolve. Kenneth Ferguson illustrates how this process works in his "Equivocation in the Surprise Exam Paradox" (1991 ). The Surprise Exam Paradox may be summarized as follows: given that a surprise exam must necessarily be taken by students within some finite time frame (such as a quarter-, semester, or year) and given that a surprise exam must necessarily mean that the testing date cannot be predicted beforehand by students, then a true surprise exam is an impossibility. The description behind the paradox begins with recognition that students cannot possibly be given a surprise exam on the final day of class. This is so because if the surprise exam had not been given through the next to the last day of the class, then every student in the class must be able to predict that the exam has to be 67


given on the final day of the class. Obviously, the test would come as no surprise to anyone. As a consequence, the final day of the class may be ruled out as an option for giving a surprise exam. Knowing that the next t6 the last day bf class is now the last opportunity for giving a surprise exam, students will be able to predict with certainty any surprise exam scheduled for day using the same logic used before. Employing the same repetitively, each day of the class may be and the proof of the paradox is completed. Since its introduction (O'Connor, 1948), scrutiny of the paradox has not revealed any flaw in the fundamental presentation of the Had there been a flaw here, the paradox would have beeri caused by equivocation as described in earlier sections of this thesis (i.e., an inadvertent error in the presentation of the logic). After ruling out a mistake in the logic, Ferguson has revealed that the paradox consists in .a shift of meaning in the use of the term "surprise." This shift in meaning categorizes the paradox as the product of an equivocation fallacy. Detailing the steps Ferguson follows to debunk the paradox is unimportant to our discussion. However, sum-68


marizing the strategy he uses to attack the paradox is crucial. This strategy contains three distinct steps: 1) The set of conditions necessary to produce the paradox are listed rc,J 2) The set of conditions that need to be in place in order to deliver a true "surprise" exam are listed [C2] 3) An element-to-element comparison of [C 1 ] and [C2 ] is conducted. Any discrepancies in this comparison are expected to rise to the paradox. Within the execution of this strategy, a shift emerged in the use of the word "surprise." All other conditions remained unchanged. The shift in meaning discovered by Ferguson is very subtle, and it is worth mentioning here that where meanings are concerned, the degree of subtlety in shift of meaning is inversely proportional to the amount of equivocation that needs to be removed. Miniscule shifts characterize huge amounts of equivocation. Perhaps it is already that what we have called Ferguson's "strategy" is really a competent application of Shannon and Weaver's equation for measuring the amount of equivocation in a single transmission on a level of meaning and interpretation. Whereas the MTC defined equi-69


vocation as the difference between the of the message sent and the contents of the message that is received, Ferguson has shown us that at a more abstract level, the equivocation in paradoxical conditions may be defined as the difference between the current conditions of paradox and the desired conditions free of paradox. Armed with a broader understanding of equivocation and a clearer distinction between equivocation and ambi-guity, we may now turn our considerations to the processes that remove equivocation from messages. These processes make up the basic structure of the recursive inquiry The Processes of Establishing Meaning: Karl Weick's Social Psychology of Organizing The task before us now is to summarize an informa-tion-processing means by which a receiver may remove equi-vocation from sites in a message in order to isolate the information content and cull meaning from that content. As has been shown, equivocation may be present due to unintended human error or may even be present as a strategy by a source to secure a purposeful advantage. This advantage may be simply that of self-preservation against what are perceived to be possible negative conse-70


quences or, more malevolently, that of deceiving a receiver from discerning a more "truthful" message. It has also been shown that equivocation may be located within the code of the message itself, within the meaning of individual concepts (e.g., words or phrases) in the message, or within larger concepts and contexts formed by the message. In this connection, equivocation can mask the information-bearing elements of a message, the ambiguity within the message, and the contexts and meanings of the message. In his book The Social Psychology of Organizing (1969, 1979), Karl Weick has posited an interlocking system of processes which together form a means for identifying and then removing equivocation from messages, or what he calls "informational inputs." A summary of this interlocking system is the subject of this subsection. At the same time, the context of Weick's system of processes--that of the organization--will be summarized and then woven into the subject of this thesis--the recursive inquiry. In both the original and revised editions of Weick's book, the treatment of equivocation springs from the development of the more central concern of "what is 71


an organization?" In addressing this more primary ques-tion, Weick endeavored to fashion a model of general application that captured the "living and dynamic" attri-butes of what is commonly known as an "organization," or in his words the "social collective." This model was developed in contrast to the more venerable models of organizations that tended to focus on the static, unchanging, and more permanent attributes of organizations. In contrast to this thinking which tended to stress organizational structures (e.g., the seemingly omnipresent "organizational chart"), Weick's model recognizes that organizations consist simulta-neously in two dimensions: activity (function) and the structure (form) necessary td support the activity. Of course, Weick is rejecting a structuralist's perspec-tive that stresses organization of the activity first. This structuralist approach, he argues, restricts what the organization can do--the organiza-tion's activities. In contrast, Weick champions the notion that the organization (structure) ought to emerge as a consequence of organizing (activity): there are processes which create, maintain, and dissolve coilectivities, that these processes constitute the work of organizing, and that the ways in which these processes are continuously executed ARE the organizatioh (1969, p. 1). 72


Weick employs three premises to derive his system of processes. The first of these premises is: "organi-zation" is 11that which emerges out of the process of organizing'' (1969, p. 3). The other two premises and subsequent construction of Weick's model come out of a synthesis of three families of theory. These are: 1) George H. Mead (1956) [Social Interactionism] 2) Georg Simmel (1950) [Kantian German Idealism] 3) w. Ross Ashby (1956) [Cybernetics artd Systems Theory] The second premise concerns itself with the purpose or intent behind these organizing processes (i.e., "What do these 'organizing processes' accomplish?"). As George H. Mead noted, general social activity-consists of indi-vidual acts and An act is an impulse that maintains the life process by the selection. of certain sorts of stimuli it needs. Thus, the organism creates its environment. The stimulus is the occasion for the expression of the impulse (Mead, 1956, p. 120). Combining this with Georg Simmel's notion (1950) that any individual's "attention" is the "selection faculty" by which relevant stimuli are chosen, Weick concludes that "man notices those stimuli which permit him to do what he wants to do" (1969, p. 26). The second premise in its entirety says that the organizing pro-73


cesses, or behavioral activity, in which organizations engage exist for the purpose of first creating and then enacting the environment that defines the limits of the organization. In Weick's own words, [organizations] create and constitute the environment to which they react; the environment is put there by the actors within the organization and by no one else the environment is a phenomenon tied to of attention, and that unless something is attended to it doesn't exist (1969, p. 28). The substance of any environment, then, is comprised of what has been learned about the environment, how this learning is responded to, and what new learning is being attended to. In this view, the environment is depicted as a fuzzy phenomenon having clarity only in proportion to what is known about it. The fuzziness or clarity as well as the extent or dimensions of the environment depend exclusively ori three factors: 1) what is already known, 2) what is being known, and 3) what is not known but could become known. The final premises addresses the questions: "How is an environment created, and how is an environment enacted?" Consistent with Simmel and Mead, Weick con-ceives of the "known" environment as comprised entirely 74


of information and all those meanings derived from that information. This view is also consistent with the findings of Katz and Kahn (1966). Part of the purpose of organizing activities, then, is to respond to that which is "known" about the environment. However, the much larger part of organizing activities serve to expand and refine what "can be known" about the environment and respond to these newer meanings. For all intents and purposes, any environment is an amorphous abstraction composed of infinite amounts of information. The "known" environment is that information which through one's attention has been resolved into meaning. Conversely, the "unknown" environment is that significantly larger body of information which remains outside 6f one's attention and, as a consequence, cannot yet be given any meaning. This leads to an intriguing question: can an environment be known in its entirety or, at least, known in a substantially larger way? At least two obstacles limit our expanded understanding of the known environment: 1) The first obstacle is imposed by the boundaries formed by the imperfections and undisciplined hab-75


its of our human attentions--short attention spans and easily distracted attentiveness. 2) The second obstacle is imposed by the intrusion of equivocation. As .has been stated before, this intrusion obscures information, and since the "stuff11 of the unknown environment consists in information, equivocation pbscures that environment. We may now say that organization emerges as a consequence of organizing activities. These organizing activities are concerned with the creation (establishment of meaning) and enactment (response to meanirig) of the environment. Subverting these goals are unfocussed attention and equivocation. So within the construction of those processes that compose the activities of organizing, provisions need to be made for both controls over attention and the removal of equivocation. Weick identifies the activities of organizing as falling into an information processing system. As Weick himself poirits out, this general system is a cybernetic system that is composed of three subsystems which are simultaneously interlocked and interdependent. These three subsystems are named: 1) variation, 2) selection, and 3) retention. These names and concepts are borrowed directly from Donald T. Campbell's Evolutionary Model 76


(1965). This borrowing is direct evidence of Weick's understanding that biological and sociological adaptations are intimately related to one another, and both reflect the kinds of activities humans are engaged in when they interact with or "enact" their surrounding environments. Each these three processes will be introduced in general After its general description, each process will be tied to the example of the "quality question" discussed in the Preface to this thesis. By doing this, it is hoped that two purposes will be accomplished: 1) A practical example will serve to illustrate the interrelatedness of each of the subsystems and 2) Weick's 3-component information processing system will be shown an effectivemodel for describing how the recursive inquiry works. The Variation subsystem is understood to be synonymous with what may be called the Enactment subsystem. Weick uses the terms "variation" and "enactment" interchangeably. The kinds of activities that comprise the Variation/Enactment subsystem are those that determine how the members of the organization "choose" to interact with the generalized environment. 77


Variation embodies the notion of "diversity." The greater the diversity of interactions with the environment, the richer will be the return to the receiver in terms of useful and relevant bodies of information and their subsequent meanings about the generalized environment. If the environment is viewed as a nonspecific source of an immeasurably large number of and variety of messages, then the Variation/Enactment subsystem consists in those processes that seek out and attend to certain specified messages. This means, of course, that as a result of a selection process, the myriads of messages that are not attended to are essentially ignored. Like gravity before Newton's time, these messages are there, but because are unattended to, they are for all intents and purposes nonexistent. The Variation/Enactment subsystem may be likened to a filter and an aperfure to the system itself. Because of the substance of the filter, certain messages from the environment-source will be ailowed through. Because of the size of the aperture, only certain amounts of information from the environmentsource can be attended to. Together, this "filter/ aperture11 subsystem will determine what sort of and 78


what amount of information will be processed by the larger system. It is expected that what constitutes the filters and the size of the aperture will include many of the biases, prejudices, assumptions about the "real world" -environment, and weltanschauung of the receiver. The more restrictive these values and biases are on the general approach to information gathering, the more restrictive the information flow will be. One advan-tage to restricting the information flow is that sig, nificantly less equivocation obstructs the processing of information. For receivers, a restricted flow of information tends to "confirm" or "reinforce" _rather simplified models of the surrounding environment. Conversely, the more one opens the flow of infor-mation, the more equivocation obstructs the processing of information. Presumably the more adept a receiver becomes at managing the processing of equivocation, the more skilled the receiver becomes at perceiving and understanding the surrounding environment. There is a trap, however. A receiver may conceive him/herself as one sensi-tive to a vast range of information gathering possibi-lities. That is, he or she may want to be as open to 79


information messages as possible. However, wanting to be an open receiver and actually being an open receiver are two entirely different matters. As will be discussed later, an unrestrictive filter/aperture system is determined by the number of restrictions put in place by the receiver. For instance, a receiver that desires to .view the environment as openly and as unbiased as possible is actually placing this desire as a restriction on the Variation subsystem. By emplacing this restriction, the receiver is, in actuality, filtering out messages and that may contradict the restriction. In short, the receiver is confirming his or her "open and unbiased" interaction with the environment by obstructing messages transmitted from the environment-source that contain information portraying the receiver as "closed and biased:,." Within the quality-question example described in the Preface, the Variation subsystem is seen as the formulation of the quality question itself and the ground rules and restrictions placed upon the information gathering. Evidence suggests that the Variation subsystem was unusually "open" to a wide range of information inputs. While this meant that a substantially large view of the environment was being taken, equally large 80


amounts of equivocation were being taken in too. It is perhaps possible that the question originally called a recursive question contributed heavily to opening up this conceptual filter/aperture system. Had there been more restrictions placed upon the gathering of information on quality by the organization, it would be expected that an equally smaller view of the environment would have occurred and the impact on the organization of this information processing would have been diminished. We have described the Variation subsystem but have not yet discussed how equivocation is treated within this subsystem. While more will be said after the general system overview, the Variation subsystem must remove at least some of the equivocation in any message before passing it along to the next subsystem. The equivocation addressed by this system is the sort damages the encoded In other words, equivocation in the purest of Shannon and Weaver's definition disguises the presence of relevant information to a receiver. It can cause a receiver to mistake an information-rich transmission for an information-deficient burst of noise. Therefore, the Variation subsystem must remove the kinds of equivocation intrusions that may cause a message 81


to be mistaken for an irrelevant input and to be ignored when, in fact, the message may be heavily laden with relevant information and should be selected for further processing. Like Variation/Enactment processes, Selection processes are similar to a set of filters through which information inputs from the Variation subsystem are sifted so that meanings can be assigned to the message contents according to some relevant set of criteria. Through selection processes, validity, relevance, and usefulness values for these meanings are determined. Through restrictions set up within the Selection subsystem, certain information inputs may be scrapped from any further processing. Other information inputs may be assigned meanings and then passed along for further processing. The criteria used by a receiver to establish Selec tion subsystem restrictions are normally quite fluid. For instance, information scrapped at one time may be deemed highly relevant at another time when Selection criteria change. Weick divides Selection criteria into two types: 1) those criteria relevant to the internal welfare of the organization (or receiver) and 82


2) those criteria relevant to the external welfare of the organization (or receiver). Weick defined "internal welfare" as those modifications necessary to control the organization's to perceived change in the external environment. By contrast, "external welfare" refers to the modifications or changes made upon the environment through interaction with the environment by the organization. Broadly, it may be said that restrictions in the Selection subsystem that are perceived to enhance the internal welfare of the organization enhance the organization's ability to adapt to external and, to some degree, unexpected internal changes. Restricti6ns that are perceived to enhance external welfare enhance the organization's ability to create change--to empower the organization to make things happen. Weick notes that the healthy organization seeks to establish an overall information processing strategy that balances both internal and then external matters over time. It is, of course, counterproductive to develop a Selection subsystem that incorporates internal and external restrictions simultaneously. As noted in the Preface, the information being gathered to address the quality_ service question was 83


interpreted first along lines of establishing stronger external ties with the organization's clientele--i.e., understanding how new levels of quality in service could benefit and impact the organization's external world. It soon became apparent that to enact this new knowledge among the clientele, the organization had to address internal This new recognition led to reinterpreting the knowledge about quality in service to the internal day-to-day activities of the organization. As the irtformation was viewed in this new light, some crucial insights were gained and enacted vithin the organization that dramatically improved inter-and intra-departmental performances. This cyclical swing from external to internal Selection processes was an on-going affair throughout the most dynamic times of our organizational change. It is. also safe to say that had the organization NOT gone through the patterns of swings, the organization would have reached a pathological stage endangering its very existence. On the one hand, fixation exclusively on the Selec tion of meanings based upon external welfare would have led the organization to interactions with the clientele that simply could not be supported by the remainder 84


of the organization. It would be an organization of promises and no follow-up. It would be a dreamy, fictitious organization being portrayed to its clientele as it "could be" rather than as it "truly is." On the other hand, fixation upon internal welfare yields an organization that runs like clockwork despite its ponderous administrative size. Usually the clockwork breaks down, however, when the organization interacts with its necessary evil--the customer. Internally fixated organizations frequently adopt the tragic motto: "business would be great if we just didn't have to deal with customers." If the primary purpose of this Selection subsystem is the establishment of meanings, this purpose is accomplished by the removal of equivocation embedded at the level of meanings. The Retention subsystem is the .last of the three organizing processes. Weick does not intend for the term retention to be confused with we might call a "cere bral archive" nor a ''storage bin of policies and procedures." Instead, Retention processes forge and fashion the frequently elaborate ways and means by which the organization responds and adapts to change. 85


In this the Retentiori subsystem guards against undesirable or contradictory intrusions of meaning that may disturb, endanger, or threaten to destroy the organization. An example of this was related above. Sometimes certain informational inputs from the are ignored as irrelevant to the information gathering as defined by the restrictions imposed upon all inputs. inf6rmational may be discarded before entry to the entire system a:t the Variation/Enactment stibsystem or, enduring these restrictions, may be discarded during the Selection subsystem. At another time, any of these informational inputs may be regarded as relevant to the Oftentimes precisely for the they.were discarded, these messages are reinserted into the information processing system. The Retention subsystem simply refers to a type of "self-cleansing" system which periodically reviews the contents of all selections retained by the Retention process and resolves any internal inconsistencies that will inevitably result over the course of time. The Retention subsystem re-steadies the course of the organization when evidence suggests it is off course. 86


Of these processes, Weick notes, [they] affect subsequent actions; they are frequently edited; they are protected .in elaborate ways that may conflict with and selection; they are coercive only to the degree that members are informed of their contents; and they contain items that frequently are opposed to the selfinterest of the persons who must implement these items (1969, p. 58). Embedded within the a6tual Retention subsystem are two smaller subsystem loops which together are essential to the control of the larger subsystem processes: 1) the Retentiori-Reorganization subsystem and 2) the Choice Point subsystem. Of particular interest to us is the Choice Point subsystem which is located at the very end of Weick's three-component information processing model. Essen-tially, the'Ch6ice Point is the final review of any information input. This final .subsystem screens for one of two choices: 1) a "noticing choice" or 2) a "doing choice" (1969, p. 59). The Noticing Choice selection both recognizes that additional amounts of equivocation remain in the message which need to be removed and that the message needs to be resubmitted to the larger organizing process 87


at an appropriate point where the remaining equivocation may be removed. The Doing Choice simply enacts or action based upon the newly emergent meaning by: 1) incorporating the new meaning into the already existing general corpus o.f meanings--the "library" of shared meanings within the organization's membership and 2) responding by either adapting internally or altering the external environment to the consequential changes imposed by the new meaning on subsequent organizing activity. The-Choice Point that all-important juncture within the entire information processing system where a message's information has become meaning and where meaning has gi ven wayto change. When either exterrial or internal emerges 6ut an important kind of feedback occurs. As Norbert Wiener points out, feedback is "the property of to adjust future conduct by past performance" (1967, p. 47). Without this capacity, past experiences contain no relevant value beyond the timespace they fill. Without this capacity, memory becomes impossible. Without this capacity, what Wiener 88

PAGE 100 as "higher order feedback" (1967, p. 48) becomes equally impossible. According to Wiener, higher order feedbacks regulate whole policies of behavior. When these policies of behavior are attendant upon the acquisition and processing of information and meaning, learning itself occurs. Since learning is directly related to 6ur investigation into the recursive inquiry and since learning is critically dependent upon higher orders of feedback, we may expect that the Choice Point--Weick's site for feedback--is critical to the descriptive power of his model for our purposes. concerns this subsystem require the removal of. those uncertainties centered. around the comparison of riew meaningswith.currently extant meanings. These of equivocatlon matters were exemplified earlier in the consideration of the presence of equivocation in puns, paradoxesr and contradidtions. While these types of equivocations may be the substance of some Retention processing, the Retention subsystem deals with a host of other equivocation types. In general, the purpose of the Retention subsystem is the development of those meanings that emerge as a result of relationship between lower orders of 89

PAGE 101

meanings. In the quality service example, this was characterized as the processes the organization engaged in when new meanings of quality _and service were added to what was "already known" about quality and service. This is not accomplished by adding new meanings to the total number of old meanings. To combine new meanings with extant meanings, the organization had to develop a descriptive meaning that tied together the relati6riship new meanings and the meanings. Developing this descriptive meaning frequently required the removal of large. amounts of equivocation. This was particularly true when a comparison of new and extant meanings did not reveal an superficial relationghip. Charles (1958) provides an irtteresting example of how difficult it can be to justify extant meanings together by a more generalized, composite meaning. In 11The Nature arid Functions of Official Norality," Warriner (1958) studied a small rural Kansas community's attitudes toward drinking. On the surface, the residents seemed to be holding to a hypocritical, contradictory moral code. On the one hand, residents loudly and profusely decried public displays of drunkenness. Perpetrators of public drunkenness were ridiculed, demeaned, 90

PAGE 102

and socially rejected. On the other hand, there was widespread drinking and drunkenness conducted in nearly all of the residents' private homes. Of particular interest to Warriner's study was his recognition that these seemingly inconsistent attitudes were not perceived as in-consistent by the residents themselves. In fact, they perceived the two attitudes as perfectly compatible although they could not articulate what moral code aligned the attitudes. Warriner sought to describe what moral code needed to be in place in the residents' minds to justify the consistency between these contrary attitudes. Warriner's task characterizes the problems the Retention subsystem must overcome in order to incorporate meanings into the larger body of extant meanings. Equivocation plays a direct role in these problems: The more obscured the relationship between meanings is, the more equivocatioh is present. If this equivocation can be removed, then the new meaning is tied into the extant corpus of meanings and a new, more generalized, descriptive meaning replaces the previous, now obsolete and less-generalized descriptive meaning. If this equivocation cannot be removed from the input, then the new meaning may exist separ-91

PAGE 103

ately, or extant meanings may be discarded to accommodate the new meaning, or the new meaning may be discarded altogether. Regardless of the change, the extant body of meanings must necessarily change. One measure of the extent to which change has occurred-is the change which occurs in the internal functions and form of the organization or in the external interactions with the environment. In this latter sense, the presence of change in the way the organization interacts with the environment will change the environment. This is referred to by Weick as Enactment. We have now come full circle. Enacting change into the environment causes the engagement .of the Variation/Enactment subsystem arid the begins all over again without interruption. In any dynamically learning organization, the information processing system is cyclical--that is, recursive. New meanings about the environment necessarily broaden the understanding of the environment and also become inc6rporated into future interactions with the environmerit. In turn, these future interactions yield still newer meanings which, through still additional recursive cycles, begins to describe how a recursive inquiry may modeled. 92

PAGE 104

Two questions beg to be asked. The first one concerns the recursive attribute of Weick's organizing system. Our tells us that humans are not incessant, greedily hungry information processors. The model suggests uninterrupted, recursive information processing. How, then, does a receiver stop the recursive cybernetic system? The system stops processingmessages when the information being considered is from the system. This may be done by a receiver at any time. Boredom, for instance, may be thought of as a statein which a receiver simply ignores all but a few rudimentary at the entry point of the Variation/Enactment Normativeinteraction with the environment has been shut off. Similarly, a sort of laziness may cause a receiver .to interpret meanings literally during the Selection subsystem rather than hassle with confusing multiple meanings. In fact, it might be more correct to say that the information processing system becomes recursive when and only so long as a receiver maintains an "openness" or "interest" in gathering new informa-tion and seeking new meanings. In the absence 6f these attributes, the information processing system becomes lineal for the 93

PAGE 105

most part rather than recursive. We may say that the normal state of the information processing system is lineal, and recursiveness is the more unusual state. The second question pertains to the equivocation that is present in each component of the system. Up to this point, each subsystem.has been shown to be sensitive to a different kind of equivocation. Moreover, it has been stated that one function of each subsystem is to remove the appropriate equivocation during its pro6essing of the information. how is this equivo-cation "removed"? The answer to this question forms the substance of the next subsection. Removing Equivocation-Assembly Rules and the Double Interact Each of the three component subsystems that make up Weick's larger-organizing system have been shown to include processes that remove specific portions of the total amount of equivocation from the informational inputs introduced into it. Which portions of the total amount of equivocation are removed is dependent upon which subsystem is processing the information: 94

PAGE 106

1) the Variation/Enactment subsystem removes those equivocations that obscure a) the message's code, b) the elements within the message, and/or c) the message from being recognized as a message 2) the Selection removes those equivocations that obscure meanings and the interpretations of meanings within the being considered 3) the Retention subsystem removes those equivocations that a} cause contradictions, logical fallacies, and inconsistencies b) obscure relationships between meanings--i.e., the meaning of related meanings The question before us now is: How do these subsystems remove equivocation in order to produce an output containing recognizable and useful information and mean ing? Weick understood that the removal of equivocation from messages could not be a6complished until the amount of equivocation that needs to be removed is first determined. At first, this may seem like an unnecessary task. However, it will be recalled that one of the 95

PAGE 107

problems associated with equivocation is no two receivers will agree to the amount of uncertainty in any given message. One may view a message literally and straightforwardly. Another may view the same message as highly ambiguous and nonstraightforward. Each will process the differently. The latter receiver will take longer and will interpret "deeper" meanings than the former receiver. The former receiver will not remove much equivocation while the latter receiver must employ processes capable of removing large amounts of equivocation. So establishing the amount of equivocation in a message is critical for determining how a receiver will go abo.ut processing the message. _But there is another reason for determining how much equivocation is present in a message. In An Introduction. to Cybernetics (1956), w. Ross Ashby addressed the issue of equivocation in what he called the Law of Requisite Variety. Simply, the Law states that "variety Qan 9nly be destroyed by an equivalent or greater amount of variety'' (1956, p. 207). The term "variety" was intended to include any variable factor that contains randomness, chaos, or uncertainty. 96

PAGE 108

In terms of equivocation, the Law of Requisite Variety states that processes which are to combat equivocation must be at least equal in uncertainty to the equivocation being Unless the amount of equivocation is known, the composition of the processes required to deal with it cannot be known. Again the Law of Requisite Vatiety states that if the composition of the processes underestimates the actual amount of equivocation present, then the processes cannot remove the required amount of equivocation to make the information useful. This task of establishing the quantity of equivocation to be removed through processes of equivalent or greater randomness arid uncertainty was termed Registering the equivocation. Weick reGognized a second problem. If disorganization was associated with r'egistering the equivocation present in a message1 then tha actual removal of the equivocation necessarily required highly organized processes. In reality, the elimination of equivocation from messages, information, meanings, and the meaning of meanings requires a two-sided process. One side is random, disorganized, and uncertain enough to register the amount of .equivocatibn present. Once this amount 97

PAGE 109

is established, the second side that is ordered and organized removes the equivocation. The registration of equivocation is accomplished by what Weick calls Assembly Rules. The removal of equivocation is accomplished by Interlocked Behavior Cycles, or "double interacts." Assembly Rules are rules of limitation or constraint imposed by the receiver on the amount of equivocation that will be tolerated in the message or information input. The relationship between the number of necessary Assembly Rules and the amount of equivocation to be removed is an inverse one. Very few Assembly Rules will register very large amounts of equivocation. Conversely, greater numbers of Assembly Rules will register commensurately smaller amounts of equivocation. Another way of understanding this inverse relationship is to suggest that the more leeway a receiver is given in the interpretion of a given message, the more equivocation the receiver will perceive and therefore register. Obviously, the reverse is also true: The more restricted a receiver is in the interpretation of a given message, the smaller the amount of equivocation will be perceived and registered. 98

PAGE 110

If we return to the Variation/Enactment subsystem and the metaphorical "filter/aperture" that dictates how much and what kinds of messages are admitted to the overall system, we may say that the receiver's eral attentiveness or concentration as well as the gen-eral biases, prejudices,. assumptions, and weltanschauung the filter. affect whether the aperture is more open or more closed. As we inquired into the matter.of quality service, few if any restrictions placed upon our gathering of messages and intormatioti. were just as likely to find and valid information from=our customers as we were the waitress in a restaurant, the mech-anic at the corner garage, the minister-of the local church, or our spouses. With these varied informationsources, the aperture to our information processing system was very open. Because of the variety of message inputs, high levels of equivocation were also present in the messages we processed. Had we been more restricted in our information gathering, a different outcome would have occurred. Say, for instance, that upper management had said, "That's a great idea to learnmore about quality ser-vice! Let's randomly survey our most frequent custom-99

PAGE 111

ers and see what they like about us! Then we can give them more of what they like." This procedure requires that a series of restrictive Assembly Rules be put into place. The effect of the procedure will affect how information will be gathered and processed throughout the system. It will also restrict what information is or is not deemed relevant. The significant benefit to adding Assembly Rules restrictions is that significantly less equivocation will be present in the messages. Another restrictive Assembly Rule that pervades organizational activity is Time. If information is needed within a predetermined time frame, then less equivocation is a necessary factor. There simply is not enough time available for the information processing system to register and remove large levels of equivocation. The downside to time.restrictions is obvious: Message variety. must be sacrificed. On the other hand, freedom from time restrictions allows for a greater freedom for gathering and processing diverse message inputs. There is also commensurately more equivocation to contend with. Before turning to the removal of equivocation, two interesting questions surface about Assembly Rules and equivocation. Given any message (M1 ), does the 100

PAGE 112

construction of Assembly Rules dictate how much equivo-cation is present in M 1 ? or is a fixed amount of equivocation in M 1 that dictates how the Assembly Rules should be. constructed? A comprehensive treatment of these questions will be reserved for Chapter 3 and specifically for the dis-cussion on Blaine Goss's Information/Noise Ratio (1989). In the meantime, two relevant point, may be made: 1) If we assume that some receiver constructs any two hypothetical Assembly Rules and then receives M 1 these restrictions wili only be able to register a certain amount of equivocatibrt (E1 ) and no more. We say that M 1 will processed as containing E1 equivocation. If at a later time this receiver adds a third Assembly Rule (e.g., time has become a factor), then a commensurately smaller amount of equivoca:tion (E < E 1 ) can be 2 registered. M 1 no longer contains E 1 equivocation. From the same receiver's standpoint, M 1 now contains only E 2 equivocation. 2) N6 single message a fixed amount of equivo-cation. As seen above, Assembly Rules can affect the amount of equivo6ation qontained in a message. Also, the receiver's perception of equivocation 1 01

PAGE 113

affects the amount of equivocation contained in a message. Equivocation is like "trouble"; a person really looking for it can find it anywhere and in just about any amount. Just as the number of Assembly Rules dictates the amount of equivocation present in any message and registers that amount, the registered amount of equivocation in the message will dictate how many Interlocked Behavior Cycles will be necessary to remove the equivocation. Weick defines _Interlocked Behavior Cycles as a series of double interacts. A double interact consists in an exchange and feedback about the exchange. When the results !rom one double interact forms the exchange for the next double interact, the two double interacts have become interlocked. The completion of one double interact forms one cycle. Successive.interlocked cycles form the pattern Weick describes for removing equivocation from messages and meanings. The number of double interacts that are necessary to remove equivocation in any input is in direct proportion to the amount of equivocation previously registered by the Assembly_Rules. The removal of large amounts of equivocation requires an equivalently large number of Interlocked Behavior Cycles. 102

PAGE 114

A double interact may be thought of as the testing of a possible option (the exchange) and the receipt of the testing results (the feedback). In the case of the Zen one solution to the paradox was given to the master. This exchange was met with feed from the master in the form of a response. As described iti the earlier discussion, the information gained from this communication loop was used by the initiate to compose the substance of the next exchange which, in turn, was responded to by the master. This repetitive routine is not only recursive, but it also exemplifies Interlocked Behavior Cycles as equivocationremoving processes. It will be useful now to summarize this long chapter section. In order to understand how the recursive inquiry works, the model presented by Karl Weick in his Social Psychology of Organizing (1969) was described. Examples of recursive inquiries from earlier portions of this thesis were plotted onto Weick's model in order to emphasize the many points of similarity. Having introduced concepts that describe how a recursive inquiry works, the final section of this chapter will introduce work by Gregory Bateson (1964) that assists in describing what a recursive inquiry accom103

PAGE 115

plishes. By way of introducing the final section and concluding this section, a couple of salient ideas will be noted. The notion of feedback has been used in a couple of contexts in the presentation of Weick's model. Wiener (1967) noted that feedback was indispensible to the conduct of future actions. Together with memory, feedback provides a means by which the experiences of the past mold the responses to the future. In other words, feedback and memory are requirements for learning. Like all good information-processing models, Weick's model is also a learning model. What is particularly intriguing about considering Weick's model as a learning model is the descriptive power with which it clarifies how information is converted into change. As a message moves from one end of the system to the other, a sequence is built that follows a message being transformed into information which is then transformed into meanings which in the Retention subsystem cause a change to occur in the entire system. It would be incorrect to suggest that change within the system occurs because some brand new meaning has just popped out of the information-processing system. Important as this new addition may be, the real change 104

PAGE 116

comes as a result of the meaning that is derived from the relationship the new meaning has to the extant meanings. Theie are two different types of meanings that coexist in the Retention subsystem. In this two-tiered hierarchy, the more numerous meanings are those processed from messages. The other meanings are relational. These meanings may be derived from the relationships between and among lower-ordered, messagederived meanings. Also, other relational meanings may be derived from relationships between and among lowerordered relational meanings. Change occurs when a relational meaning changes due to the addition of any new message-derived meaning. More significant change comes about when a significant change in a relational meaning occurs. This will happen when some new message-derived meaning does not fit neatly into the extant body of meanings. By extension, massive change will occur when a properly processed, new messagederived meaning contradicts the extant body of meanings. Change due to relational meanings are directly related to the amount of equivocation present in the Retention subsystem. This relationship was illustrated earlier by the discussion on equivocation as it was related to Ferguson's Surprise Exam Paradox. It may 105

PAGE 117

be said from this that major changes occur when: 1) Significantly large amounts of equivocation in the Retention system are present and are removed, and 2) A significant change has occurred in Relational meanings as a result of the removal of this equivocation. Two consequences of change in _Relational meanings are that future interactions with the environment will be different, and the itself will be understood differently. Out of these two consequences, a third consequence emerges: the information processing system itself will change, and information will be processed differently. We may call these consequences a rough sketch of learning and intellectual growth. Finally, eVidence that change has taken place can be measured ai a charige or difference in the ways and means by new interactions with the environment occur. This was discussed as the second element of Weick's Choice Point. New information and new Relational meanings will cause the system to respond differently in future interactions with the environment. Bateson's Hierarchy of Learning model was specifically chosen to assist in explaining what a recur106

PAGE 118

sive inquiry does because the Hierarchy is built upon these same three premises: 1) Information and, more importantly, meanings are ordered according to hierarchies 2) Hierarchical levels of information and meanings cause commensurate hierarchical levels of change and 3) Change is discernable as responses to the rele-vant Gregory Logical Categories of Learning: Hierarchy for Mapping Change Response In order to address the question "What does a recursive inquiry accomplish?," it will be necessary to estab-lish some sort of or chart on which we may place various types of learning phenomena including the recur-sive inquiry so that we may compare them together. Greg-ory Bateson provided such a map with his Logical Categories of Learning and Communication (1964). Both Weick's model and Bateson's hierarchy share at least one crucial assumption: Change due to learning can be measured as a change observable through response. This assumption is based upon a series of transformations that begin with information. This series may be summar107

PAGE 119

ized as follows: New information is processed into new meanings which cause changes in we called relational meanings. Change in relational meanings causes a change in the way a receiver perceives and understands the relevant environment. As a consequence of this process of learning, the receiver must respond differently in subsequent interactions with the environment. Moreover, both Weick and Bateson agree that any change in an individual's response to information necessarily indicates that some sort of change due to some sort of learning has occurred. While Weick does not take the implications of this con6lusion any further, Bateson begins his hierarchy at this very point. While neither a receiver nor an outside observer may reliably investigate what is actually happening inside a human brairi as it processes information and culls meanings, both a receiver and an outside observer may reliably investigate shifts and changes in their own or others' responses. Small shifts and changes we may attribute to relatively small in relational meanings. Larger shifts and changes may be attributed to commensurately larger changes in relational meanings. Bateson's Logical Categories of Learning and Communication are built out of the idea that changes in 108

PAGE 120

response ahd derivatives of changes in response are reliable evidence of changes in Learning and derivatives of changes in Learning. Before describing Bateson's Five Levels of Learning, a few overview comments are in order. Bateson's Hierarchical Learning structure is modeled after the Theory ofLogical Types expounded by Russell and Whitehead in Mathematica (1910-13). This Theory asserts that in any formalized line of logic or mathematical development, no 6ategory of classification can specific.member of its own classification. Stated in other ways, the riame of. the thing is not the thing That is, the name is 6f a different logical the thing itself. Alfred Korzybski (1941) stated this notion in his famous quote, "The map is not the territory." Any class of items, then, is of a higher logical type than any item in the class. Moreover, any class of items is of a lower logical type than the class which names the relationship it has with other classes. Because they are derived from a structure of logical types, Bateson's Levels of Learning avoid one of the plagues common to other hierarchical structures that map learning. There are no "gray areas'' that occur 109

PAGE 121

between the interstices of the Five Levels. Each Level is precisely self contained and a clear distinction separates 6ne level from another. Any communication transaction or portion of any transaction can be assigned to one and only one classification level. With Bateson's hierarchy, a consistent description of a communication transaction can be reached. Perhaps this notion of logical typing has already been recognized as a name for the distinction we have drawn between message-derived meanings and relational meanings. Beyond this, the Zen koan example portrayed another application of logical It will be remembered that the goal of the exercise was for the initiate to define a rule or "algorithm" which encompassed the range of previous correct responses to the "one hand clapping" paradox. This rule.was considered a valid end to the exercise if it shown capable of gen erating other correct dbviously, the range of correct responses form a logLdal type lower than the logical type by the rule which generates them. Logical typing is important to our understanding of the recursive inquiry. Bateson also reminds us that logical typing is an all-pervasive attribute of very 11 0

PAGE 122

normal communication. In fact, logical typing may be applied to any communication transaction. For instance, interpersonal conflicts often arise when one individual's playful gibe is misinterpreted by another as a flagrant insult. This, Bateson points out, is an instance of two individuals interacting on different logical type levels of meaning. In the Palo Alto Group's work with schizophrenics, blatant disregard was shown by patients for interpreting messages at the logical type level of meaning intended by the message's source. It was discovered that one-distinguishing characteristics of schizophrenic patholo9ies was an innocent insistence on interpreting any_messages perceived as "threatening" in an entirely literal manner. One:patienf, f6r example, left his work early without supervisory permissi6n and drove home. An employee happened to call 'him at h6me and asked sarcastically, "Well, how did you get THERE?" The patient's reply was, "By automobile" (195-6, p. 209). Whether or not the patient knew and disregarded the sarcasm, this kind of transaqtion is characterized by two communicants operating on different logical type levels of meaning. Bateson (1964) that within the coding of messages is information that clues a receiver about the 1 1 1

PAGE 123

logical type level which the source intends the informa-tion to be interpreted. Bateson calls these clues context markers. These markers not only serve to distinguish each of the Five Learning but they also serve to distinguish how messages within each level may be hierarchically arranged. With these preliminary we may now summarize each of Bateson's Five Learning Levels. Level Zero Learning Withiri this Hierarchy, is defined as the change that occurs as a result of the processing of inforThe evidence of"learning is found in the type of response or the change in the type of response made as a result of the presence of information. The measure of the change in response is equivalent to the change brought about due to learning. Inorder to establish a or foundation upon which to build the Hierarchy, _Bateson defined Level Zero Learning as the invariant response of a receiver to a given message. That is, of t&e frequency I with which a message is sent, each time it is transmit-ted, the response will alJays be the same. There is 112

PAGE 124

Zero change in the receiver's response, and that response is 100% predictable. Messages and information are not "evaluated" nor "considered," and "choices11 are not made at Level Zero. Metaphorically, messages and information are processed as inviolable instructions or procedures to be carried out. Responses are 11wired into11 the receiver. At Level Zero, the Retention subsystem is conspicuously missing the presence of Relational meanings. A programmed computer exemplifies this type of learning. For any given information input, the computer's program conducts an invariant set of procedures before leading to an output. All informational inputs are treated according to pre-established processes. Seem ingly random or evaluative piocesses are actually preprogrammed processing loops that are mimicking purposive processes. Zero Level Learning is not intended to be limited to the electronic or inanimate worlds. Zero Level Learning is well represented in the organic and biological worlds. For instance, when information inputs signal dangerously high levels of heat, any amoeba will respond by retreating from the source with 100% reliability. This same response will occur in 100% of all other cases 11 3

PAGE 125

that the amoeba receives and interprets information about threatening levels of heat. Like the amoeba, the vastly more complex human being will retreat from a heat source for the same reasons when the information sensors in the skin transmit dangerous levels of heat. Autonomic biological systems, involuntary functions, and reflex behaviors are further illustrations of infer-. mation processing at Level Zero. Moreover, all genetically "wired in" responses are also instances of Zero Level Learning. In biological systems, this inescapable one-to-one relationship between information and response is critical to the survival or perpetuation of the brium of the organism. We may also say that Zero Level Learning is in evidence in chemical systems. The reliability of response characterized in Level Zero is a highly desirable attri wheri it comes to the manufacture of sophisticated plastids, alloys, and drugs with highly rigid specifications because reliability means that replication can be controlled. An entirely predictable end result is crucial in the production of these special products and, for that matter, any product that involves the replication of a desired chemical result. 1 1 4

PAGE 126

Because molecules combine in predictable ways under specified conditions, we may say chemical reactions represent a Zero Level Learning system. Two additional comments are worth mentioning. First, one-to-one relationships between information and response do not necessarily mean that Level Zero Learning represents what we referred to earlier communication transactions. The contrary is actually true. Highly complex expressions of communication transactions are For-instance, the DNA double helix is an information source that contains a finite but incredibly vast number of message inputs. Proteins are the receivers in this biochemical message and information exchange. That any protein-receiver can carry out the instructions encoded into any given DNA site in the same manner at any given time fulfills the conditions for Level Zero Learning. The elaboration of this exchange is, of course, among the most complex information pro-cessing systems in the biological world. 1982, pp. 112-115.) (Cf., Campbell, Since no "evaluation" of information inputs is pos-sible at Level Zero, information is interpreted by the receiver as is. This matter has a significant impact on the biological world. In a genetic context, for exam-11 5

PAGE 127

ple, if a message site is damaged severely enough, a protein will no longer "recognize" it as a message nor "recognize" that it once contained information. The protein has no memory for what genetic material was contained at the damaged site. That information input is now irretrievably lost to the system. On the other hand, if a genetic message site is damaged in a manner that results in a mutation, we may say that a new set of genetic instructions have the old set. If the newset of instructions are "recognized" as genetic messages by the protein, that protein will not distinguish any differences. So far as the protein receiver is concerned, the mutation is neither new nor old, neither good nor bad. The mutation simply is the current set of instructions to be carried out. We may conclude this discussion of Zero Level Learning by saying that at this messages and information are processed without regard to the context in which tpat information is transmitted. Moreover, context mar-kers and certainly meanings are not necessary nor relevant to the carrying out of Zero Level Learning. Communication is entirely lineal: The receiver simply accepts and performs the requirements laid out by the information. There is no circular feedback loop to let the source 11 6

PAGE 128

know that the information has been received and understood. Level Learning This Level is defined as a change in the type of response to information characterized in Level Zero Learning. Since responses in Level Zero are typified by fie one-to-one relationships to the information being processed, Level I is typified by a change in the specificity of the response to information. The more ver.nacular meaning of the term "learning" is used in this Level. A Level I receiver "learns" a new response and replaces an old response with this new reponse when a given message input is transmitted. Bateson stresses that a Level response is not one that a receiver selects from a repertoire of possible responses. Rather, a response learned at Level I will either replace a Level Zero response or another Level I response which had itself previously replaced a Level Zero response. Level I Learning defines the kinds of learning normally associated with habituation, Pavlovian conditioning, rote learning, and instrumental rewards or punishments. As such, Level I receivers learn that a 11 7

PAGE 129

certain new response to a given information input has certain desirable consequences (or, perhaps, less undesir-able consequences) that are not associated with the old response. These consequences reinforce the replacement of the old response with a new response. Characteristic of Level I Learning is the disruption, extinction, or inhibition of the newly learned response over time when the reinforcements or consequences are removed. It is at Level I that a message's context first emerges as a contributing factor in how the information is to be interpreted and Subsequently responded to. In order for something like conditioning to work, a mes-. sage must riot only contain its basic complement of infer-mation but it must also contain a contextual meaning that the receiver to respond in a desired new way. Associated with the basic information content, then, is the expectation that a response will in some way a positive benefit to the receiver. This may be stated in another way: At Level I, a receiver's response is mediated by a meaning composed of a message1s information content and context. With the addition of message context to the information pro-cessing scheme, Bateson's context markers come into play as a means of discriminating possible contexts from one 11 8

PAGE 130

another. Also, Level I marks the first time that the messaging scheme includes feedback. The receiver's response must be reinforced--it must prompt a return response or it cannot supplant a previous response. From this we may say that if Level Zero Learning is lineal from source to receiver, then Level I Learning is circular. Source and receiver are bound together by a feedback loop. Level II (Deutero-) Learning This Level names the learning that occurs which results in a change in Level I responses. Also referred to as "Deutero-Learning" by Bateson, Level II Learning marks the first appearance of "choice" among meaning alternatives. Depending upon which meaning is selected, any message input will elicit one of a number of possible responses. Additionally, Bateson refers to "learning to learn" as an applicable description of Level II. By this reference, Bateson is conveying one of two notions. First, that independently of external reinforcements (the absence of Level I conditioning), a new meaning has been derived from new or given message or information input, or secondly, that an extant meaning has been generalized 11 9

PAGE 131

to incorporate new meanings or previously unrelated meanings in novel contexts. Evidence that these internal learning processes have taken place t6 be seen in the manner in which the receiver responds to message inputs and contexts. In the former case, the receiver may be expected to respond to a new message in an unpredictable way, or the receiver may respond to a familiar message in an unexpected way. In the latter case,. a receiver may respond in a familiar in an entirely unfamiliar riontext, or the receiver may respond in the same way to contexts that previouslywere responded tp differently. While a change in a receiveris response is often caused by a of meaning, a receiver may simply choose to change the itself. This occurs frequently at Level II. For example, peers .may bombard a receiver with messages to conform to a social norm. While the messages by themselves may not create meanings that prompt the receiver to respond with conformity, the receiver may nonetheless choose 'to respond with conformity. The selection made by the receiver in this case is not a choice among possible meanings but is rather a choice among possible 120

PAGE 132

This creates an interesting problem which did not elude Bateson's attention. At Level II, then, changes measured in a receiver's may be attributable to one of two very different causes: 1) A change occurred in what we called Relational meanings and carne about as a result of. choices made in the interpretation of message-derived meanings or 2) A deliberate choice was made to select a response different from that response which would have emerged from Relational and message-derived meanings. Whenever a receiver's response comes about as a consequence of a change in meanings, we may say that the response was conformed by the meanings. More importantly, we may say that the response is consistent with the extant body of meanings. Also, we may say that a meaning-conformed response has no need to be selected, or chosen because it emerges of its own. For this reason, if a receiver's response comes about as a consequence of a choice among responses, then the meaning-conformed response is necessarily de-selected in favor of another response. When this happens, then a disparity exists between the response and the extant body of meanings. This disparity describes the all-too-1 21

PAGE 133

human inconsistencies that emerge when what we SAY and what we DO are altogether different or may even be con-tradictory. Besides creating a disparity between meanings and response, a more far-reaching information processing problem occurs when a receiver chooses to select among responses. This selection of a response from a set of possible responses precludes that sort of criteria will be used to make the selection. Regardless of the criteria used, the receiver will prioritize the responses according to how desirable the outcomes of each response are perceived to be. Respon-ses perceived to have less desirable outcomes are less likely to be chosen. When a receiver perceives that all responses carry. undesirable consequences, then the receiver is most likely to choose a response with the least undesirable consequences. When a receiver falls into the trap of "overriding" a meaning-conformed response by selecting a response according to its perceived desirability, then the self-validating characteristic of the content of Learning II has the effect that such learning is almost ineradicable (Bateson, 1964, p. 301). To understand this "self-validating characteristic," we need to recall that in Weick's information processing 122

PAGE 134

model, the receiver's response accomplishes two purposes . Any response will determine how the receiver interacts with (i.e., enacts) the environment and will form the feedback connection that closes the information processing loop and recursively restarts the processing system again. As the feedback loop, the response will determine what selection filters and aperture opening will be used for admitting (and by consequence ignoring) message inputs. If a receiver's response is a chosen response, then the receiver's interaction with the environment is not based upon the body of extant Relational and messagederived meanings. Rather, the receiver's interaction with the environment is necessarily predicated on the perception of desirable outcomes. Also, chosen responses will bias the information processing system itself. Messages validating the response will be .more readilyadmitted to the system at the exclusion of messages contradicting the response. These new messages will, by their very nature, create highly equivocal meanings when compared to the extant meanings although they will "seem" to be low in equivocation when they enter the system. The illusion of Level II Learning is that it is an end unto itself. The more a receiver holds to select-123

PAGE 135

ing responses, the more the environment seems to validate the receiver's perception of it and the less complex the environment seems to be. In actuality, the selected responses are biasing the information processing system itself into selecting only those messages that do not contradict the selected response. The illusion that the environment is less is predicated on an information processing system that is more restrictive in the types of message it to process. Because of these self-validating attributes, skilled Level II learners hold ferociously to the interpretation/ response schemes they have created. New situations and conditions are more likely to be fitted into an existing interpretation/response scheme than is a new scheme to be developed along side of or in contradiction to an older scheme. Bateson calls this self-validating restriction of messages and information inputs as the 11segmentation and punctuation of the stream of action and experience into [convenient] contexts" (1964, p. 2-93). Level III Learning Earlier it was mentioned that one of the desirable characteristics of Bateson's Levels of Learning Hierarchy 124

PAGE 136

was the clear distinctions made between learning levels. It was pointed out that this attribute made categorizing message and information exchanges reasonably easy and reduced confusions due to "gray areas" found in other message categorizing schemes. Since the distinctions between Levels II and III are so pronounced, it may be useful to explore these differences between learning levels. There two types of learning systems that populate each level. At Level &ero, there is one type of system that may be biol6gical, or mechanical which is incapable of than Level Zero responses. Level Zero. The learning ceiling for these systems is That is, they do not have the apparatus necessary to respond at a higher level of learning. A second type of system that may also be biological, electronic, or mechanical is one that conducts Level Zero activity but has the capacity to conduct activity at higher levels also. Our amoeba represents the first type. Pavlov's dogs represent the second type. Likewise at Level I, there are those learning systems that have a .Level I learning ceiling and that may conduct activity at Levels Zero and I but at no higher 125

PAGE 137

level. There are also those learning systems that have a higher learning ceiling than Level I that are free to conduct activity at Levels Zero, I, and at least II. Some electronic and mechanical systems exhibit Level I attributes such as computers, thermostats, and machine engine governors, but to Level I seems to be the ceiling for electronic and mechanical systems. Most animals have Level I ceilings, too. A few biological learning systems seem capable of conducting activity at Level II (cf., Bateson, 1966). Excluding humans, Level II is the learning ceiling for all bioLogical systems. When we speak of a learning "ceiling," a reference is being made to the incapacity of the information processing system to process messages and information and to derive meanings at the next higher learning level. We may also say that the system'is incapable of perceiving that there is a next higher learning level. On the other hand, it is also pointless for the amoeba to concern itself with Level I activity when Level Zero information processing is entirely sufficient for its survival. The learning ceiling for humans is Level III, but a problem that intrigued Bateson surfaces here. For those receivers who fall into the self-validating trap 126

PAGE 138

described in the Level II section above, Level II activity seems sufficient unto itself. In other words, when the human information processing system learns to reinforce its own responses by validating the responses with information that confirms but never denies them, the receiver has developed a self-perpetuating information system that excludes contradictory messages and information. In effect, these Level II receivers have imposed an artificial learning ceiling at Level II. As a consequence of this ceiling, these receivers will be mistakenly led to believe Level III does not exist and will have difficulty imagining how Level III activity works. As the description for Level III unfolds, it will be recognized that few humans are regularly active at Level III. Level III is defined as the change in Level II responses. Level III will also be shown to consist in the kinds of learning activities which have been identified as recursive inquiries. Bateson notes that access to Level III is not contingent upon intellectual acumen, special giftedness, nor wishful thinking. Rather, Bateson states that an individual "is driven to Level III by 'contraries' generated at Level II 127

PAGE 139

(1964, p. 305). Based upon our earlier investigation of equivocation and paradox, we may say that the "contraries" Bateson refers to are the highly equivocal dis-parities that emerge between extant meanings and selected responses or the discovery that messages and information from the environment are more subtle and equivocal than the information processing system can accommodate. If Level III Learning requires a change in Level II responses, then we may also look for a change to occur in the acquisition, interpretation, and use of information at Level II. Referring to this type of change, Bateson says, something of this sort occurs in psychotherapy, religious conversion, and in other sequences in which there is profound reorganization of character (1964, p. 301 ). In order to understand what Level III responses are, it will once again be necessary to review 0hat is occurring among meanings in Weick's Retention subsystem. At Level II, one of two possible selection processes is made by a receiver in order to achieve a Level II response. One possible process selects among possible meanings and leads to a new relational meaning which is then expressed by the related change in response. A second possible process selects among possible res-128

PAGE 140

ponses and leads to a response chosen independently of any extant relational meaning. Relational meanings refer to those interpretations that connect meanings derived from individual messages or information inputs. In this respect, relational meanings define the relationship these message-derived meanings have to one another. A Level II has a host of relational meanings tying together variously related message-derived meanings. We may expect, for instance, that a relational meaning would bind together meanings associated with "intimacy." Another Relational meaning would bind together meanings associated with "conflict." Stillanother would bind together meanings associated with "social events" and so forth. At Level III, the receiver synthesizes the relationships among relational meanings highly generalized meanings. In drawing the relatedness among relational meanings, these new meanings could be called interrelational meanings. But these new meanings actually do more than interrelate relational meanings. These new meanings also reflect a purposiveness or an orderliness that comes about when highly equivocal "contraries" are integrated 1 29

PAGE 141

by meaning. Over time, the Level III receiver's activity is to integrate more and more relational meanings into fewer and fewer of these Mega-meanings. In the truest sense, the response from a Level III receiver is an enactment on the environment. Also, a Level III response will tend to open the information processing system to larger, more disparate message and information inputs from the environment rather than restrict message inputs. Chapter 3 consists of a more worthy treatment of Level III Learning and elaborates upon these introductory comments. Level IV Learning Level III learning contains the kind of activity that integrates meanings of lower logical type into meanings of higher logical type. If one to try to imagine what-the Level III ceiling looks like, the skilled Level III learner would have integrated meanings into a bare few Mega-meanings. At this point, Bateson suggests that a change in the types of responses characteristic of Level III would necessitate "the combination of phylogenesis with ontogenesis" (1964, p. 293). Bateson also notes that this 130

PAGE 142

Level IV does not exist in any known terrestrial organism. It is also doubtful that any human communication system could come close to capturing what such a combination of phylogenesis with ontogenesis would look like or function like. Concluding Remarks Chapter 2 of this thesis has been devoted to introducing and elaborating upon the work of Claude Shannon and Warren Weaver (1949), karl Weick (1969, 1979) and Gregory Bateson (1964). There were two purposes for introducing these works: 1) The first purpose was to tie the major attributes of the activity called the recursive inquiry to current communication thinking. 2) The second purpose was to flesh out further our understanding of the recursive inquiry through somewhat familiar descriptions of well-conceived models. Since the recursive inquiry by definition addresses highly uncertain, equivocal, and ambiguous messages and information, Shannon and Weaver's Mathematical Theory of Communication was used to establish definitions for key terms as they related to fundamental communication 1 31

PAGE 143

concepts. Because the MTC itself makes clear distinctions in its use of terminology, it was particularly helpful in distinguishing the differences between "uncertainty," "equivocation," and "ambiguity." These distinctions were crucial to understanding how a receiver goes about determining which parts of a message obscure its meaning and which parts enhance its meaning. Another important characteristic of the recursive inquiry is that of establishing useful meanings out of highly uncertain, equivocal, and ambiguous messages and information. In an attempt to understand how these sorts of messages and informational inputs are made sense of, Karl Weick's model for Information Processing was introduced. Pivotal for the selection of this work was the emphasis Weick paid to the role equivocation plays in the obscuration of meaning and the role response plays in making his information processing system recursive. The final important characteristic of the recursive inquiry is the role it plays in learning. In understanding the power a recursive inquiry can play in the learning process, it was necessary to introduce a learning categorization scheme. Gregory Bateson's Logical Categories of Learning and Communication was chosen since it is built upon shifts and changes in the manner in which 132

PAGE 144

a communicator responds to informational inputs, message derived meanings, and relational meanings. This scheme fit nicely into Weick's scheme for processing information and responding to that processing. With all of the major pieces conceptually in place, Chapter 3 will: 1) Begin with some introductory remarks on perception of equivocation and will proceed with a discussion of Blaine Goss's Information/Noise Ratio. From this, we will derive an expression that relates information and equivocation to the establishment of meaning. This relationship will help us understand what a recursive inquiry can accomplish, and the interplay of limits and boundaiies on information and meaning 2) Introduce "patterns" and, in particular, the Moire Phenomenon will be discussed as it relates to the combining of patterns into new patterns and defining the constituent elements of patterns 3) Introduce Peter Senge's Vision/Current Reality Gap metaphor as it relates to the activities and purposes of the recursive inquiry and both individual and organizational learning 133

PAGE 145

4) Continue the discussion of patterns as they enhance the descriptions of Learning Levels II arid III and as they relate to a receiver's responses at these levels. This discussion will focus on patterns as relating to collaboration within Level II and III contexts 5) Conclude with a general discussion of the implications of the recursive inquiry to Level III Learning. 134

PAGE 146

CHAPTER 3 AN EXPOSITION OF THE RECURSIVE INQUIRY AND THE GENERATION OF PATTERNS Introductory Remarks It will be recalled that the goal of a recursive inquiry is to take a highly equivocal message and through processes described in Chapter 2, to remove the equivo-cation so that meaning and application may be reached. We have already noted that no two receivers per-ceive the same amount of equivocation in any given mes-sage. One of the two receivers will most likely per-ceive more equivocation in the than the other will. This is of importance insofar as a receiver can only remove as much equivocation as is perceived. This perceived amount was called registering the equivocation in the message. If I am working alone on a problem that I perceive to be high in equivocation, then I had better construct an information processing strategy that has very few Assembly Rules and includes a great deal of processing time. If I do not construct this type of strategy, then I cannot hope to maximize the amount of 135

PAGE 147

equivocation I can potentially remove. In actuality, my work on the problem consists of two critical attitudes toward equivocation: 1) I must recognize (i.e., perceive) how much equivocation is in the message ahd 2) I must construct an information processing strategy that is tailored to remove the amount of equivocation I have recognized. If, for some reason, I mismatch my information processing strategy to the amount of equivocation I perceive, then one of two outcomes will occur. If I perceive a great deal of equivocation but construct a rigid number of Assembly Rules restrictions, then I will remove too small an amount of equivocation compared to what I perceive. Unless I make an adjustment in the strategy later, I will probably respond to the problem in too simplistic a manner. I will not have given the problem its due consideration. On the other hand, if I perceive little equivocation in a message but construct an information processing strategy that is designed to remove large amounts of equivocation, then I will fall into Hamlet's trap: I will be incapable of taking action--always seeing more in the problem than is truly thereo 136

PAGE 148

Now if I am working on a highly equivocal problem with another person, information processing becomes at least one degree more difficult. Our first order of business must be to achieve some sort of agreement on the amount of equivocation we both perceive. A classic example of this occurs in conflict. Let us assume that a married couple is disagreeing over the amount of quality time they are spending together each week. One spouse feels that given a host of contingent factors and in comparison to other past busy times, that the amount of quality time has been quite good. These considerations make the content of the disagreement "cloudy" at best to the spouse. The other spouse considers the content of the disagreement quite clear: The amount of quality time may be more, but it is still not enough. Unless the one spouse can convince the other that this is not a straightforward disagreement and that there are gray areas to be understood, then the maximum amount of equivocation that will be removed in this disagreement is the amount distinguished by the second spouse. We may summarize this conclusion by saying that the maximum amount of equivocation any two or more receivers can remove from an equivocal input is the maximum amount 137

PAGE 149

of equivocation they both perceive together. This limit always be determined by the individual who perceives the least amount of equivocation in the input. However, this limit may be modified to the extent that the limiting receiver is persuaded or convinced to perceive more equivocation than was originally perceived. In our consideration of the quality service question, we founa that a substantial amount of was devoted to 11sharing11 results with our .department members. One of the ciitical of this ing time was "discussion of what remained to be learned or what was.needed in order to learn more. In the example given in the Preface, this became the Now that we know what we know, ho"w can we improve the quality of our service? This discussion proved to persuade members that we had still more equivocation to process. If a perception of equivocation is critical, so is the attentionor concentration of the receiver. Attentiveness menti6ned eariier as an important part of how messages were selected for input into the processing system itself. If a message is not first 11attended to,11 then the message cannot be selected for processing at all. Any messages not 11attended to" are automatically de-selected from consideration. 138

PAGE 150

If attentiveness refers to the sensory means by which a receiver selects certain messages and ignores the vast majority of other messages, then we may refer to the receiver's concentration .as the sensory means by which a receiver attends to the processing of the message. This creates a distinction between attention and concentration as well as relates the two concepts together. We may also define concentration as the absence of distraction during the processing of a message. In a similar manner, we may define attentiveness as the absence of distraction during the message selection process. Distraction may be then, as sensory intrusions that disrupt the selection and processing of messages. In the literature, sensory intrusions are attributed generically to noise,. and noise is a type of equivocation. Whether noise is auditory,visual, tactile, or a form of malaise, its presence iri the information processing system is omnipresent. Obviously, the more adept a receiver is at reducing the presence of noise, the more adeptly the receiver will process the message and remove the other forms of equivocation. In short, reduced noise smooths the processing of messages. Conversely, 139

PAGE 151

increased noise levels not only add additional equivo-cation into the processing system, but they also impede the progress of message processing. Further consideration of this matter is worthy of discussion as it relates to how certain messages may be considered recursive inquiries or not and how the the conduct of the recursive inquiry may be impeded or facilitated. This chapter will proceed with a discussion of Blaine Goss's (1989) Information/Noise Ratio. From his Ratio, we will develop concept into a formal relationship and then discuss the limits of a message's information content and meaning this formal relationship suggests. Blaine Goss's Information/Noise Ratio and the Information and Meaning in Messages From any receiver's point of view, the information contained in a message will be variably damaged by both the amount of equivocation and noise present in the exchange. Noise is a form of equivocation but unlike other forms has a measure of control associated with it. Like other forms of equivocation, the presence of noise will reduce the amount of information contained in the message. Controlling noise greatly aids informa140

PAGE 152

tion processing, and this control is a form of equivo-cation removal. Noise can be caused by external factors or by the receiver's own internal conditions. In its common form, noise usually refers to external intrusions that are to be found in the environment. However, in considering noise as a form of equivocation, it is desirable to expand the consideration of noise to include internal forms. Internal noise is intended to refer to those intru-sions that come about due to the internal."chatter" that is characteristic of the flow of normal human conscious-ness and both voluntary and involuntary bitilogical functions. Like externai noise, the amount of.the "chatter" that is present in a communication transaction affects the processing of information at any time. Either one of the two forms of noise tends to rob the receiver of attentiveness and concentration in the processing of messages. Blaine Goss ( 1989) points out that any individual has a finite amount of attention . and concentration that can be devoted to message inputs. Over time, these reserves of attention and concentration diminish, and a receiver becomes more susceptible to the intrusionof noise. The inatten-tive and "loses" concentration. 1 41

PAGE 153

Goss uses this observation to describe the flux that occurs in normal communication. When a larger percentage of attentiveness is drawn away by noise from the process of selecting messages, less concentration is available to be devoted to the processing of the message and the smaller the amount of information will be processed. From this Ratio, we may combine some of the observations we have made about a message's informational content and about noise.and equivocation. These combinations will allow us to derive an expression from Goss's Ratio that will show how the interplay of iriformation and equivocation affects meanings. This information/noise relationship is in constant flux as is to all who found themselves wandering off from a conversation that requires them consciously to refocus their attentions back to the conversation. This wandering off is just as likely to be the result of external influences (as in an argument that is brewing at the far table of a restaurant or in the general noisiness of a chaotic classroom) as internal influences (as in a daydream, a sense of discomfort, or a sense of general weariness). 142

PAGE 154

If the amount of information in a message drops as the amount of noise increases and if the amount of information present in the message affects the meaning of that information, then we may extend Goss's Informa-tion/Noise Ratio to say that a message's meaning will be directly affected by the flux ot information and noise. This relationship is particularly important to a recursive inquiry when a message's meaning is the result of a signi-. ficantly long-term time information/noise flux will have a variable but significant impact on the process over time. We may show. our extension of Goss's Ratio schema-tically as follows: if M 1 = MEANING derived _by the receiver from some message1 and I 1 = INFORMATION CONTENT in message1 and N 1 = NOISE intrusion in the processing of message1 then, the following expression may be made: ( 3. 1 ) For meanings are derived from long-term processes, we may say that the meaning be related to the aver-age perception of the information content over time and the average amount of noise intrusions during that time. 143

PAGE 155

We may expand this_expression by taking into con-sideration our earlier observation that noise is composed of both external and internal noise factors: N 1 = external NOISE (Ne) + internal NOISE (Ni) We may also expand our information variable. Any message contains two types of information: 1) a fixed amount of "content" (or "primary") information, which consists in the purpose of transmitting the message, and 2) "context" (or "secondary") information, which consists in determining how the receiver is to interpret the content information. Therefore, I1 = Primary content (Ip) + Secondary context (Is) Our newly expanded expression now look like this: = ( Ip + Is) (N + N.) e l. ( 3. 2) At any given frozen moment in time, the meaning and information content of any message will be depen dent at least upon the amount of noise present during the processing of the information. As noise factors are reduced, then meaning must necessarily increase. Control of external noises is often easiest to achieve, and this control is critical if a receiver is trying to maximize an understanding of any message and its information contents. If the processes that forge 144

PAGE 156

meanings are facilitated by quiet environments, then they are doubly facilitated by the quieting of the senses. If external noises are reduced to or near zero and internal noises can be quieted, then some interesting things happen to meanings. As the noise factor reaches closer and closer to zero, then the meaning value sky-rockets. This, of course, is the claim of yogis and mystics (cf., Yogananda, 1946, pp. Credence for this conclusion begins to emerge when we consider the following examples. Suppose that an employer has just received an unsigned message from a very distraught employee threatening to quit. Seated alone in an office, the employer reads the message, puts it down, and considers it. A moment later, the employer picks up the message again and recognizes the author of the note by the curled edges of the paper. If we assume for a moment that the noise factors have remained constant, to what do we attribute the shift in meaning of the message from the first glance to the second glance? If we return to the expression, theri if M 2 > M 1 and if we assume that Ip has not changed then the new meaning has to have emerged from an increase in I --contextual, or secondary information. s 145

PAGE 157

But where did this information "come from?" It came about from the removal of other equivocation from the message. If this is the case, then our expression needs to include a variable for those sorts of equivoca-tion that are not noise. if EE-N = all EQUIVOCATION except NOISE then our full-fledged expression reads: = (I + I ) p s ( 3 3 ) In our example of. the employee's threat to resign, new meaning arose out of a reduction in non-noise equivoca-tion and the resultant increase in contextual informa-tion due to the removal of equivocation. Next, let us assume that a man is preparing to pro-pose matrimony to a woman over dinner in a busy and noisy restaurant. As he utters the words, a clumsy waiter nearby drops a tray of dirty dishes on the floor. Dis-tracted, the woman hears, "Will you bury me?" In a few moments when the restaurant quiets down, she hears him repeat the message, "Will you marry me?" Once again, meaning has shifted with time. This shift is reflected in a reduction of external noise (and 146

PAGE 158

possibly some internal noise as well) noise and an increase in content (primary) information. We may conclude from both of these examples that any removal of equivocation (including noise itself) .. will necessarily result in an increase in available infor-mation. We may also notice that as the removal of noise distractions increases and as the noise factor reduces closer and closer to zero, then the greater an impact the.reduction of equivocation becomes in the freeing up of new information. At this point, let us consider an entirely hypothetical case and recall the yogic and mystic claim. Assume that over time some receiver learns to attain and maintain a condition where external and internal noise factors drop to near absolute zero. Assume that over this same time frame, the receiver relentlessly removes vast volumes of equivocation from a message until mere traces of equivocation remain. Under these conditions, the denominator of the will approach zero. As it does so, the numerator approaches near infi-nite sizes. Since the message content in this case is a con-stant, then as the numerical value of the numerator approaches infinite proportions, both meaning and con-147

PAGE 159

textual information reflect these same dramatically larger proportions. Four interesting conclusions may be drawn: 1) The most adept receivers at processing information into meanings will be adept .at. disciplining their own attentiveness and concentration and at managing both external and internal noise factors. 2) To a receiver, any message' amount of information will vary. The variation in this amount of information is exclusively dependent upon how skilled the receiver is in managing all forms of equivocation and how much consideration is given to the processing of the information. 3) Any receiver Skilled in .the processing of information can "free up" and process more information from any message than could have been purposefully encoded into the message by the.source. 4) Within the context of a recursive inquiry, any highly equivocal message, problem, or consideration contains within it the potential to yield vast amounts of information. New information is created by two possible sets of circumstances: a) When new meanings emerge and all forms of equivocation are simultaneously held in check and 1 48

PAGE 160

b) When a reduction in any form of equivocation is maintained over time. Before moving ahead into the discussion of patterns, a few final comments are necessary. Goss's Information/NoiseRatio describes a relationship that occurs in an individual playing the role of a receiver. Building upon this Ratio, we have mostly reviewed circumstances in which the and the intentions of the source were ignored. With this in mind, there are a few problems to consider. When, for instance, we speak of "meaning," shifts in the "amount" of meaning will reflect any of the receiver's biases, prejudices, and errors.of perception. For instance, if an individual is inclined to interpret messages "literally," contextual (secondary) information inputs will be treated. as nonexistent. In doing this, the receiver literal1y drops a "ceiling" of limitations on the amount of meaning that may be attained from the message. Once content (primary) information has been processed into meaning, no additional meanings are possible because no additional meanings are soUght. Conversely, if an individual is inclined to create vast amounts of contextual (secondary) information in even the most innocuous and straightforward of messages, 149

PAGE 161

imaginary contextual information inputs can easily intrude into the individual's processing of information--inputs that do not even exist. This was referred to earlier as Hamlet's trap. Secondly, in the absence of a feedback loop with the source that suffices to keep the receiver on course or aligned with the intended meanings, the receiver is left with an altogether subjective judgmetit call about what is a "meaning." It is a frequent enough occurrence that finds an individual detoured into some cul-de-sac of thinking because an incorrect "meaning" was formed among the individual's basic assumptions. These errors of meaning eventually surface1 but the obvious question that arises is: How can an individual recognize errant meanings and fix them? Inevitably, any -receiver is forced make at least a few subjective assumptions throughout the processing of messages. In the absence of any definitive confirmation system with the is, in fact, the case in recursive receiver is forced to make even greater numbers of subjective assumptions. Whether these assumptions prove correct or incorrect, they are the foundations upon which the receiver creates meaningsG 150

PAGE 162

The next section introduces the idea of pattern. Patterns and the relationships that give rise to patterns are particularly useful in aiding the receiver to test the veracity or falsity of these assumptions and related meanings. Relational Meanings, Responses to Relational Meanings, and Patterns Throughout this thesis, we have spoken of two types of meaning. There are highly specific meanings derived from single message inputs, and there are factotum mean-ings which are derived from specific meanings or from other factotum meanings. These latter meanings we named relational meanings because they relate specific meanings or other relational meanings together. In the previous secti6ri, we derived a relationship (3.3, p. 146) from Goss's Information/Noise Ratio that shows the interplay between equivocation, information, and meaning. This relationship suggests that when large amounts of equivocation are present in a message and if that large amount of equivocation can be removed over time, then the amount of information content grows enormously. 151

PAGE 163

Our experiences tell us that we as individuals determine how much equivocation is present in any message through our perception of the message. If I am told to 11meet you at 8 o'clock tonight, .. then it is of no expedient value to me to perceive vast amounts of equivocation in the message. I will tend to perceive the equivocation as small--i.e., "where shall we meet?" My strategy will be to process this message as having a very specific meaning. My response to the message betrays how much equivocation I have perceived in the message and, inexorably, how I have processed the message's meaning. Whenever I respond to messages as containing minimal equivocation and quite specific meanihgs, then my response falls under the Level I category of Bateson's hierarchy. We have already seen that relational meanings contain relative amounts of equivocation larger than the amount of equivocatibn in the meanings from which they are derived. For instance, r may respond to the question "who are you?" by relating together specific meanings about my professional life: "I am a sixth grade teacher at East Middle School in Aurora, Colorado." I may be just as inclined to respond by relating speci-152

PAGE 164

fie meanings about my home life: "I am a happy husband and proud father of two neat children.11 Once again, my response to the message has revealed both my perception of the amount of equivocation in the original input and how I have processed meaning. In Bateson's hierarchy, would be Level II responses. However, I might have responded to the question 11who am I?" in a more philosophical context.:.:.-one that reveals I am wrestling with an extremely large amount of equivocation. I may respond by saying, "I honestly don't know. I am coritinually myself that same question." This kind of response not only suggests a recursive inquiry, but it also suggests that the response is of a different order than the more distinct Level II responses. We conclude from this latter example that I am seeking a relational meaning which I am deriving from a host of lower relational meanings that include my "professional" and "home life" relational meanings as.well as other Level II meanings. What all this suggests is that at some point, the derivation of relational meanings the removal of enough equivocation that the processing becomes recur-153

PAGE 165

sive and Level III meanings emerge. But at what point or under what conditions does this change occur? Consideration of this question will necessarily include consideration of our fundamental question: What does a recursive inquiry accomplish? In order to help distinguish the many levels of relational meanings, we will introduce the notion of pattern. Pattern will be defined as the response to the relationship that exists between any two or more meanings. It is tempting to confuse pattern with relational meanings but there is a clear distinction. Relational meanings are the results of certain information processing consttucts while patterns are associated with to these results. There is always a relationship between the response and the meaning, but the response is not the relational meaning itself. An example is in order. Tw6 good friends are sitting together over lunch. Tom asks Dick how work is going. Dick responds by saying that work is going great and that he just received.a nice bonus. Torn relates this information. to a larger meaning that Dick always tends to flaunt his accomplishments. In response, Torn notes that he, too, received a handsome bonus recently. Dick relates Torn's response to a larger meaning that 154

PAGE 166

Tom is competitive and generally does not like to be outdone. Dick responds by saying that his bonus came along with a plaque of commendation. Tom responds by saying that he received some additional perks along with the bonus. A larger pattern of boasting emerges. Both Tom and Dick's responses are patterns that may be termed "boastful." While their responses are similar, they each are reactions to a different relational meaning: Tom boasts because Dick flaurits his accomplishments while Dick boasts because Tom is competitive. Whether or not these reiational meanings are true or false is immaterial. The responses signify that each friend considers the relational meanings to be true. Now at another time, Tom and Dick may respond differently. While their normative pattern may be to boast, we may assume that there exists a time when Tom still holds fast to his understanding that Dick flaunts his accomplishments but too.weary to engage in a boasting contest. In this case, Tom is responding to a substantially different relational meaning: one emerging from the psycho-physical informati6n of discomfort in combination with the "flaunting" meanings normally associated with Dick. 155

PAGE 167

We may assume further that Dick will notice that Tom has not responded in his customary fashion. This new information somewhat contradicts his relational meaning that says Tom is competitive and does not like to be outdone. For Dick, a new relational meaning emerges from these two contradictory meanings and too, responds differently. "Are you okay?" he asks. This new response does not necessarily mean that Dick has abandoned his belief that Tom is competitive and does not like to be outdone. The new response merely means that Dick is responding to a different relational meaning. Because patterns are not relational meanings, we cannot automatically look at an individual's response pattern and conclude from what relational meaning or meanings the pattern emerged. But when patterns are compared against one another over time, the substance of relational meanings begins to come into better focus. We may say, for instance, that if a certain pattern becomes predictable in a given set of circumstances, then the individual is responding to a single relational meaning. Deviation from the pattern in the same set of circumstances necessarily suggests that a different relational meaning is being responded to. 156

PAGE 168

To understand the relational meaning behind the pattern, we must turn to the individual and rely upon his/her ability to communicate the substance of that relational meaning. Like Dick, we probably would also ask, "what is wrong?" So long as the individual can truthfully communicate the relational meaning, then we may say that pattern and meaning may be explicitly correlated, and that at least in this case, pattern and meaning may be considered Of course, any individual's ability to communicate truthfully a relational meaning is dependent .upon how little or how much equivocation remains in the rel.ational meaning, how adept the individual is in communicating, and how credible and trusbmrthy the individual is. When trust, credibility, and commuriication skill are high and equivocation is low, we may say that pattern is nearly synonymous with relational meaning. These individuals are "open books" or "w.ear their hearts on their sleeves. 11 Like the old Hindu saying, \o7hen one 1 s life perfectly reflects one's beliefs, then words become entirely redundant. More often than not, patterns and relational meanings are separated by a metaphorical "gap." While more of this will be discussed in a later section, it is 157

PAGE 169

important to note that the clarity or obscurity of a relational meaning does not affect whether a pattern emerges or not. What precise pattern emerges depends on how clear a relational meaning is, but even obscure relational meanings prompt pattern responses. In fact, we may say that the minimal requirement for pattern to exist is recognition of ielationship regardless of whether anything else is understood about the relationship. When the minimal requirements for pattern exist, then the metaphorical "gap" that separates pattern from relational meaning is at its widest. The recursive inquiry addresses the means by whith the equivocation causing this gap may be reduced. Before examining this further, it will be necessary to digress and discuss anothei characteristic of patterns. Our fictional Tom and Dick were described in relation to two different types of patterns. One type of pattern referred to an intrapersonal response to a relational meaning (e.g., "Tom. is competitive") Another type of pattern which we called "boasting" referred to the interplay of intrapersonal responses at an interpersonal level. This second type of pattern is formed by the interlocking of more basic patterns. 158

PAGE 170

If we summarize the larger "boasting" pattern, we might say that Tom and Dick were engaged in the production and sustenance of a pattern of escalation. The production and sustenance of this larger pattern was made possible by smaller constituent patterns that mirrored and amplified one another. One prompted another boast which, in turn, prompted another larger boast which, in its own turn, prompted a still larger boast, et cetera. In Naven (1937), Gregory Bateson identified this interlocked pattern as one of two archetypal patterns found naturally in a host of social interactions. Schis mogenesis is the term he used to name any escalatory archetypal patterns of interaction are mirror images of one another and form a larger pattern that escalates, this larger pattern is called symmetrical schismogenesis. Bateson found symmetrical archetypal patterns at all levels of social activity. They were found at the roots of addictions, psychopathologies, boasting matches, feuds, international arms races, and all forms of competition to name just a few of its elaborations. The other archetypal pattern of schismogenesis is called complementary schismogenesis and is character-1 59

PAGE 171

ized by constituent patterns that are contradistinctive by nature. Patterns of dominance are responded to by patterns of submission; exhibitionism is responded to by spectatorship; dependence promotes nurturance responses. Progressive escalation of complementary schismogenesis likewise produces exceedingly undesirable results. If a wife submits to her domineering husband, the pattern does not proceed with the husband de-escalating his dominance. The Opp6site occurs. With his wife's submission, the husband responds with more dominance. If unchecked, the wife's successive levels of submission prompt successively greater dominance and eventually reduce her to little better than a slave. Opposing dominance with her own expression of dominance will not this If the wife resists, the husband's escalation of dominance is often unexpectedly aggressive and may lead to unpredictable abuses. Bateson's obvious concern was to understand the conditions under which the escalations characteristic of schismogenesis patterns could be managed, or dissolved altogether. Our investigation has shown that the constituent patterns which make up these larger, archetypal patterns may be changed when new relational meanings replace older relational meanings. 160

PAGE 172

These new meanings in combination give rise to the larger pattern change. An overly simplistic example of this was given when Tom was too weary to respond to Dick's boast. To some degree, then, we all have control over our patterns of response. The degree of control we have our patterns of response is directly dependent upon the amount of sense we have made of our relational meanings. Since the amount of sense we make of our relational meanings is directly dependent upon how much equivocation has been removed from the relational meanings, the key to managing patterns of response is to manage levels of equivocation. Bateson's hierarchy was built upon responses to messages and informatLon and meahings . Since pattern has been defined as responses to relationships between information and meanings, we may reconstruct the hierarchy and Levels of Learning according to our concept of patterns and still maintain a consistency in the integrity of the hierarchy. This reconstructed hierrchy (Barnett, 1993) Sharpens the distinction Levels II and III. In light of patterns, Level II Learning may be summarized as the purposeful management of patterns. Level III Learning, on the other hand, 161

PAGE 173

may be summarized as the creation of new patterns by purposefully managing equivocation. Before we can understand this difference between the management of and creation of patterns, we need to understand a bit more about the combinatory properties of patterns and the constituent components of patterns. Moire Phenomena: The Combinatory Properties and Constituent Parts of Patterns Moire Metrology is the name given to the scientific study of what has widely become known as the Moire Pheno-menon. Moire Metrology concerns itself with patterns and the characteristic properties of patterns. This science includes the study of the interactions between and among pattetns, the combinirig of constituent patterns to form new and the relationships that occur in these interactions and combinations. While Moire Metrology has almost exclusively concerned itself with regulated physical patterns such as light and sound waves, it has already yielded interesting and surprising results that are germane to our consideration and defi-nition of pattern. Two illustrations will serve us at this point. 162

PAGE 174

Bateson himself provides us with the first illustration (1979, pp. 88-89). Let us-assume two different but humanly inaudible high frequency pitches (one"m," the other "n11). These two inaudible sounds patterns will combine to form a single complex sound pattern. If we could take a representative picture of this resultant pattern, we would see huge crests and troughs and calm spots among The resultant pattern, of course, bears little resemblance to its two constituent One oddity,-however, is that at a point m X n, a huge crest will occur. At these moments, the sound pattern may become audible to the human ear. A beat may be heard resulting from the interaction of two inaudible sound sources. We may draw some important conclusions from this illustration. It would be understandable that any human hearing the regular beats would mistakenly interpret the beats as a singular pattern. As we already know, the beats not the pattern. Rather, they are a tiny fragment of the actual pattern--the vast majority of which exists in an inaudible realm. Moire Metrology tells us that under a given set of conditions, any two or more patterns will combine to form only one unique resultant pattern. However, 1 63

PAGE 175

it is possible to combine a host of different sorts of patterns in different ways to form resultant patterns that are similar to a unique resultant pattern. Let's go back to our audible beep-beep-beep beat. If my task is to recreate this beat pattern exactly as it was formed and if I have mistakenly assumed that what I hear is the entirety of the pattern, then I may recreate the audible beat precisely by attaching an electronic device to a tiny speaker. By adjusting the electronic pulses and the audible I can create a perfect mimic of the beat. Obviously, I have not fulfilled my task, though. To a creature with more sensitive hearing, the complex sbunds of the combined human-inaudible patterns would bear little or no discernable relationshiP to the mundane sound of my electronic impulse/speaker device. As in the physical realm, human response patterns combine with other response patterns to form complex, often unpredictable resultant patterns. Our earlier discussion of the dominance-submission pattern is but one of many possible illustrations of this. As humans, we are highly sensitive to this tendency to unpredictability. We spoke earlier of the Level II Learner who when responding to a new meaning tends 164

PAGE 176

toward choosing among possible responses that one response which seems to be most advantageous to the individual. Now we may say that a Level II Learner might de-select certain responses because he/she simply does not know how the response will fit into the larger picture. In a context of patterns, we may say that a Level II Learner tends to select those response patterns which perception or experience suggests combine into more favorable resultant patterns even if the response patern is in contradiction to the new meaning which prompted the selection. But Moire Metrology has shown us that patterns can often be more complex than our senses can perceive. Moreover, can combine in ways more unpredictable than our experience can have encountered. This leaves the Level II Learner with a conundrum: Knowing that how I respond will the larger pattern in which I must live, how can I respond in a way consistent with my understanding and that maximizes my wellbeing in the larger pattern? This question, more or less, may encapsulize what Bateson referred to earlier as the "contraries generated at Level II" which drive a Learner to Level III. (In passing, it may be noted 1 65

PAGE 177

that this question might well be one of the more fundamental kinds of recursive Let us now turn to our second illustration concerning patterns. Since the time when Newton split a beam of white light with a prism, we have known that any beam of sunlight can be broken down into a spectrum of colors. Later, scientists learned that the spectrum band extends beyond the visible colors in the spectrum and includes infrared and ultraviolet "invisible" colors. Each representative of the spectrum was found to have its own unique wavelength pattern, and each representative was found to be invariably arranged in a hierarchical manner from right to left according to the frequency of the wavelength pattern. While at first it is tempting to conclude that the simple sunlight beam is of constituent patterns and is therefore like other patterns .that are composed of constituent patterns, a sunlight beam is not like other patterns in this way. First of all, we may say while the pattern of light is the result of the combination of its constituent patterns, we cannot say that this resultant pattern reflects the high degree of complexity characteristic of more common resultant patterns. If any-166

PAGE 178

thing, we may say that the pattern of light is an ultrasimplistic "summary" of a vast spectral continuum. This makes for an extraordinary attribute of certain patterns: As large numbers of constituent patterns are combined, the resultant pattern actually simplifies rather than makes more c6mplex their combined expression. The attribute of simplicity in mathematical and theoretical expression exemplifies. this attribute of resultant pattern. Newtori's Laws of Motion, for instance, are remarkably simple expressions of pattern that generalize over a vast host of constituent patterns. This is not to say that these sorts of resultant patterns--though simple in expression--are equally simple to derive or create. In fact, the converse is true. Because so many constituent patterns are involved, the combining of them involves the necessary removal of equally vast amounts of equivocation. We have already shown that the combining of specific meanings into relational meanings involves the removal of equivocation and that the combining of relational meanings into new relational meanings involves the removal of still larger amounts of equivocation. As large numbers of relational meanings are combined 167

PAGE 179

into mega-meanings, a commensurately large amount of equivocation will need to be removed. We may now say that the combining of large numbers of constituent patterns into a simpler resultant pattern will result in what we will call mega-patterns. Two other examples of mega-patterns that we have already encountered include symmetrical and complementary schismogenesis. Before leaving our of the light patern, another important attribute of its pattern needs to be covered. With sufficient heat, we may cause calcium, sodium, iron, and brass to give off light. We recognize these emissions as light patterns because they may be broken down into precisely the same constituent patterns arranged in precisely the same order as any form of light. However, calcium, sodium, iron, and brass spectra consist in slight variations in cerain constituent patterns. These internal constituent shifts do not change the emission into something other than light. Rather, we recognize that the integrity of the light mega-pattern remains intact but that the mega-pattern's constituent patterns have been shifted or transformed so that the mega-pattern can be used in a new and different way. Metaphorically, we may 168

PAGE 180

say that this shifted or transformed expression of any pattern or mega-pattern is the "creative" use of it in new circumstances. The Zen initiate cited throughout this thesis is, of course, seeking a mega-pattern. The telemarketing department described in the Preface also sought a mega-pattern as it engaged in the activity of finding out how to provide quality service. The recursive inquiry was introduced as an activity by which a telemarketing department or Zen initiate could find solutions to their highly equivocal concerns. Based upon our discussion to this point, we may conclude that the product of a truly recursive inquiry is the discovery of and application of meanings that form mega-patterns. This provides one answer to the question posed earlier: What does the recursive inquiry do? There is also another part to this question's ans-wer, and this will require a consideration of Peter Vision/Current Reality Gap (1990). Peter Vision/Current Reality Gap and the "Equivocation Gap" If we consider that the 9oal of any recursive inquiry is to derive mega-patterns, then we must say that the 169

PAGE 181

activity of any recursive inquiry is to remove the equivocation required in the formation of relational meanings that prompt mega-pattern responses. In order to understand more clearly what this means, we will turn to the Vision/Current Reality Gap--a concept that plays such a central role iri Peter Senge's The Fifth Discipline (1990). This Gap is what Senge invokes to describe the relative distance that exists between where an organization envisions itself to be and where it actually is. Senge argues that when this distance is large, a form of creative tension persists that strives to pull the organization toward the vision with energies and efforts in direct proportion to the distance of the Gap. Conversely, when the Gap is nonthreateningly small, then the vision for the organization is very close to the current status of the organization. In this latter case, there is little or no creative tension at play in drawing the status quo to the innocuous vision. Senge's point is that the Vision/Current Reality Gap encapsulizes the dynamics. which are characteristic of those vibrant organizations referred to in the current literature as "learning organizations" (cf., Argyris, 170

PAGE 182

& Schon, 1978). In Senge's own words, these learning organizations are where people continually expand their capacity to create the results they truly desire, where new and expansive patterns of thinking are nurtured, where collective aspiration is set free, and where people are continually learning how to learn (1990, p. 1 ). Senge introduces this Gap by having us imagine both of our hands inside a wide elastic band tightens as we spread our hands apart and that loosens as we bring our hands together. If we hold our left hand in a fixed position that represents the organization's status quo, then we may be free to move our right hand to variable positions that represent different sorts of visions for the organization. The further away from the fixed left hand our right hand moves, the more challenging the vision of the organization becomes. As the right hand moves further away to more challenging visions, the tension in the elastic band becomes commensurately stronger and one of two things happens. Either the left hand is drawn toward the more challenging vision and the status quo is changed, or if the left hand remains immovably fixed, then the right hand and its challenging vision is snapped backward to the staid status quo. 171

PAGE 183

The creative tension Senge refers to is, of course, represented by the elasticity of the band. This creative tension is the term Senge uses to summarize the "learning" activities that move the organization from the status quo toward the vision. The absence of creative tension in an organization signifies a fixation with the maintenance of the status quo against significant visionary changes. Given this synopsis,_ we may now begin a series of substitutions to show that the explication of the recursive inquiry in this thasis serves to describe how an individual or any collective of individuals moves from a status quo to a vision. Moreover, we may now show that Senge's Vision/Current Reality Gap is really an "Equivocation Gap'' and that it is the removal of equivocation that moves current reality toward vision. Finally, we may show that the recursive inquiry is a Level III Learning activity. Kenneth Ferguson's "Surprise Exam Paradox" (1991) provides an effective platform from which to build these relationships. Within the given conditions of the puzzle and the application of those conditions to a classroom setting, a contradiction arose suggesting that despite the instructor's desire to administer a surprise 172

PAGE 184

exam to students, it is quite impossible to actually give one. Within a context of pragmatic experience, however, it is easily imaginable that a surprise exam can be given. So the paradox emerges. Within our information processing structure, the "current reality" consists in the collision of two relational meanings that contradict one another. Now at this point, we may follow one of two courses: 1) We may ignore the paradox, or 2) we may seek to dissolve it. If we follow the first course, then within the Retention subsystem of our information processing system, the contradictory relational meanings will persist, and our response will probably be reflected through some pattern we have developed to counterbalance unimportant contradictions. We might, for instance, shrug our shoulders, tuck the curiosity away into a "trivia file", and retrieve it as a tidbit fit for use at some future cocktail party. If we follow the second course, then within the Retention subsystem, a new relational meaning will be sought that somehow recognizes the relationship between both paradoxical constituent meanings. Once again, the response will reflect the activity going on within the 173

PAGE 185

information processing system. As Ferguson relates the solution to the puzzle, he used an analytical process similar to that which we have named a recursive inquiry. From both Shannon and Weaver and Weick,.we have seen that the derivation of a new relational meaning requires the removal of equivocatio'n. In the case of paradox, the amount of equivocation to be removed is normally very high. Throughout the recursive inquiry's cycles of information processing, equivocation is continually being removed and new relational meanings are continually being forged. But the obvious questions at this point are: When do we know we have arrived at the new relational .meaning we want? or, more bluntly, When do we know we are done inquiring? Ferguson's description of his solution aids greatly in understanding the answers to these questions. It will be recalled that Weick stated equivocation needed to be registered before it could be removed. It will also be recalled Shannon and Weaver showed that we need some sort of comparison of messages in order to measure equivocation. With these notions in mindi a new relational meaning cannot be derived unless some minimal amount of information is known about it. Earlier, we mentioned that this minimal amount of information 174

PAGE 186

could be nothing more than a ''suspicion" that a relationship exists. If no relationship at all can be detected, then no equivocation can be measured and no relational meaning can be derived. In fact, by definition, where there is no relation to be detected, there can be no meaning. In Ferguson's case, he recognized that in order to dissolve the paradoxical tension between the two relational meanings, a new descriptive relationship had to be found that would allow both contradictory relational meanings to coexist without contradiction. Once having distinguished this restriction on the new meaning, Ferguson could commence with a recursive inquiry. In Senge's terms, Ferguson had created a "vision," and this vision is vital to keeping control over the registration and removal of equivocation throughout the recursive inquiry. Without detailing the minute steps Ferguson went through to resolve the paradox, we may summarize his study of the problem by saying that as he "discerned" new understandings of the problem, these newer meanings were compared against the desired relational meaning. This continued comparison is critical to any recursive inquiry because as we have found, the removal of any amount of equivocation will result in new meanings. If 175

PAGE 187

these new meanings are correctly derived, then they too will help to define what the desired meaning will be. In other words, if consideration of the two contradictory meanings yields a new relational meaning that is not the desired relational meaning, then we must know that the desired relational meaning must necessarily encompass all three relational meanings. Ferguson leveraged these discoveries to gain a clearer understanding of the problem. We might say that through this process of recursive "triangulation," he was inevitably led to suspect that definitions ought to be checked for consistency. Once this line of information gathering began, the shift in meaning of the term "surprise" was uncovered. In terms of the desired ne\o; relational meaning, Ferguson had discovered that the only way these two contradictory sets of conditions could coexist was to accommodate a shift in the meaning of "surprise." Once this desired relational meaning has been achieved, we may say that the Gap which initially separated the original conditions from the envisioned conditions has been closed. Furthermore, we should point out that the shrinkage or expansion of this Gap is com-176

PAGE 188

pletely determined by the removal or addition of equivo-cation in the meanings being processed. It is hoped that this illustration has served to tie together the recursive inquiry with the information processes that create meanings and that Senge's Vision/ Current Reality Gap has served_to simplify conceptualiza-tion of the recursive inquiry. A second illustration will be drawn in the final section of this chapter. This second example is expected to do two things: 1) tie the recursive inquiry together with the derivation of new relational meanings and the consequent expression of these new relational meanings in mega-patterns and .2) separate the_ recursive inquiry from Level II Learning and distinguish it as Level III Learning. Collaboration Activity at Learning Level II and as Recursive Level III The term collaboration as applied to decision-making processes concerns itself with two or more individuals engaged in communication that will result in a consensus decision or course of action (Eisenberg & Goodall, 1993, p. 271). In order to achieve this consensus, it is not necessary for the individuals to agree with the decision or course of action. Consensus through collaboration 177

PAGE 189

does mean that each individual feels comfortable that they have been heard and have had an equal opportunity to participate in the process. As Ed Schein describes, "If there is a clear alternative which most members subscribe to, and if those who oppose it feel they have had their chance to influence, then a consensus exists" (1969, p. 56). One of the premier advantages to collaborative decision-making is that members of the process feel more genuinely "bought into" the final decision or course of action. Within the context of our thesis, we may say that collaboration is a communication process that seeks to achieve.a shared relational From this shared relational meaning, we would expect that there would be a high degree of corresponding response patterns from the membership of the decision-making body. The members would tend toward a high of predictable and similar forms of interaction with the environment. One of the assumptions of collaboration is that the achievement of shared meaning (shared decision-making) will necessarily result in aligned and consistent patterns of response (shared courses of action) among the members. Relying on an assessment of our own prag-178

PAGE 190

matic experience, we recognize that there are problems with this assumption. Just because two individuals share a common understanding does not automatically mean that the two individuals will respond in a concerted manner. If fact, these two may entirely agree upon a relational meaning: My wife and I mutually enjoy quiet music in the background as we eat dinner. Our response patterns are different: She selects Barry Manilow, and I select Claude Debussy. Now collaboration may work the other way: We seek to unify courses of action (patterns of response). During jury deliberations, members seek to reach consensus on the guilt or innocence of the defendant. While on the surface this deliberation seems like a collaborative decision of a shared meaning, it is not. Juries do not seek unanimity on why an individual is guilty or innocent. Juries seek unanimity on whether an individual is guilty or innocent. Why one member determines that a defendant is guilty and why another member determines that the same defendant is guilty is immaterial. The problem, of course, to this sort of collaborative effort is that the relational meanings behind the common action are not necessarily aligned. In organi-179

PAGE 191

zations, misaligned motives for common efforts can have critical consequences. If these motives are self-promoting, for instance, the collaborative effort is based upon superficial consensus and not on wholehearted consensus. My and I mutually agree to listen to Barry Manilow because she wants to and because I do not wish to hassle over a petty decision. Have we collaborated? The last juror agrees with the others to a guilty verdict because he/she wishes to get home in time for the 5 o'clock News. Did the One of three legislators who together are preparing a piece of legislation agrees to delete language that weakens the law because the one needs the support of the other two and because the other two know that the language as it exists will disenfranchise important special interest groups of theirs. Is this collaboration? All three of these considerations of collaboration reveal the Level II trap that Bateson (1964) described and which we cited earlier in this thesis. Once again, this is the tendency for receivers to sever the response pattern from the fundamental relational meaning. It will be recalled that receivers sever this relationship normally because some motive or set of motives suggest 180

PAGE 192

that a different response patternwill be more beneficial in the long run. These motives are other relational meanings coexisting sideby-side with the fundamental relational meaning but not tied to it. These coexisting relational meanings give "purpose" to our choice of a different response pattern and "override" the direct response pattern of the original relational meaning. Let us reconsider the legislator. A relational meaning has prompted a piece of legislation. The desire of the legislator is to see the legislation passed. In order to achieve this desirable pattern, the legislator must alter the original response (the original piece of This altered response is inconsistent with the original relational meaning but is consistent with an overriding relational meaning. What this means is that the legislator has not prepared the best piece of legislation possible--at least according to his/her undertanding of the rieed. Instead, what has been prepared is a piece of legislation that has a greater likelihood of being passed. This helps to illustrate one of the problems of any cooperative decision-making process: Namely, how glorious visions of what could happen evaporate into more mundane realities. 1 81

PAGE 193

Is there an alternative to the pitfalls of Level II activity? At Level II, we have seen that individuals tend to ''choose" patterns of response in relation to meanings they have derived. This carries two serious consequences. when individuals a pattern of response, they sever the relationship between the fundamental relational meaning and the response. Sec ondly, when a pattern .is chosen, the individual tends to lose control of the pattern. In other words, the chosen pattern tends to combine with other response patterns in often unforeseen, highly unpredictable and uncontrollable ways. Both of these consequences are not matters of control of pattern so much as they are matters of control of equivocation. We may characterize Level II activity as that concerned with choosing meanings choosing patterns. Level III activity concerns itself with the management of equivocation. This kind of activity is concerned with creating meanings and creating patterns. To understand what this means, let us consider how collaboration works at Level III. First of all, we know from our Level II discussion of collaboration that it is of superlative interest to achieve common relational 182

PAGE 194

meanings. Secondly, we know from this same discussion that it is of superlative interest to maintain a common response pattern consistent with our common understanding of shared relational meanings. For collaboration to work at Level III, these two attributes must be considered simultaneously of equal importance. The first .step is to determine what constitutes the desired new relational meaning being sought by the collaboration. To our telemarketing department, this desired meaning was formed into a question: How do we improve the quality of our service by 10%? To the Zen initiate, the desired meaning being sought was the question: What descriptive conditions are necessary to generate the series of answers to the question, "What is the sound hand clapping?"? To Kenneth Ferguson, the desired meaning being sought was the question: What "surprise exam" conditions are necessary for both pardoxical sets of conditions to coexist without contradiction? This action coincides with the establishment of what Senge's calls the establishment of the vision. The creation of this relational meaning can and should include all of the critical relational meanings of the collaborating members. As we considered the quality question corporate-wide, members from the billing and 183

PAGE 195

information systems departments wanted to understand what it meant for them to improve the quality of service by 10%. So too did our CEO and CFO want to understand how they could improve the quality of their service by 10%. All of these considerations and concerns became constituent relational meanings that were to be included in our envisioned meaning. Identifying all of the constituent meanings was vital to our progress later. Obviously, identifying these constituents helped us to get a clearer understanding of the vision and what the envisioned new meaning had to encompass. Less obvious to us was that identifying these constituents aided in identifying the amount of equivocation we would later need to address. While .we never used the term "equivocation," we did realize that to achieve our vision, we needed to do "a great deal of work" to incorporate all of the constituent concerns in the achievement of our goal. This "great deal of work" has been shown to be the activity of removing equivocation and creating meanings. We may say that in general, the number of constituent relational meanings that must be included in the desired meaning will be directly proportional to amount of equivocation that will need to be removed in subsequent activity. 184

PAGE 196

Still less obvious to us at the time was that identifying constituent relational meanings increased the diversity and level of participation of the collaborating members. It will be recalled that when two or more individuals are involved in removing equivocation, the smallest amount of equivocation perceived in the meaning will form the boundary of how much equivocation can be removed by the group's effort. For us, we had a corporate-wide understanding that we were all up against a big chore. In our naivete at the time, we all mutually perceived that we had a large amount of "equivocation" to remove. This agreement is important to Level III collaboration and cannot be minimized. The collaborating group must deteimine how complex, far-reaching, difficult, time-consuming, and unclear the desired vision is. The decisions about these concerns determines how wide the Equivocation Gap is perceived to be. At this point, the first response pattern of Level III collaboration is to assess what is already known about the new meaning. This establishes the baseline of meanings and further identifies constituent relational meanings germane to the desired meaning. In our company's case, this was accomplished by sharing what we already knew about quality service. Normally, this sharing tends 185

PAGE 197

to add new relational meanings to the stock of shared meanings within the collaborative group. Also, it is quite normal for members within the group to create their first shared relational meanings. One member might well recognize a new meaning from constituent meanings shared by two others. Oft.en, this session becomes a real brainstorming activity. With these new baseline meanings in place, a Level III collaboration can easily disintegrate into a Level II activity. This happens when the new meanings are mistaken for the envisioned meaning, and the group members are encouraged to "go out there now and do 10% more quality service." This undermines the envisioned meaning and mistakes the response patterns for the mechanisms by which the envisioned meaning is achieved. This is the Level II trap. More properly, Level III control of the response pattern would be redirection of it to the removal of more equivocation from the Equivocation Gap. This is an important distinction because the redirection of the response to the removal of equivocation begins the recursion processes. In the this critical juncture was portrayed as a rephrasing of the original question: [Now that we know what we know about service,] 186

PAGE 198

How do we improve the quality of our service by 10%? Our mutual corporate energies were being focussed back toward the main learning endeavor. But why is this redirection important? It is important because the redirection of the response to the removal of more equivocaion keeps resonse patterns of the collaborating members aligned and in concert. While it may be tempting to turn the collaborating members loose on applying what ihey have learned, this subverts the goal, which is to create the vision. This scenario duplicates what we saw at Level II as members of the group shared a set of common meanings but responded in diverse and uncoordinated patterns. In short, members from making response choices that counteract the newly created meanings. In our company, redirection was a difficult thing to do. We were continually wanting to implement our new discoveries. When we did this, we had to restart the whole process over again. Worse, we found that some of our discoveries turned out to be busts, and we would have done better to wait until we were further along into the recursive inquiry process before implementing anything. 187

PAGE 199

When we did redirect, a curious thing happened. While we did not know it at the time, as we returned to our day-to-day activities, we not only consciously worked at learning about our envisioned meaning, but we subconsciously enacted our shared combine of relational meanings in the carryi,ng on of our conscious activities. In other words, when-Level III collaboration is in effect, the change that occurs due to new shared meanings becomes an inseparable of the pattern, or manner, in which the is continued. This was a crucial discovery about the recursive inquiry: It is not necessary to consciously implement new meanings; rather, the new relational meanings more or less implement themselves into the fabric of the continuing learning effort. But this is entirely a matter of faith. It is entirely a of faith in both the recursive inquiry process and the individuals conducting the inquiry. For our telemarketing department, this meant that our commitment to and utter belief in the learning process had to result in the desired outcomes. Because of our wholehearted commitment to understanding and learning about our vision, the desired vision must necessarily follow. This faith is a distinguishing differ-188

PAGE 200

ence between Level II and Level III activity, and this same faith is one of the distinguishing differences between choosing new meanings or patterns of response and creating new meanings and patterns. With the redirection response, the Level III collaboration becomes a recursive inquiry, and the recursive inquiry completes a single cycle. At the end of this cycle, the collaborating members have commonly-shared relational meanings and are commonly-engaged in processing messages and seeking new relational meanings. This unified activity registers and removes additional equivocation. With subsequent sharing sessions, still more equivocation is removed and newer meanings created. These newer relational meanings clarify the substance of the vision by defining new relationships contained by the envisioned meaning. With each succeeding recursive inquiry cycle, the body of shared meanings increases, the Equivocation Gap decreases, and the closer the collaboration effort comes to reaching its vision. At that hypothetical point when the Equivocation Gap reaches zero, then the collaboration effort will have shared in the derivation of the vision. When this desired meaning is redirected, then the effort will have "created" the vision. 189

PAGE 201

Whether this recursive inquiry process is conducted by an individual or by a collective of individuals, the resultant outcomes may very well be what Georg Simmel (1950), George Mead (1956), and Karl Weick (1969) were referring to when they suggested that the object of social activity is the "creation" of environment. This discussion brings to a close Chapter 3. It has been the purpose of this chapter to disclose what a recursive inquiry does and what may be gained by the use of a recursive inquiry. Furthermore, this disclosure caused us to consider the role of behavioral response patterns and some of the unusual attributes of patterns in general including the propensity for patterns to combine in novel and unpredictable ways. Finally, a description of collaboration as a coliective activity at Learning Levels II and III was considered. This helped to distinguish the differences between those two Levels as well as allowed us to place the recursive inquiry among those activities exclusive to Level III Learning. The final Chapter of this thesis will consider some of the heuristic possibilities further efforts on this theme may provide and will open with some concluding comments about the recursive inquiry. 1 90

PAGE 202

CHAPTER 4 CONCLUDING REMARKS AND HEURISTIC DIRECTIONS Five Speculative Footnotes to the Recursive Inquiry This thesis has a case for the recursive inquiry as a conducive means by which tq tackle particu-larly difficult, complex, and highly uncertain problems. The common attribute of these types of problems is the high levels of equivocation they possess. The recursive inquiry has also been presented as an interlocked set of processes which may be used in a context that features an intrapersonal, interpersonal, or a combined configu-ration that removes the shrouds of equivocation from obscure, information-dense messages and meanings. This presentatio-n has described what a recursive inquiry is, what a recursive inquiry consists in, what a recursive inquiry does, how a recursive inquiry works, and how a recursive inquiry compares to other information processing activities. What remains to be presented is how the recursive inquiry may be applied heuristically. This subject forms the largest part of Chapter 4. 191

PAGE 203

By way of introducing these heuristic matters, two observations and three cautions will be offered as footnotes to our considerations of the recursive inquiry. These are offered as footnotes because no empirical data has been gathered that either refutes or supports them. My own observations of them in recursive inquiry contexts, and my reflections upon them are solely the reasons for considering them. The first of the two observations concerns an individual's learning skill in such Learning Level III activities as we have presented them. These activities would include the creation of the environment, derivation of mega-meanings, innovative creativity, and the consideration of paradoxes and ontological matters. It may be difficult for us to imagine turning over a highly complex matter to an individual who does not possess a commensurately high level of education. Complexities are often mistaken by Level II thinking to require special training in abstract thinking. Abstract thinking, in turn, is considered to be closelyrelated to university training and, in particular, to advanced university training. In the Oxford English Dictionary (1971, p. 11 ), the word "abstract" is a Latin composite that means "drawn 192

PAGE 204

from, or separated." This means that a certain charac-teristic, form, function, or other attribute is drawn from or separated from its original context and used in a new or novel new context. Gregory Bateson uses the term "abduction" to capture this sense of abstrac-ion: We are so accustomed to the universe in which we live and to our puny methods of thinking about it that we can hardly see that it is, for example, surprising-that abduction is possible, that it is possible to describe some event or thing (e.g., a man shaving in a mirror) and then to look around the world for other cases to fit the same rules that we have devised for our description (1979, p. 157). This sense of abduction is not only an integrally impor-tant part of what we mean by but abduction is what the recursive inquiry The Zen initiate is searching for that 11description" of rules which define the series of answers to the koan. Our telemarketing department discerned a level of "description" for quality service that enclosed customer interactions as well as many interpersonal interactiOns. In both of these cases, abduction and abstraction are entirely possible within the scope of any individual regardless of education skill level. What seems to mat-ter more than education level is an individual's ability to focus concentration, to stay on task for long periods 193

PAGE 205

of time, and to be curious. Effective Level III collaboration happens more because of highly intense interest of the members than because of educational background. The second observation is that self-aggrandizement short circuits or undermines the recursive inquiry. Specialized interests, ulterior motives, "hidden agendas," and self promotion may be conceptualized as relational meanings that are surreptitiously added to the mix from which the desired relational .meaning is to be derived. Whether or not these are undisclosed meanings (and they usually are), these relational meanings are treated as though they are actually within the mix. Whereas the recursive inquiry subsists on shared meaningsand the shared "discovery" of meanings, selfcentered meanings often pose contrary or contradictory meanings which within our discussion add new levels of equivocation . If these new levelsof equivocation remain undisclosed in interpersonal settings and unrecognized in intrapersonal settings, then unregistered equivocation will continue to shroud higher level relational meanings. It is that interpersonal group collaborations, undisclosed self interests within the context of our considerations may help to describe those kinds of pathological group dynamics as "group think" and other 194

PAGE 206

"defensive routines" as identified by Chris Argyris (1985). The first of the three cautions concerns issues of morality and ethics surrounding recursive inquiry activity. In particular, higher levels of relational meanings result in what we called mega-patterns of response. We must ourselves the questions related to how beneficent or malevolent is the mega-pattern being sought. After all, the case could be argued that both Mahatma Gandhi's "Satyagraha" (nonviolence movement) and Adolf Hitler's "Lebensraum" (the Ary"an cultural and agricultural expansion) were laudableexamples in principle of a desirable Level III In the actual pattern of response, the former movement was instrumental in securing independence for India. The latter movement, of course, led to the enactment of the Holocaust and the patterns of policies so deeply associated with the ugliness. of the Third Reich. How can one tell the difference between beneficent and malevolent Level III activity? We would expect that since the patterns of response reflect the relational meanings being derived and subsequently enacted, observation of these response patterns over time should reveal some sorts of consistencies that should help us perceive 195

PAGE 207

the relational meanings giving rise to these patterns. With this in mind, we may speculate at five measuring rods for gauging the general trends of the Level III inquiry: 1) More beneficent outcomes should bring about a stronger sense of unity of purpose among the members of the recursive inquiry and should be revealed through noncontradictory response patterns even though the tendency of the inquiry is to incorporate more diverse message inputs and more diverse rela tional meanings. 2) More beneficent outcomes shpuld encourage higher levels of disclosure and trust among members of the inquiry. Both of these response are essential to the removal of more difficult levels of equivocation. Additionally, a growing trust and confidence in the recursive inquiry system should emerge. 3) Beneficent outcomes should also be reflected in an increasing number of adherents and dependents of the adherents who speak of the enacted change with across-the-board consistency. We might say that the feedback from these third party dependents ought to be reliable gauges by which to compare 196

PAGE 208

what the members of the inquiry say is happening and what actually is happening. 4) Beneficent outcomes should bring about a trend to tolerate and include more diverse and critical opinions as more equivocal meanings are being derived. This tendency reflects Weick's recognition (1969) that as more equivocation is being registered, the more necessary it is to incorporate random elements. 5) Beneficent outcomes should bring about more and more unforeseen applications of the mega-pattern. The nature of mega-patterns is the "discovery" of broader applications and more unified meanings than originally perceived. It would also be expected that as new applications are envisioned, the megapattern should temporarily recede in clarity (the Equivocation Gap has grown once again with the inclusion of a new relational meaning and, hence, more e:quivocation) until equivocation removal pro-. cesses once again reduce the Gap. With mention of the Equivocation Gap, a second caution is one named first by Peter Senge (1990) in his discussion of the Vision/Current Reality Gap. During those times when the gap between the Vision and the Current Reality grows, it is a natural Level II temptation 1 97

PAGE 209

to lower the specifications of the desired relational meaning. Stated another way, it often appears safer, easier, and less risky to stick with the more familiar Level II activities and allow Level II patterns to determine future patterns than to trust in unproven Level III processes, relational meanings, and patterns. As noted before, falling into this temptation will undermine any further progress in Level III activity. The final. caution is that there will be human casualties in a conversion to Level III activity. Jan Carlzon (1987) describes what may be construed as a Level III recursive inquiry activity in his autobiography, Moments of Truth. The organizational context in which this activity occurs is the decentralization of Scandinavian Airlines Systems (SAS) through the inversion of the corporate structure, the redefinition of company's purpose and market base, and the normal day-to-day focus of each individual employee's functions. Pervading this story is both the triumph of change over an ineffectual status quo and like any history of human triumph, the tragedy of those individuals who were lost in the change. In Level III activity, as mentioned earlier, certain individual attributes are necessary for an individual to participate. In addition to those previously ennum-1 98

PAGE 210

erated, we learn from Jan Carlzon that resistance to change and faintheartedness in the face of uncertainties and risk are lethal. With the closure of this introductory section, we will explore of the applications and questions that additional research into the recursive inquiry shed light on. The first of these general surveys will be in four areas of Communication Theory: 1) Constructionism, 2) Social Cognition, 3) Transformative Leadership, and 4) Bateson's Logical Categories of Learning and Communication. Thereafter, we will consider applications and questions that apply to Intrapersonal and Interpersonal Communication. This thesis will conclude with a few and questions that apply to the matters of patterns and mega-patterns, Organizational Learning, and Elaborations into Current Communication Theory This thesis has drawn heavily upon the work of Shannon and Weaver (1949), Karl Weick (1969, 1979), Greg ory Bateson (1964) and the Theory of Logical Types (Whitehead & Russell, well as other theoretical traditions. Some of these traditions, particularly those of Social Interactionism, Cybernetics, and Systems Theory 199

PAGE 211

have direct kinship to the theme of this work because they directly impacted the development and evolution of the thinking behind the main works cited above. While the work of Shannon and Weaver, Weick, and Bateson have dominated the explication of this work on the recursive inquiry, this is not to suggest that other theoretical traditions do not have important things to say about the recursive inquiry. Rather, my investigation into the recursive inquiry brought me face-to-face with four rather formidable questions. These three theoretical traditions were useful in answering these questions. The first question I encountered was: What makes certain messages highly obscure and others straightforwardly clear? (This, of course, turned out to be a recursive question.) Shannon and Weaver's Mathematical Theory of Communication provided a distinctively clear definition for equivocation and provided clues as to the distinctions between ambiguity (desirable uncertainty) and equivocation (undesirable uncertainty). the MTC provided very useful definitions for the central matters associated with the exchange of information. The second question I encountered was: How does an individual make sense out of obscure messages? Karl 200

PAGE 212

Weick's Theory of Social Organizing provided a particularly useful model for explaining how messages are turned into information, how information is converted into meanings, and how meanings become behaviors. Pivotal to Weick's Theory is the removal of equivocation to achieve meaning and this seemed to tie back nicely to Shannon and Weaver. The third question I had to answer was: What makes a recursive inquiry different from an ordinary inquiry? One consequence I knew was that a larger amount of learning resulted from a inquiry so I sought a description for learning. Bateson's Learning Hierarchy filled the bill and also related behaviors (responses) to learning. This hierarchy had the additional value of dovetailing nicely into Weick's work. The final question I faced was: How are new meanings and change related? Bateson's interest in patterns and l'1oire Phenomena prompted my own. investigation into the application of Moire Theory and the formation of patterns from constituent patterns. In a synthesis of these four descriptions, I found a reasonably comprehensive means of giving substance to many of the attributes of the recursive inquiry. However, as I considered the recursive inquiry further, 201

PAGE 213

I perceived it cuts a wider swath across Communication Theory in general than I could have anticipated. There are other attributes of the recursive inquiry which constraints of time and space prohibited me from pursuing in this thesis. However, in the interest of providing additional heuristic directions to those who may be interested in understanding more of the recursive inquiry, I would like to touch upon some of these theoretical relationships. Constructionism Since the recursive inquiry involves the derivation of relational meanings and since these relational meanings tend to classify the constituent meanjngs from which they are much of the language used to describe the recursive inquiry links it with Constructionist Theory. For instance, according to Kenneth Gergen, one of the four ,fundamental assumptions of social constructionism is that "reality is socially constructed by interconnected patterns of communication behavior" (1985, p. 266). This coincides with the description we have used to explain how response patterns create the environment and simultaneously cause new inputs to enter the the information processing system. 202

PAGE 214

Of greatest relevance, though, may be the elaboration of the Retention Subsystem's derivation of increasingly general relational meanings and constructs of reality. This matter was only lightly touched upon in the thesis. As relational meanings grow more general, these meanings tend to grow fewer in number as they incorporate ever larger numbers of constituent meanings. As relational meanings are synthesized together, then more hierarchically general meanings are derived. These highly generalized meanings integrate vast numbers of other meanings and tend to "order" or "organize" one's reality. What this points to is that successful Level III recursive inquiries tend to .make for an environment composed of highly cogent meanings. It would be suggest that our discussion of the recursiveinquiry fits neatly into Constructionist thinking. For instance, our description of the recursive inquiry and the Hierarchy of Learning in which it fits seems to suggest that at Level II, the constructs of reality are experience-dominated and conditional. At Level III, the recursive inquiry suggests that reality is not constructed out of experience but is actually created. 203

PAGE 215

Social Cognition Since Social Cognition "concerns how people think about the social world and how they think they think about the social world" (Fiske & Taylor, 1984, p. 17), our discussion of the recursive inquiry concerns these interests. In particular, work in understanding how contexts in which messages are sent will impact the derivation of relati.onal meanings. This matter of contexts of messages was not extensively covered. Message contexts, however, have a causal facet to them. Whether or not an information input is interpreted differently at two separate times is entirely dependent upon whether or not the context in which that information has been sent has shifted. Contextual matters have two important impacts on our discussion. The causal and conditional characteristics of contexts will influence how much or how little equivocation is perceived and will influence how an individual responds to meanings. It would be expected that Social Cognition would provide valuable insights into the relationships between and among message contexts, the registration (and subsequent removal) of equivoca204

PAGE 216

tion, the derivations of relational meanings, and the subsequent patterns of response. Transformative Leadership We have discussed the recursive inquiry as though it may be present or absent depending entirely upon the skill or desire of the individual or individuals involved in the communication transaction. This is somewhat true in the case of an intrapersonal recursive inquiry. In a group or organ{zational context, though, the presence or absence of the recursive inquiry is mostly dependent upon the leader facilitating the activity. The leader desiring Level III activity and outcomes must be cut from a special kind of cloth. Peter Senge (1990) notes that an organization committed to superlative forms of learning exhibits an attribute which he has come to refer to as metanoia, or a "transcendent shift or movement of the mind" (i990, p. 13}. In this particular case, Senge is referring to an organization consisting of individuals learning to find connections and interconnections with the world rather than individuals looking at themselves as separated from the world. 205

PAGE 217

According to Senge, "through learning we reperceive the world and our relationship to it" (1990, p. 14). Also, the learning organization is one "that is continu-ally expanding its capacity to create its future" (p. 14). The processes we have shown to compose the recursive inquiry fits what Senge has described as the shift in the processes of learning. Obviously, the leader of such an effort cannot help but be of a different sort than that leader who runs a more traditional organization. In Leaders (1985), Warren Bennis and Burt Nanus have identified this new sort of leader a "transformative leader." The role and function of the transformative leader as a necessary part of the recursive inquiry process needs to be drawn in more elaborate detail. Logical Categories of .Learning and Communication Finally, absolutely no work has been done to establish an internally ordered hierarchical structure within any of the first four levels (assuming that Level IV by definition cannot be hierarchically organized internally) of Bateson's Learning Levels. It has been implied throughout this thesis that each of the first 206

PAGE 218

four levels has contained within it an internally consistent sub-hierarchical ordering. At Level III, for example, we spoke of an "envisioned relational meaning" that we assumed was of a hierarchically higher order than the constituent relational meanings were. How ought these meanings be classified? Such an ordering, by definition, ought to also order the responses to these meanings. Through such an ordering, we.might begin to draw relationships between patterns and meanings. Additionally, we might begin to answer the following three questions: 1) What is the range of learning and communication complexity represented within each Level? 2) What are the differences in the descriptions of hierarchically adjacent Levels (i.e., what are the fundamental differences between the Level I logical type and the lowest Level II logical type?) 3) Are there "inherently hidden" levels within Levels? For instance, biochemical/genetic Level Zero meanings and patterns were illustrated earlier as occurring between DNA sources and nucleic protein receivers. But in a crab, for instance, what constitutes the message sequence and the relational meaning that causes a claw pattern and not a leg pattern to occur? 207

PAGE 219

or what causes a 11bottom side11 pattern of a claw to be expressed in its proper location and not two bottom sides? or what message sequence and rela-tiona! meaning causes the crab's top anterior and posterior patterns to be symmetrical from right to left but asymmetrical from anterior to posterior? Do these types of meanings and patterns suggest that extremely high degrees of purposiveness can exist within Level Zero structures or that higher Levels may within lower Levels and are nested within or have nested within them other Levels? Elaborations into Intrapersonal and Interpersonal Communication Throughout this thesis, examples of the recursive inquiry have been provided that position it in both intra-and inter-personal communication settings. No effort has been made to draw distinctions between these two communication contexts as they relate to the recursive inquiry. The basic process of the recursive inquiry is no different for the Zen initiate contemplating a koan quietly in a monastic cell or the telemarketing department or the husband and wife struggling to make sense of their daily routines. 208

PAGE 220

Also, by having defined what the recursive inquiry is, what it does, and how it works, we have by definition excluded a host of intra-and inter-personal communication transactions from what we have called Level III Learning. We have assumed all along that when appropriate and possible, Level III activity is superior to Level II activity when it to problem solving and With this in mind, the recursive inquiry has many possible applications to intra-and inter-personal communication issues. In the next two sections, it is hoped that a short discussion of some of these issues will spawn thoughts about and applications of the recursive inquiry in these two critical fields of study. Intrapersonal Communication Since the major outcome of the recursive inquiry is making sense of meanings by seeking relational meanings, the use of the recursive inquiry in an intrapersonal context is to make sense of the self and to make sense of the self in conjunction with the surrounding environment. Many of the operations attributed to the recursive inquiry are to be found among the attributes identified in the Intrapersonal Communiction literature (Cunningham, 1991 ). Among these are the transactions 209

PAGE 221

with the surrounding environment through perceptions which form the grist of the messages and information which are attended to by processes that convert information into meanings or that convert meanings into new meanings (Cunningham, 1991, pp. 600-601). Consequently, we may liken the recursive inquiry to a special type of "reflective thinking." For many of us, reflective thinking is a passive activity during which we review theevents of the day, daydream, chastise ourselves for faux pas, reconsider the manner in which we handled matters, or simply monitor the flow of our streams of consciousness. In many profeSsional contexts, reflective thinking is viewed as a waste of time. Reflective thinking takes time away from doing somethin_g "productive." If we tie the recursive inquiry to reflective thinking, we are discussing an aggressive process--not a passive activity. Moreover, the recursive inquiry is a highly controlled, integrated process quite opposite to the random, uncontrolled intrusion of thoughts that occur during daydreaming or reviewing the stream of consciousness. Frankly, most. individuals are unskilled and untutored in how to conduct a reflective recursive inquiry (cf., Johnson & Hackman, 1994)o 210

PAGE 222

Perhaps the biggest obstacle facing an individual who is seeking to find answers to highly equivocal questions is not so much how to conduct a recursive inquiry but how to discipline the mind to concentrate on the problem. Related to this obstacle of controlling internal "chatter" is that of external noise control. Unless these disciplines are taught, an individual will find that a recursive inquiry is an which more equivocation is being added to the problem than is being removed.-It will be an exercise in confusion. Additional investigation into these concerns may provide valuable insights into how individuals learn about themselves and how they might attain a greater sense of internal consistency and harmony. In The Psychology of Invention in the Mathematical Field (1945), mathematician Jacques Hadamard presents a four-step conceptual model showing how creativity occurs over a variety of diverse endeavors. Derived from records left by composers, artists, poets, scientists, and mathematicians, Hadamard describes the display of creativity as involving: 1 ) Deliberate preparatory conscious work, 2) Incubation of the work in the subconscious mind, 3) Illumination of the work, and 4) Verifying and Precising the work. 211

PAGE 223

These four steps of the creative process can be tied into patterns of response that an individual displays while conducting a reflective recursive inquiry. This is not to say that all creativity is the result. of Level III-type activity. Undoubtedly, many forms of what we culturally acknowledge as creativity is not Level III work, but certainly those extraordinary instances of 11creative discovery11 are probably instances of recursive inquiry Additional investigation of the relationships between the recursive inquiry and creative expression could yield important clues as to how creative discoveries occur and how individuals can be trained to maximize the impact of their creative potential (Johnson & Hackman, 1994). Finally, another intriguing line of investigation would be found in what Bateson (1964) identified as the life-transforming events that cause an individual to jump from Level II to Level III Learning. Cited earlier (p. 127), Bateson suggested that these sorts of sequences of events resulted in a 11profound reorganization of char acter" (1964, p. 301). He illustrated these kinds of events with certain psychotherapeutic breakthroughs and religious conversions. 212

PAGE 224

Investigation into this issue would shed valuable light on both what constitutes the "profoundly reorganized character" and how profoundly traumatized individuals may be assisted in making healthy sense of their extraordinary circumstances. Interpersonal Communication There are many possible applications here. The recursive inquiry is almost a natural strategy for inclusion in group problem-solving contexts. Much of what has been illustrated in this thesis emphasizes applications to the improvement of interpersonal communication. One of the.more beneficial applications, however, concerns how the recursive iriquiry may aid in conceptualization and management of interpersonal conflict. To introduce this application by way of I recall a time when both my wife and I were embroiled in conflict. While I do not recall what event or events prompted us to defer from our traditional habits of managing conflict, I do distinctly recall that in a marvelous moment, both of us simultaneously recognized that the other spouse had a very good point. 213

PAGE 225

This state of affairs presented us with a dilemma: We both had conflicting, but good points to make on the matter. What did this mean? We spent the largest portion of the evening thereafter trying to make sense of our points. We were, in short, discovering a "shared" meaning that was necessarily larger than either of the meanings we individually held and that would effectively change the shape of how we would respond to this issue in the future. The normal characteristics of our conflicts gave way to an excitement of discovering something new together. Perhaps the most intriguing facet of this incident was the heightened sense of interpersonal intimacy that comes about when personal agendas are set aside and two individuals engage in a common effort to discover something that neither one knows in advance. As we talked about it later, we both agreed that for this unique time frame, we both sensed that we were involved in a common activity that made us feel like our individual best interests were best served by identifying our common interests. If we assume that a great if not all of interpersonal conflicts consist in the perception of irreconcilable meanings, then management of conflict ought to include strategies for reconciling dissonant meanings. 214

PAGE 226

Since we have seen that the recursive inquiry seeks to achieve relational meanings, we ought to assume that the recursive inquiry has direct application to interpersonal conflict. The matter for investigation, of course, centers around how one gets from a condition of conflict to a condition of shared reconciliation. Because of its nature to seek shared meanings, the recursive inquiry could have interesting impacts on traditionally antagonist-based activj.tjes as the preparation and passage of legislation in a two-party political system, labor/management, corporate acquisition, and international and, in the courtroom, both on the preparation and presentation of legal cases and the deliberations of the jury. Elaborations into Pattern and Mega-patterns It was in Naven (1937) that Gregory Bateson first introduced the archetypal pattern of and two of its offspring: symmetrial andcomplementary schismogenesis. It will be recalled that schismogenesis is the term used to describe any pattern of escalation. That is, a pattern combines with a like pattern. to produce a larger, more elaborate pattern of itself. 215

PAGE 227

Symmetrical is generated when one pattern (e.g., anger) generates a response of the same pattern (i.e., anger), which is likely to generate a larger expression of the previous two (i.e., more anger). Complementary is generated when one pattern generates a response of an oppositefitting pattern. This complementary pattern generates expression of the original pattern and so forth. Margaret Mead (1979) identified some of these complementary patterns as nurturance/dependence, dominance/submission, and spectatorship/exhibitionism. Assuming that human behavior patterns.combine in a manner similar to patterns found in the physical world, we may gain important about general human behavior by studying these patterns and the manners in which they combine and express themselves. Both Bateson and Mead have shown us behavioral patterns can be employed in describing cross-cultural anomalies .of human behavior. With this in mind, generalizable descriptions for understanding how certain human behavior patterns are likely to combine and how these new combined patterns are likely to perform in future riontexts could be gained. (e.g., How does a pattern of self-promotion lead to a pattern of unscrupulous use of others?) 216

PAGE 228

In a more direct application of the recursive inquiry to patterns, addiction may be defined as a pattern of schismogenesis. Regardless of the type of addiction, the dismantling of the addiction pattern requires a shift in the relational meanings that are causing the addiction response. The recursive inquiry would be such a process for accomplishing this of causal meanings. In fact, we may go so far as to say that unless the relational meanings contributing to the addiction pattern are transformed and new relational meanings are derived, the addictiori pattern cannot be broken. In our discussion of patterns and combinations of patterns, we ought to expect that within larger social considerations, large social patterns should be discernable. The identification of what we have called mega has been demonstrated by John Naisbitt in his book Megatrends (1982). Traditionally, these general patterns somehow rise to the surface of public awareness and become matters of concern for the corporate, legislative, judicial, or educational institutions. In many cases, these mega-patterns, like the decay of the inner city and the flight of its inhaoitants to the suburbs, are formed from exceedingly complex and variable constituent patterns. Of concern is the methods 217

PAGE 229

by which these mega-patterns are addressed. While the recursive inquiry seems adept at handling highly complex and equivocal concerns, can some transform of the recursive inquiry be used at the societal level to shape an undesirable mega-pattern into a desirable mega-pattern? To illustrate, let us consider the body of Federal Assistance programs. Government assistance in areas of social concern has tended to become tax-generated financial support. In a general sense, Federal financial support through a financial assistance program tends not to solve the problem it was intended to address. In fact, the reverse seems to emerge,as a pattern. Take. the War onPoverty, for instance. Subsidies since the 1960s have increasedand so too has the of poverty. We sense a pattern bf complementary schismogenesis. Financial support _generates additional poverty which generates financial support and the associated bureaucracy to administer the support. This additional support generates additional poverty and so forth. This pattern is reflective. in other arenas. Efforts to reduce the Federal Deficit through Federal financing have increased the Deficit. Increased financial support for schools has led to lower and lower overall performance scores. Federal programs to reduce unwanted pregnancies 218

PAGE 230

have led to unwanted pregnancies. Postal Service increases have led to poorer overall postal service performance and reliability. Increased subsidies for the War on Drugs have led to an increase in illicit drug trafficking. Considering such mega-patterns, an investigation into these equivocation-dense messages might begin to show how the recursive inquiry and other as yet unconceived Level III activities may transform or help us gain control over undesirable mega-patterns. Elaborations into Organizational Learning Strides are being made in Level III meanings and patterns already in the world of business. Just a bare few of these have been cited. These strides are largely due to intense stresses to perform in a competitive arena where survival is difficult. Those that do survive will be aided by mastery in the ability to identify, embrace, and enact continual change. The business realm is also vibrantly active with ideas--another outgrowth of competition. Organizations are beginning to discover that they possess the skills, expertises, and resources to educate themselves. As more organizations move toward a learning environment, 219

PAGE 231

the business world may soon rival education institutions as the major provider of employment skills to high school graduates. Already many businesses are assuming the responsibilities to re-educate their employees in the latest scientific research and in the development of busines-related, social/interactional skills. In some instances, the business world seems to have succeeded in solving local problems and concerns eluding the solutions of legislatures and social services. Despite the failure of "quality circles" to take hold, newer innovations--the greatest of which may be called "dialectic collaboration"--are providing the impetus for business to confront highly complex problems eluding other organizations. Still, the major obstacle facing the business realm is Time. Rigid stresses and requirements mandated by deadlines limit much Level III activity. Still, the business and organizational setting seems a specifically useful place to inVestigate the possibilities of the recursive inquiry. Elaborations into Metaphysics Perhaps the richest source of Level III study may be found in arenas growing less accessible to communication scholars. Bateson, an avowed atheist, bemoaned 220

PAGE 232

the tendency of contemporary science to disenfranchise itself from consideration of religious and sacred issues and phenomena. Co-authored with his daughter, Mary Cath-erine Bateson, and published posthumously, Angels Fear (1988) explores the possibilities of an "epistemology of the sacred" which attempts to confront matters of concern that simultaneously confound the mind and soul. He had become aware gradually that the unity of nature he had affirlned in Mind and Nature might only be comprehensible through the kind of metaphors familiar from religion (Bateson & Bateson, p. 2). In the process of investigating the differences posed by Level II and Level III meanings and patterns, it has been reiterated that while an individual may be proficient and skillful with Level II processes, an individual does not automatically 11graduate" into Level III processes. There is:not a guarantee either that desire alone permits a person to "scale a ladder" to Level III. We know very little at this point about what conditions are required for an individual to transcend from Level II to Level III. Remembering that Bateson himself suggested that Level III was accomplished by a profound reorganization of the character, we would that Level III patterns involve some dislocation or displacement of Level II 221

PAGE 233

patterns. As we search for expressions of these kinds of Level III patterns, we would be hard pressed to find clearer examples of the recursive inquiry than to consider the four Gospels of the New Testament. From four quite different vantage points, we have a glimpse of four quite different individuals struggling with the same recursive question: Who is this Man, Jesus Christ? Earlier, reference was made to the control of the wild and unpredictable mind as a crucial skill to understanding of the self through the recursive inquiry. Hindu and Zen masters have wrestled with this very matter for millenia and have discerned techniques for teaching such a skill. masters of the art have documented to have gained control over autonomic physical functions--some of stopping the noise of the heart. Under these conditions, we can begin to imagine the circumstances necessary to test the mathematical expression we derived: When the intrusion of noise and the removal of equivocation approaches zero, then information and meanings become vastly large. In a remarkable way, the recursive inquiry and the aeons-old religious inquiry seem to bear many resemblances. It would be unfortunate to disregard as irrelevant or as superstitious a millennia-old tradition that has 222

PAGE 234

distinguished the relevance of calm, meditative, contemplative, prayerful attitudes to the conduct of spiritual inquiry. While other relationships between metaphysics and themes within this thesis may be drawn, the primary point is that from an epistemological standpoint, confirmation or nonconfirmation of its content and conclusions can be and ought to be tested against metaphysical phenomenon as well as against other contexts of phenomena. Concluding Remarks The recursive inquiry has been introduced in this thesis as a type of communication activity that allows an individual or group of-individuals to make sense out of highly equivocal, complex, or confusing messages and meanings. This activity has been conceptualized as a set of three interlocked processes that admit externally transmitted messages, decode those messages into information, derive meaning from the relationships between information inputs, and respond to those meanings through behavioral patterns. These processes consist in the registration of equivocation and its subsequent removal. It has been shown that the removal of equivocation results in information or, in the case of removal of 223

PAGE 235

equivocation from the relation of information inputs with one another, meaning. This exploration into the recursive inquiry has actually been an exploration into making sense of communication. Much of our communication thinking implies that the sender of messages is liable and responsible for how much or how little understanding results from communication transactions. By :i,mplication,. this places the receiver in the more role. It is a premise of this thesis that the receiver can determine the amount of information and understanding that wiil result from the communication transaction. We-may even go so far as to say-that information understanding can be reached independently of interaction with the message source. In the realm of the recursive inquiry, then, the receiver becomes of fundamental concern. The receiver is the one who distinguishes what is to be communicated, to be disclosed, and to be understood. It is the intent of this thesis to encourage further investigation into what can be known by a receiver and what skills of communication are required for an adept receiver or group of receivers to make sense out of the most equivocal of messages. 224

PAGE 236

REFERENCES Argyris, (1985). Strategy, change and defensive routines. Marshfield, MA: Pitman Publishing. Argyris, c., & Schon, D. A. (1978). Organizational learning: A theory of action perspective. Reading, MA: Addison-Wesley Publishing Co. Aristotle. (1932). The rhetoric of Aristotle. (L. Cooper, Trans.) New York: Appleton-Century-Crofts, Inc. Ashby, W. R. (1956). An introduction to cybernetics New York: Wiley. Barnett, c. S. A patterns-based hierarchy of learning: A reconstruction of Bateson's logical categories of learning and communication. Unpublished. Presentation, Colorado Speech Communication Association Convention, April, 1994. Barnett, c. s., Harrington, D., & O'Neill, K. The existence, role, and function of recursive questions in hierarchical learning: An exploratory study. Unpublished. Presentation1 Colorado Speech Association Convention, April, Bateson, G. Naven: A survey of the problems suggested by a composite picture of the culture of a New Guinea tribe drawn.from three points of view. Cambridge: Cambridge University Press. Bateson, G. (1964). The logical categories of learning and communication. Reprinted in Steps to an ecology of mind (1972) New York: Ballantine Books. Bateson, G. (1966). Problems in cetacean and other mammalian communication. Whales, dolphins and porpoises (1966) 569-799, ed. by Norris, K. S. Berkeley, CA: University of California Press. 225

PAGE 237

Bateson, G. (1969). Double bind, 1969. Reprinted in Steps to an ecologyof mind (1972) 271-278. New York: Ballantine Books .. Bateson, G. (1979). Mind and nature. New York: Ballantine Books. Bateson, G., & Bateson, M. c. (1988). Angels fear: Towards an epistemology of the sacred. New York: MacMillan Publishing. Bateson, G., Jackson, D., Haley, J., & Weakland, J. H. (1956) . Toward a theory of schizophrenia. Behavioral Science, 1 (4), 251-264. Bavelas, J. (1978). Personality: Current theory and research. Monterey, CA: Brooks/Cole. Beavin Bavelas, J., Black, A., Chovil, N., & Mullett, J. (1990). Equivocal communication. Newbury Park, CA: Sage .Publications. Bennis, & Nanus, B. (1985). Leaders. New York: Harper & Row. Campbell, D. T. (1965). Ethnocentricity and other altruistic motives. Nebraska symposium on motivation, ed. by Levine, D. Lincoln, NE: University of Neb-raska Press. Campbell, J. (1982). entropy, language, and life. New York: Simon & Schuster. Carlzon, J. (1987). Moments of tiuth. New York: Ballinger Publishing. Compact edition of the Oxford English dictiona+y (1971). Oxford: Oxford University Press. Cunningham, S. B. (1991 ). Intrapersonal communication: A review and critique. Communication Yearbook, 597-620. Eisenberg, E. M. (1984). Ambiguity as strategy in organizational communication. Communication Monographs, 227-242. 226

PAGE 238

Eisenberg, E. M., & Goodall, H. L. Jr. (1993). Organizational communication: Balancing creativity and constraint. New York: St. Martin's Press. Ferguson, K. G. (1991 ). Equivocation in the surprise exam paradox. The Southern Journal of Philosophy, (3), 291-302. Fiske, s. T., & Taylor, s. E. (1984). Social cognition. New York: Newbery Award Records, Inc. Geneen, H., & Moscow, A. (1984). Managing. New York: Avon Books. Gergen, K. J. (1985). The social constructionist movement in modern psychology. American Psychologist, 266-75. Goss, B. (1989). The psychology of human communication. Prospect Heights, IL: Waveland Press. Hadamard, J. (1945). The psychology of invention in the mathematical field. New York: Dover Publications. Hofstadter, D. R. (1979). Godel, Escher, Bach: An eternal golden braid. New York: BasicBooks. Johnson, c. E., & Hackman, M. z. (1994). Creative communication: Principles & applications. Prospect Heights, IL: Waveland Press. Katz, D., & Kahn, R. L. (1966). The social psychrilogy of organizations. New York: Wiley. Korzybski, A. (1941). Science and sanity. New York: Science Press. Lewin, K. (1938). The conceptual representation and measurement of psychological forces. Contributions to Psychological Theory 1, 4, (4). Mead, G. H. (1956). Social psychology. Ed. by Strauss, A. Chicago: University of Chicago Press. 227

PAGE 239

Mead, M. (1979). End linkage: A tool for cross-cultural analysis. In About Bateson. Ed. by Brockman, J. New York: Dutton. Naisbitt, J. (1982). Megatrends. New York: Warner Books. O'Connor, D. J. (1948). Pragmatic paradoxes. Mind, 358-9. Pearce, W. B., Cronen, V. & Harris, L. M. (1982). Methodological considerations in building human communication In Human communication theory: comparative essays. Ed. by Dance, F. X. E. New York: Harper & Row. Schein, E. (1969). consultation: Its role in organizational development. Reading, MA: AddisonWesley. Senge, P. (1990). The fifth discipline: The art and practice of the learning organization. New York: Doubleday. Shannon, c., l Weaver, W. (1949). The mathematical theory of communication. Champaign-Urbana, IL: University of Illinois Press. Shulman, M. (1955). Max Shulman's guided tour of campus humor. Garden City, NY: Hanover House. Simmel, G. (1950). The sociology of Georg Simmel. Trans. Wolff, K. H. New York: Free Press. Sluzki, c. E., Beavin, J., A., & Veron, E. (1967). Transactional disqualification: Research on the double bind. Archives of General Psychiatry, 494-504. Warriner, C. K. (1958). official -morality. 165-8. The nature and functions of American Journal of Sociology, Watzlawick, P., Beavin, J., & Jackson, D. D. Pragmatics of human communication: A study of interactional patterns, pathologies, and paradoxes. New York: W. W. Norton. 228

PAGE 240

Weblin, J. E. (1962). Communication and schizophrenic behavior. Family Process, 1, 5-14. Weick, K. (1969). The social psychology of organizing. New York: Addison-Wesley. Weick, K. (1979). The social psychology of organizing (revised). New York: Addison-Wesley. Whitehead, A. N., & Russell, B. (1910-13). Principia mathematica. Cambridge: Cambridge Press. Wiener, N. (1967). The human use of human beings. York: Avon Books. Yogananda, P. (1946). Autobiography of a yogi. Los Angeles: Self-Realization Fellowship. 229 New