'I'-I-IIIIIIIP- ii \iiiiMimiiiiiii ”£818 scum STATE unwensmr uam‘mfis 3 1293 0168 067 This is to certify that the dissertation entitled Virtual Organization and Perceived Availability: The Case of Manufacturing Assistance Providers presented by Scott Melbourne Preston has been accepted towards fulfillment of the requirements for Ph.D. degree in Communication fl) “mm I Major professor Date June 26’ 1998 MSU is an Affirmative Action/Equal Opportunity Institution 0- 12771 LIBRARY Michigan State Unlverslty PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE me WM“ VIRTUAL ORGANIZATION AND PERCEIVED AVAILABILITY: THE CASE OF MANUFACTURING ASSISTANCE PROVIDERS by Scott M. Preston A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1998 ABSTRACT VIRTUAL ORGANIZATION AND PERCENED AVAILABILITY: THE CASE OF MANUFACTURING ASSISTANCE PROVIDERS By Scott M. Preston With the advance of information technologies and mediated channels for communication, new ways of organizing are possible. Despite human need and proclivity for face to face communication, strong ties develop and are maintained (at least in part) through mediated channels tied by bonds of trust and common interests rather than proximity and formal structure, systems based on virtual organization transcend the boundaries of time and space. The process of virtual organization suggests that the apparent dichotomy between proximate and mediated communication should be replaced with a model that remakes the forms communication can take into complementary categories Of experience rather than mutually exclusive contexts. Virtual organization is a process of social construction (Gergen, 1985) for a system grounded in shared rules and resources (Giddens, 1981) that creates and maintains the availability of all task-relevant actors and objects in the minds of its members. TO understand this process and its outcomes as they relate to perceptions Of availability, elements of cognitive and structuration theory are integrated under the perspective of social construction. Cognitive theory addresses how behavior production occurs through memory’processes, while structuration theory explains how the categories, norms, and routines used by individuals arise and are reinforced by social interaction. The content of memory at all levels above the most basic is a product of social construction; new memory traces are determined by the structural lens through which social actors perceive the physical world. Structures are the expected patterns of physical sensations that elicit meaning at multiple levels of abstraction. The key structure created by virtual organization is the actor's perception of available others who share the same purposes, as embodied in a transactive memory system (W egner, 1987). It is this perceived network of available others that makes virtual organization possible for individual actors. This dissertation seeks to answer three questions: What structures are created by virtual organization? How do these structure come into being? What changes does virtual organization bring about in existing structures? These questions are explored through an instrumental case study (Stake, 1995) of manufacturing assistance providers who were recruited by a university project (NEM Online) developing online resources and services to improve the global competitiveness of Michigan manufacturers. Drawing on the author’s personal experience as a member of NEM Online, archival documents Of NEM Online’s development, two network surveys, and interviews with assistance providers, this dissertation describes the world of assistance providers and identifies how they come to perceive certain resources as available. Perceived availability is identified as the mental labels actors have for objects and people, which are cued by the categories people apply to inquiries and reinforced by their movement through an interaction space defined by time schedules instead of territory. The theoretical perspective is found to be sufficient for analysis at the program, relationship and sequence levels of control, but in need of more detailed theory to frame research at more concrete memory levels. COPYRIGHT BY SCOTT MELBOURNE PRESTON 1998 ACKNOWLEDGEMENTS This dissertation has benefited from the hard work of my dissertation committee, Jim Dearing, Ken Frank, Steve McComack, and Sandi Smith. Their wide range of interest and provocative teaching inspired me to think big yet stay connected to the components Of my vision. Without their challenging questions and careful critique, this dissertation would not be before you now. TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES LIST OF KEY TERMS NOTES 'ON USAGE CHAPTER ONE: NEM ONLINE AND VIRTUAL ORGANIZATION Mediated communication The Internet as a focus of research The Michigan Industrial Extension Partnership Virtual organization The emergence of NEM Online Communication and participation in NEM Online NEM Online and virtual organization CHAPTER TWO: THEORETICAL REVIEW Social constructionism Cognitive theory Structuration theory Review of theoretical relationships Issues for research Structuration of mediated relationships Theoretical research questions CHAPTER THREE: CASE STUDY DESIGN Points of contact Research plan Operational definitions Data collection Network surveys Interviews Archival Observation vi page no. ix x xi xii 25 25 34 45 58 61 65 73 73 75 77 79 8O 85 86 86 Data analysis Network surveys Interviews Archival Summary Emergent research questions CHAPTER FOUR: EMERGENT RESEARCH QUESTIONS: THE WORLD OF ASSISTANCE PROVIDERS Missions of assistance providers (Emergent l) Conservatism of assistance providers Proactive and reactive searching Building a network Of resources Surveillance of environment Responding to questions and needs Defining the information environment (Emergent 2) Requirements of place Importance of local resources Importance of gatekeeping Perceptions of face to face communication Leaming to use online resources Lack of time to learn new technologies Facilitating mediated communication Relationships and affiliations (Emergent 3) Affiliation requirements and opportunities Need to use what is available in environment Access limitations Required communication What inspires assistance providers Desired situations Personal versus exchange relationships Resource clusters CHAPTER FIVE: 87 89 90 91 92 92 95 97 102 105 105 108 114 120 120 122 123 126 127 132 134 135 135 135 136 138 139 139 141 144 THEORETICAL RESEARCH QUESTIONS: PERCEIVED AVAILABILITY AND THE STRUCTURE OF INFORMATION ENVIRONMENTS Categories for mediated contact (Theory 7) Categories applied by assistance providers Key services of assistance providers Categories for information seeking (Theory 2) Criteria for evaluating resources Location vii 151 153 153 160 164 165 167 Recognizing relationship requirements (Theory 1) Attributes and perceptions of voice mail Face to face as foundation for mediated communication Redefining regions via mediated communication (Theory 3) The effects of proximity Importance of meetings Geographic affiliations Identifying acceptable channels for communication (Theory 6) Expectations for response Evaluating the types of resources Channel characteristics and learning costs Patterns of communication during information seeking (Theory 4) Electronic mail usage Mass messages sent and received by assistance providers World-Wide Web use Perceptions of databases NEM Online’s contact with assistance providers Expectations for interaction (Theory 5) Advantages and disadvantages of resource types Cross-cutting Obstacles Summary of alternative explanations for issues CHAPTER SIX: CONCLUSIONS Evaluation of theory Review of theoretical elements Summary of theoretical contributions Evaluation of NEM Online Targeting assistance providers Establishing value of content Achieving perceived availability LIST OF REFERENCES viii 170 171 173 176 177 178 180 183 184 187 188 191 192 196 197 201 205 211 212 218 222 225 225 226 239 241 241 243 244 246 Label Table 2.1 Table 3.1 Table 4.1 Table 4.2 Table 4.3 Table 5.1 Table 5.2 Table 5.3 Table 5.4 Table 5.5 Table 5.6 Table 5.7 Table 5.8 LIST OF TABLES Description page no. Key definitions 62 Data sources for emergent and theoretical research 87 questions Number and type of respondents to data collection methods 98 Number of office staff and the use of gatekeepers 122 Clusters identified in correspondence analysis for 148 1997 survey data Local resources cited on 1997 survey, by type of resource 182 Assistance providers’ expectations for response 186 Access and use of phone, fax and email by 194 assistance providers Bookmarks cited by assistance providers 199 Assistance provider Internet use and familiarity with 203 databases Communication between NEM Online and assistance 207 providers during update of Referral System profiles, November 1997 — January 1998 Time required to complete updates of assistance providers’ 208 Referral System profiles, November 1997 - January 1998. Advantages and disadvantages of human and document 213 I'CSOLII'CCS Label Figure 1.1 Figure 2.1 Figure 2.2 Figure 4.1 Figure 5.1 Figure 5.2 LIST OF FIGURES Description Types of availability for interaction Negative feedback loop Hierarchy of control Clusters of most-cited resources and respondents to 1997 survey Clusters of resources sharing origin and form and respondents to 1997 survey, first iteration Clusters of resources sharing origin and form and respondents to 1997 survey, second iteration page no. 14 37 39 145 157 158 LIST OF KEY TERMS Action IOOp: A negative feedback loop for behavior generation via discrepancy reduction. Accessibility: The capacity for contact via a given channel. Availability: The perceptual quality of being present in social space for interaction, via physical presence or mediated channel. Categogy: A mental label that identifies specific behaviors as salient. External memory: The knowledge a person has through objects (e.g., books, files). fl)_r_m_: The various ways information can be accessed (e.g. organizational representative, a person known as an individual, a print or electronic document). _I1i§r_arc_hv Of control: The nine levels at which human attention can be focused, from least to most abstract in content. Individual memog: The knowledge a person has in his or her own mind. Qg'gip: The identifying label connecting multiple forms to each other. Rpm: The path actors follow through time and space. Resource: An accessible object or person with relevant information or expertise. Rpptin_e: A well practiced series of behaviors at lower levels of control that lead to discrepancy reduction at a higher level of control. Structure: The rules and resources that comprise and enable social systems. M: A patterning of relationships between entities, situated in time and space. Transactive memog: The knowledge a person has through other people. Virtual organization: A process of social construction (Gergen, 1985) for a system grounded in shared rules and resources (Giddens, 1981) that creates and maintains the availability of all task-relevant actors and objects in the minds of its members. xi NOTES ON USAGE There are three conventions I follow in this dissertation that require explanation. First, in order to be inclusive I alternate the gender of pronouns (e.g., him and her, she and be) when referring to generic individuals for purposes of illustration. In cases where I play through an example of an interaction, I assign names to make the example more personal and more readable. When the interaction is dyadic, one name is male and one female so that I can use pronouns without the risk of making ambiguous references. Second, I underline key terms in the text to signal a formal definition, which follows the underlined word in a sentence. For example, in defining communication the sentence would read: “Communication is the symbolic exchange of meaning between two or more actors” (no quotation marks in text). In some cases, the underlined word comes after the definition. I do this when definitions logically flow from one to the next to avoid identical sentence structure and make clear the connection between the terms. If I use the term in a sentence before providing the definition, I introduce it immediately afterwards, in parentheses or the next sentence. Third, I alternate between the word “actor” and “person” depending on the content Of the chapter. When describing results and presenting context, Iuse person. But in the theory chapters (Two and Six), I emphasize the performance aspect of social behavior in my labeling Of its participants as actors to fully convey the implications of my theoretical argument and the necessary social context of all behavior.l While I grant that each actor is a distinct individual, I reject the argument that an individual’s being or behaviors can therefore be understood in social isolation. Without the levels Of meaning made possible by the presence and attention of others to an act, the behaviors of an individual can only be understood in a trivial sense. ' The term actor can also be applied to an organization, which also act in a social system. Thus, the same points apply to these more macro entities. xii CHAPTER ONE NEM ONLINE AND VIRTUAL ORGANIZATION This dissertation investigates the concept of virtual organization through a case study of manufacturing assistance providers in Michigan. These people were chosen because they were one of the populations of interest served by a university grant-funded project, NEM Online, that sought to use electronic media to increase the competitiveness of Michigan’s manufacturers. In the course of its four year development from September 1994 through May 1998, NEM Online reinvented itself multiple times. The evolution of NEM Online from one part of a three pronged industrial extension effort to an emerging for-profit business was influenced by many changes in the world of telecommunications. This chapter introduces the emerging importance of personal mediated communication, explains the new technological environment and the concept of virtual organization that emerges from it, and describes the relevant parts of NEM Online within the industrial context Of its efforts. MEDIATED COMMUNICATION Historically, the literature on organizational communication has focused on interactions that occur in the same physical place. When mediated communication was addressed, it was treated as an independent category Of interaction or as an entirely independent phenomenon, unconnected with face to face communication. However, as new media have arisen and spread through society, this distinction between mediated and proximate communication has become less clear. As the number of media options increases, they take up an increasingly larger proportion of the time social actors spend communicating with others. The networks now accessible through mediated channels contain sufficient rewards to be worth the initial costs in the time and effort it takes to l learn. For example, personal relationships need not die when one person moves away, and immediate access to up to date market and political information gives a business a competitive advantage over firms without that contact (Goldman, Nagel, and Preiss, 1995; Davidow and Malone, 1992). The question is how to incorporate proximate and mediated processes within the same organizational framework. The false dichotomy between face to face and mediated communication (both mass and personal) dominates the literature on mediated communication. After thousands of years of only experiencing the world through our unaided senses, the twentieth century has seen the world shrink as people gained the tools to see and hear events continents away. This mediated experience is definitely different from what people experience when physically present, but the two types of experience are neither Opposites nor mutually exclusive (see Reeves and Nass, 1996). Today’s friendships are maintained through both mediated and proximate interaction (W ellman, Carrington, and Hall, 1988); in the world of computer-mediated communication, friendships often begin and develop without the actors ever being in the same physical place (Parks and Floyd, 1996). One requirement for friendship is interaction for reasons besides serendipity; with the telephone’s near-universal availability in Western society, few of us have a friend who is not available via media. Increasingly, the problem is not how to reach a distant person, but when to initiate contact so it fits both actors’ schedules. The temporal structure of human relationships is more important than the technical characteristics of the channels used for communication. If the same set of motives lead to both mediated and proximate interaction, then we should focus on function instead of form (Ogan, 1993). A 9121111131 is the means of delivering a message, such as making sounds through speech or typing words on a computer keyboard; a medium is what the encoded message passes through on its way from sender to receiver. While it is generally accepted that face to face is a channel, the medium in use is often overlooked. Communication in the face 2 to face channel passes through the medium of air; since air is a constant component of human experience, it can be difficult to recognize as a context. The term “mediated communication” does not distinguish between face to face communication and everything that passes through a medium other than air. Instead, mediated communication describes interaction patterns that take place across a range of channels, including but not limited to face to face. This contrasts with efforts to expand the mutually exclusive typology Of channels, such as Ball-Rokeach and Reardon’s (1988) explication of a third form of human communication, the telelogic, to describe mediated but not mass communication. While Ball-Rokeach and Reardon (1988) make a useful distinction, they fail to consider the extended nature of human communication across time as well as space. They have enriched our understanding of form but neglect the unifying approach of function. The temptation to create a dichotomy of mediated and proximate interaction must be resisted in favor of the view that the two are more complementary than exclusive. Only by broadening the scope of our interests can we accurately depict the complex realities Of the mediated social world. The idea that mediated channels are complementary to face to face has been around for a long time (e. g., Katz and Lazarsfeld, 1955). Uses and gratifications theorists (Blumler and Katz, 1974; Rubin, 1994) have long argued that people choose media to meet needs that were once met through face to face interaction alone. Typically, uses and gratifications researchers attempt to explain why people use different types of mass media by identifying the needs that are gratified by that media use (e. g., Garramone, Harris, and Anderson, 1986; Lichtenstein and Rosenfeld, 1984). People are assumed to be rational users who act in order to gratify needs; while actors are not always aware of which needs they are gratifying at the moment of choice (see Rubin, 1984, on instrumental versus terminal use of media), they are usually able to articulate their needs when asked to explain their media choices after the fact (Rubin, 1994). Much Of human behavior is both purposive (sometimes termed strategic) and routine, i.e., it doesn’t involve ongoing 3 cognition (Kellermann, 1992). The emphasis on an actor’s behavioral repertoire and the perceived efficacy of these behaviors at meeting the actor’s goals are found in two variables common to uses and gratifications research: The actor’s perceived utility for a channel, and the range of alternatives available to the actor for gratifying the salient need. The successful application of uses and gratifications to interpersonal motives normally met through proximate interaction by Rubin, Perse, and Barbato ( 1988) supports the claim that it is not the channel chosen that is most important, but rather the relational goals that the actors are attempting to achieve. The alternate paths forged by actors require different skills and provide different experiences along the way, but their final goal is the same. The Internet as a focus Of research One of the fastest growing topics of media research is the Internet. The Internet emerged from an internal project at the Advanced Research Projects Agency (ARPA) to network all their computers at sites across the US. as a means Of saving money and increasing computing resources (Hafner and Lyon, 1996: 41-42). This network, known as ARPANET, came online in the early 19705 and was used mainly by the military. During the 19708, computers were often sold in sets that could be networked together; however, purchasing was Often haphazard, so organizations ended up with computers from multiple manufacturers that had different networking protocols. The need to communicate across computer brands and protocols led to the development of a universal protocol called TCP/IP that transcended the proprietary protocols of computer manufacturers. This universal protocol made the Internet possible; when ARPANET became fully operational in 1982, TCP/IP was the protocol used to share information between computers at different locations (Comer, 1994: 58). This was the first indication that networking technology would turn business principles upside down, as TCP/IP and the Internet that it made possible created an open architecture that everyone could learn to use, in contrast to 4 the closest architectures of computer manufacturers, which were patented or kept as trade secrets (Comer, 1994: 56). While the Internet has many forms of content, several are particularly relevant to the above discussion. USENET is a structured hierarchy of bulletin boards where people post, read, and respond to the messages of others (Corner, 1994). There are over 5,000 topical bulletin boards (or newsgroups) on USENET, with topics ranging from computer architecture and programming to science fiction to the culture of Sri Lanka. Multi-User Domains (MUDs) are computer games that allow users to interact with each other in virtual space, creating stories and digital objects in a world maintained for their amusement. Some MUDs follow an adventure game format similar to role-playing game like Dungeons and Dragons, while others have no explicit goal other than social interaction. These virtual worlds have provided ample evidence of the significance of mediated interaction to people in their “real” lives, as the following brief overview of research will demonstrate. Recent research on mediated communication has demonstrated that strong social relationships emerge, develop, and even migrate from their origins in Internet communication. Turkle (1995) described the complex social worlds that emerge as people interact with each other’s digital personae, and analyzes the significance Of these virtual relationships from a psychoanalytic perspective. Baym (1995) investigated the process of community building in a USENET group devoted to discussion of television soap operas (rec.arts.tv.soaps). Reid (1995) studied how MUDs mediate between the user’s imagination and what she communicates to others about her imaginings in the virtual world. Kim (1998) describes the growing popularity of the online game Ultima Online, emphasizing the contrasting effects of different types of realism. The game is popular because it looks like players are in a real place, but it is frustrating for many because the game system is too realistic in its requirements for many hours of practice before player performance can improve, which gives rise to the fast growing phenomenon 5 of “player killing” by users used to action games where combat is the sole reason for playing. Finally, Parks and Floyd ( 1996) performed a survey study with participants in 24 randomly selected USENET newsgroups that found 72.2% of women and 54.5% of men had personal relationships with people they had met online, and 63.7% of the respondents had communicated with these others in another channel besides the computer. With 62 million adults (30% of the US. population age 16 and older) using online services at the end of 1997 (IntelliQuest survey, 1998), the phenomenon of Internet communication is a huge Opportunity for examining relational issues that transcend face to face interaction. While the Internet began as a task-focused means of sharing information among government researchers and academics (Comer, 1994), it has only been recently that the business community has begun to see cyberspace as a developing arena for inter- organizational communication. But now, government contractors must submit all their bids via the electronic data interchange. Sales from Internet retailers in the US. alone are expected to generate $4.8 billion in revenues for 1998, and reach more than $17 billion in 2001 (Conover, 1998). The Internet has also been recognized by marketers as an enormous channel for direct advertising. Television commercials now publicize the company’s World-Wide Web home page by listing its Uniform Resource Locator (URL) along with their logo, and many Web sites have banners advertising products and services ranging from computing tools (e.g., http://www.microsoft.com) to the US. Marine Corps (e.g., http://www.marines.com). Telecommunications firms trumpet their computer networking capabilities (e. g., network MCI) and threaten that those who hesitate to get on the information superhighway will be left far behind everyone else, socially and professionally. As AT&T ominously predicted in 1995, “You will” - like it or not. THE MICHIGAN INDUSTRIAL EXTENSION PARTNERSHIP In early 1993, the federal government issued a call for Technology Reinvestment Program (T RP) proposals to fund projects that would benefit US. business. In Michigan, industrial extension service providers from Michigan State University, the University of Michigan, the Industrial Technology Institute, and the Michigan Department of Commerce had been developing what they termed a “virtual organization” known as the Michigan Industrial Extension Partnership (MIEP). The goal of MIEP was to create collaborative ties between knowledge providers, industrial extension agents, and small businesses in Michigan as a means to improve Michigan's economy and increase the global competitiveness of Michigan manufacturing firms. MIEP submitted five related proposals to the TRP, three of which were funded: (1) Developing common methods and training field agents for industrial extension; (2) building a communication network linking knowledge-providers, small business users, and industrial extension agents in Michigan; and (3) planning grant for the Michigan Industrial Extension Partnership. The two proposals that were not funded requested funds to establish manufacturing outreach centers in West Michigan and the Upper Peninsula. These centers were intended to be the marketing arm of MIEP; when they were not funded, none of the funded projects expanded their mission to include marketing. Collectively, the funded proposals were likened to the musculature, nerve system, and brain of the “virtual organization” MIEP (Michigan Industrial Extension Partngshjp, 1993). The mission statement for MIEP emphasizes innovation in multiple arenas: The mission of the Michigan Industrial Extension Partnership is to enable small and medium sized manufacturing enterprises in Michigan to gain and maintain superiority in an increasingly competitive global marketplace. The Partnership will accomplish this mission by assisting businesses in Michigan to: [1] speed the adoption of new and rapidly evolving technologies, [2] invest in modern plant and equipment, [3] develop new domestic and export markets in a rapidly changing global 7 economy, [4] link up businesses with research, development, and industrial extension centers to develop and market new products and processes. (Michigan Industrial Extension Partnership, 1993) The communication system proposal focuses on the fourth point, creating linkages between the different stakeholders to facilitate collaborative efforts benefiting the manufacturer. These linkages are in turn intended to facilitate the other three points, which are the province of the common methods and training proposal. All of this was to be madepossible by the funding, coalition building, and assignment of roles and responsibilities that the planning grant proposal would provide via a consumer-driven design process. While MIEP was intended to function as a whole, the three funded proposals were enacted by different partners. Michigan State University (MSU) was responsible for the communications system, the Industrial Technology Institute was in charge of the training program, and the Michigan Department of Commerce (which became the Michigan Jobs Commission in 1995) directed the planning grant. As the communications network was the element that would make MIEP a virtual organization, this dissertation focuses on it. The purpose Of the communication system, originally called the Manufacturer's Support Network (MSN), was to improve the ways small and midsize manufacturers get and use information, a process structured into a three year plan. In the first year, the system itself would be designed. To this end, a team of librarians and faculty from the engineering and business colleges at MSU was assembled to design the technological system for communication, develop procedures for building up a collection of information relevant to manufacturers and assistance providers, and evaluate existing databases as potential parts of the MSN's information resources. In the second year (1995-1996), the system would be introduced to assistance providers through a series of training sessions, a directory of assistance providers in Michigan would be developed, and formal contracts with the selected databases finalized. The third year (1996-1997) was scheduled for implementation of the system with manufacturers themselves, utilizing 8 the assistance providers trained the year before as diffusers of the information resource innovation. The means by which the communications network would connect people with other people and various documents underwent a dramatic transformation from the TRP proposal to the end of the grant funding. Two trends in computers transformed the interface of the MSN. First, computer operating systems moved away from the black screen and text commands of DOS towards the icons and point and click interface of Apple’s Macintosh operating system and Microsoft’s Windows software.1 While DOS requires often esoteric text commands, Macintosh computers and DOS computers with a Windows interface show the user a screen where icons represent programs and documents. The user can move the cursor around the screen to open documents or run programs by “clicking” the left mouse button while the cursor rests on top of the icon. In 1995, Microsoft unveiled Windows 95, which integrated the Windows interface with the DOS operating system. Since then, Windows and Windows 95 have become so dominant that the US. Department of Justice has brought antitrust charges against Microsoft regarding the integration of functions often sold by competitors as separate programs (Christian Science Monitor, 1998). The second major change in computing occurred with the creation of Hypertext Transfer Protocol (HTTP) in 1991. Previous networking technologies such as Gopher, File Transfer Protocol (FTP), and electronic bulletin boards used text-based, menu-driven DOS interfaces to transfer information from one computer to another. Gopher, an interactive software application for reading files based on a hierarchical model of folders and files (Kardas and Milford, 1996: 59), allows users to search for sites and move betweenmenus to find the information they want. However, Gopher limits users to following the hierarchical structure of the files on the computer containing the relevant ' Personal computers (PCs) predominantly have either a Macintosh or a DOS operating system. Windows is a software program that gives the PC user an interface like a Macintosh. 9 information. In contrast, hypertext differs from standard text in that its units are linked to multiple others, which allows a reader to follow a non-linear progression of ideas. From any one point, a reader can jump to a number of other documents written in Hypertext Markup Language (HTML) using browser programs such as Mosaic and Netscape Navigator. Instead of the line created by traditional text, a diagram of hypertext linkages would appear more as a web. The name “World-Wide Web” (or simply “the Web”) reflects this potential for linking information in many new patterns. As HTML documents became more common, the Web was born. The original plans for the MSN were based on technologies such as Gopher and electronic bulletin boards. By 1995, it was apparent that the MSN had to be drastically revised to take advantage of the Web’s superior communication capacity. The MSN would take the form of a Web site where manufacturers could find the information they needed to address their management, marketing, and technological problems. To this end, librarians and computer scientists were hired to develop and maintain a database for manufacturing-related information, and a partnership developed with the regional telephone company (Ameritech) and MERIT, the state-wide Internet service provider for MSU, to increase the number of phone lines available across Michigan with dial-up access to the Internet through MSU’s Computer Center.2 To collect the content for the database, graduate students in engineering and management were hired to seek out manufacturing related documents for the database on the Web using search engines such as Lycos, Webcrawler and Alta Vista. The process of moving from Web page to Web page was known as “surfing;” thus, these students were Officially known as surfers. Search engines sprang up to help users find what they wanted When they did not know the Universal Resource Locator (URL) of the pages they sought. ‘ 2 To access the Internet for electronic mail or the World-Wide Web, you must have an account on a computer with an Internet address (such as pilot.msu.edu). As most people do not have a computer that is Part of the Internet, they need to call the computer where they have an Internet access account and log in using TCP/IP software that formats messages to be sent over phone lines to the Internet computer. 10 However, searching the Web for manufacturing related information was described as “trying to drink from N iagra Falls,” as most searches returned a list of hundreds of Web pages, many of questionable relevance. This made the indexing of valuable manufacturing resources a primary part of the effort. Because the information available on the Web varied dramatically in quality and detail, evaluation became a crucial task. To address these unanticipated problems, the first priority of the MSN became cataloging of the information. In addition to hiring a full-time librarian to work on the project in July 1996, the MSN began a number of pilot projects in major Michigan industries, particularly automotive, office furniture manufacturing, and wood products. These projects were the first step in recruiting domain experts for assistance in evaluating the quality and depth of the information being cataloged. These domain experts were to be drawn from the ranks of assistance providers to identify user needs, evaluate information for inclusion in the MSN, moderate conferences, and manage bulletin boards related to their area of expertise. Their inclusion in the communication component of MIEP was part of the effort to bring all manufacturing assistance providers in Michigan under the same organizational umbrella to create a virtual organization. Virtual organization Change is the driving force in the modern organizational world (T erreberry, 1968; Goldman, Nagel, and Preiss, 1995). As the pace of technological development increases, organizational decision makers find that their set ways of dealing with their environments are fast becoming obsolete. From this pressure comes the need to innovate in order to remain viable in an increasingly competitive global marketplace. Larger institutions are spending more and more effort to become self-designing organizations (W eick, 1977), but small and medium size businesses often lack the resources to identify and realize optimal alternatives using only knowledge internal to their organization. Thus, the 11 informal networks that cut across formal organizational boundaries are of pivotal importance to small and medium size businesses as a source of information, particularly in an environment of change (Saxenian, 1994). Under these conditions, virtual organization becomes a potent means of structuring experience across units for the benefit of all. Virtual organization is a process of social construction (Gergen, 1985) for a system grounded in shared rules and resources (Giddens, 1981) that creates and maintains the availability of all task-relevant actors and objects in the minds of its members. The process of creating and sustaining this system occurs regardless of the time and space barriers between the system’s members because the ties that bind participants in virtual organization together are not the result of shared concrete experience but rather similar perceptions of the world. Because participants in virtual organization share fewer concrete referents than participants in more traditional forms of organization, they rely heavily on shared abstractions, such as procedures, goals, and expectations. Prior to this century, the participants in a virtual organization rarely communicated, as time and space were near insurmountable obstacles to regular contact. But with the arrival of telecommunications technologies such as the telephone, electronic mail (email) accounts, and facsimile (fax) machines, distance and time ceased to be as imposing as before, and it became possible to make connections with like-minded others without ever meeting them face to face. The combination of electronic media and recent innovations in information processing is the most important development in business this century (Davidow and Malone, 1992; Goldman et al., 1995), and another manifestation of people gaining more control over their environment (Beniger, 1986). Thus, how information is made accessible is one of the most important and interesting questions asked in organizational communication research. The greatest change of the 19903 has been the emergence of the computer as a powerful and pervasive medium for communication. Electronic networks 12 such as the Internet, the World-Wide Web (the Web), and electronic bulletin board systems are constantly expanding in both content and users. Computer mediated communication today is an expanding array of alternative means of presenting, disseminating, and gathering information. “Cyberspace” has become part of popular and business culture, and many entrepreneurs are vying to make a fortune in this virtual world. But the driving force behind these changes is the search for ways to overcome obstacles to interaction by exploring the possibilities for organization and communication in an evolving social structure whose boundaries are not yet defined by a paradigm for use. The central problem of virtual organization is how its members are recognized by each other. The process of virtual organization does not rely on the physical cues commonly associated with social presence, but rather develops new cues and expectations that permit members to achieve their goals through mediated interaction. Giddens' (1981) concept of presence availability provides a basic model for constrained interaction. People who are available are in the same physical place and are not engaged in any activity that precludes interaction have presence availability, i.e., they are accessible for conversation. People can also be in the same place and not be available for interaction; for example, two people engrossed in conversation are bLsy: They are not available for interaction, despite being in the same interaction space. What is missing is a construct that addresses how people can be available for interaction and yet be in different physical places. To parallel presence availability, we need only add another dimension to Giddens' (1981) dichotomy, creating a two by two typology of availability by presence, as illustrated in Figure 1.1: 13 Not Available Available Present in place Busy Presence Availability Not present in Jrlace Absent Mediated Availability Figure 1.1 Types of availability for interaction Thus, for virtual organization to occur, the structure for information seeking shared by its members must include resources that have mediated availability as well as those which have presence availability. Mediated availability is prefigured by several constructs in the media and sociological literatures. One such construct is Korzenny’s (1978) electronic propinquity. Korzenny focused on the importance of perceptions of belonging to any theoretical perspective on electronic media: Perceived propinquity then is that communicative trait...which should reflect a sense of involvement, mastery over the communication environment, and which, within certain limits, should provide communicative satisfaction - when the interaction is desired. (Korzenny, 1978:8) Recent research (Fulk, 1993; Schmitz and Fulk, 1991; Corman, 1990) has demonstrated the impOrtance of the perceived qualities of a medium, as Korzenny predicted. However, Korzenny's (1978) theory only predicts relationships in the limited cases of single interactions with no expectation of future interaction. Walther (1994) demonstrated that these conditions were too constraining for an accurate depiction of mediated communication. Thus, while Korzenny’s construct is ahead of its time, his theory lacks the necessary scope to be useful. The term mediated availability is both more descriptive and derived from a richer theoretical perspective than electronic propinquity. 14 Virtual organization has appeared before under other names, the most famous of which is the invisible college (Crane, 1972). In her review of innovation in the research fields of rural sociology and mathematics, Crane (1972) found that there were shared methodologies and research foci across distance that structured the two disciplines. From this, she hypothesized the existence of a social structure that cut across institutional, distance, and temporal boundaries: The invisible college. However, Crane (1972) and other invisible college researchers (see Chubin, 1983, for a review) focus more on the outcomes of communication, such as patterns of citation and the diffusion of research and theoretical innovations, than on the actual communication sequences and relationships between researchers. Lievrouw (1989: 618) points out that the original “invisible college” existed in the absence of formal organization (Price and Beaver, 1966) rather than across existing organizational boundaries, as Crane (1972) defines it. In the case of NEM Online, the first position is more useful, as NEM Online is not transcending organizational boundaries so much as it is attempting to create a community of assistance providers that subsumes more specific networks. Crane’s failings are summed up by Lievrouw (1989: 620) as follows: Crane’s findings tend to obscure the central role of communication behavior and interpersonal processes and emphasize the mapping of institutional structures. She [Crane] uses citation data as evidence of the informal communication relations among her subjects, yet by using structural data she reinforces the primacy of structure over process. Thus, the work of Crane and other invisible college researchers tells us who has knowledge without describing how individuals acquire and disseminate it. While the existence of invisible colleges is an important first step in demonstrating the need for the construct of mediated availability, research must now address the systems that give rise to and sustain the process of virtual organization. 15 The key element of virtual organization3 is that its participants draw on both present and mediated resources. The argument of this dissertation is that mediated availability is increasingly important to the organization of social systems, and that the differences between mediated and present channels can be overcome via social construction of alternate behaviors. Advancing telecommunication options change the competitive environment and require new paradigms of intra- and inter-organizational communication and information processing. Media coverage of developments in cyberspace raises the overall awareness of technological change in society, and has economic impacts as businesses react to this new frontier. Recent work in management has focused on the crucial role of information processing technology in keeping organizations effective.4 In a virtual organization, information is a resource being moved or copied in a way that has economic impact. Goldman et a1. (1995: 7) and Davidow and Malone (1992) argue that virtual patterns of organization that depend on the ability to rapidly pull together coalitions to exploit transient market opportunities (what Goldman et al. term agfly) must replace the current mass production patterns if the United States is to remain economically competitive with other nations.5 Traditional patterns of organization for mass production organized workers so every one knew only their part of the process. In contrast, the members of a virtual organization are responsible for the entire process, not just for part of it, and thus require access to all resources necessary to keep the system in Operation. Thus, the characteristics of virtual organization are a focus on change, being customer driven and managed, and the presence of highly skilled, flexible workers working in a collaborative climate (Davidow and Malone, 1992: 7-8). 3 Virtual organization is primarily a process. However, to avoid awkward constructions I will occasionally use the term as a noun to indicate a system where the process of virtual organization is at work. 4 The focus of this literature is usually for-profit organizations, but the desired outcomes of non-profit and not-for-profit organizations require the same adjustments. 5 Saxenian (1994) describes the continued success of Silicon Valley as based on the existence of widespread informal networks for information sharing between collaborators and competitors, and contrasts this with the more rigid, formal, and of late less successful industries along Route 128 in Massachusetts. 16 What makes virtual organization unique is the structure it takes in the minds of its participants. The key to virtual organization is a broad network of collaborators and a focus on nurturing customers (Goldman et al., 1995; Davidow and Malone, 1992). But too often businesses do not know how to integrate telecommunications innovations with their existing technologies and structure, let alone use them to become agile manufacturers (Goldman et al., 1995). In response to this dilemma, a new industry has sprung up to teach businesses when and how these new telecommunications technologies can help them. But many small and medium-size businesses cannot afford to hire or train someone to take advantage of the Internet. These businesses fall back on the many different types of organizations that assist them in other aspects of their business, from writing business plans to finding a site where they can expand to training their workers in state-of-the-art techniques. These organizations have many names and many different patterns of affiliation, from municipal departments to utility divisions to university outreach. They can be grouped together on the basis of the assistance they provide to businesses and their motivation for doing so. These assistance providing organizations exist to help businesses - particularly but not exclusively manufacturers - become and remain profitable. These organizations, usually non-profit and often affiliated with government, do not exist to make money, but to keep economies strong and jobs plentiful. These assistance providing organizations attempt to grow the economy and create jobs in a given geographic area by providing assistance to businesses so they can remain competitive and encouraging the growth of new business. While MIEP aspired to being a virtual organization of assistance providers around Michigan, technological and political forces conspired to change the focus of the effort to institute collaborative assistance across space. 17 THE EMERGENCE OF NEM ONLINE In early 1995, the first of several name changes occurred as MIEP became first the Network for Manufacturing Excellence (NME) and soon after the Network for Excellence in Manufacturing (NEM). The impetus for the first change came from the manufacturers on the steering committee for MIEP, who wanted to shift emphasis from a process ("industrial extension") to an outcome ("excellence in manufacturing"). This shift emphasized that the organization rated the goal above the means used to achieve it, freeing NEM from adherence to any one approach to bringing about the desired outcome. The terms “network,” “excellence” and “manufacturing” were identified as the key elements of the proposed organization. How they were ordered was a function of the acronyms; as soon as someone pronounced NME as “enemy,” the words were rearranged to create the less menacing NEM. Excellence was the outcome, manufacturing the context, and network the means. As the communication and training components focused on applying telecommunication innovations, the computer science sense of the term "networ " was highly appropriate. In addition, NEM was to link assistance providers around the state in a loosely structured system of collaboration, a system aptly described as a "network" in the social science sense of the word. Thus, the new name served the dual purposes of focusing attention on outcomes and emphasizing the informal nature of the organization to the world. While MIEP became NEM due to the direct influence of the manufacturers on its advisory board, MSN was forced to change by an external organization’s advertising campaign. In the summer of 1995, Microsoft began marketing its new online service, the Microsoft Network, under the acronym "MSN." Given the size and prominence of Microsoft in the world of computers, the name of the communication project had to be changed. The new name was NEM Online, which both resolved any possible confusion between the Microsoft product line and NEM's communication system and made the 18 connection between NEM and its communication nerve system explicit. However, this name change occurred just before the communication system was to be advertised to assistance providers around the state, and thus left a legacy of confusion among those assistance providers who assisted in pilot tests of the training and other components of what they had known as the MSN and MIEP. In late 1995 and early 1996, NEM lost several of its partners. The Edward Lowe Foundation bought themselves out of NEM following the death of their founder, Edward Lowe. The Michigan Jobs Commission, which initially had its home page put up by NEM, became embroiled in a political struggle over its involvement with an organization that could point to out-of-state resources, and ceased active participation in the NEM effort. The head of the Michigan Small Business Development Centers (SBDCs) decided that it was not appropriate for the Michigan SBDC office to collect information about and publish a directory of assistance providers, which was to have been its major contribution to the NEM effort. While the Industrial Technology Institute (ITI) continued its training efforts under the NEM banner, it made no mention of the other partners. In effect, NEM Online became the only active part of NEM, largely because it was based in an academic institutiOn (Michigan State University, particularly the Department of Management) and was home to the principal investigators for the entire project. Because NEM retained some negative political baggage, particularly with the Jobs Commission and other state organizations, the “Online” was preserved as a means of distinguishing between the parent organization and its child project. There was talk of changing NEM Online to a name with no connection to N BM in June of 1996, but it got no further than the joke that the new acronym would be NMC, or “No More Changes.” When the SBDC network left NEM, the responsibility for creating a directory of people and organizations who could assist manufacturers fell to NEM Online. This audience of manufacturing assistance providers was seen as the logical intermediary between NEM Online and the manufacturers who were the ultimate clients. A project 19 group was assigned to build a Referral System for assistance providers, describing their organizations and providing sufficient information for a manufacturer or other assistance provider to decide if the organization described was likely to meet her need. The goal of the NEM Online Referral System was to catalog all the organizations that assist Michigan manufacturers and make that information available in a searchable database mounted on the Web. The Referral System was an electronic version of the directory published by NME in 1995 based on data collected by the Industrial Technology Institute in Ann Arbor, Michigan. The Referral System would be part of the NEM Online Web site, with individual Web pages for each assistance providing organization, its own search engine, and links to Web sites related to the assistance providing organizations in the system. The Referral System differed from the directory in its focus on describing the services provided and profiling the organization itself, including the mission, client industries, and published information of the assistance provider. The Referral System drew its intended clients from the NME directory, the membership directory for the Michigan Economic Developers Association, and a list put together by the Michigan Jobs Commission. While the combined lists had over 500 different names, it soon became apparent that many organizations had more than one person listed, some organizations were not really assistance providers in the sense of providing direct services to businesses, and other organizations were a part of a larger organization. Estimates of the true number of the assistance providing organizations in Michigan started at 450 and eventually dropped to between 300 and 350. These 300 or so organizations were thought to be the key to disseminating NEM Online to manufacturers, and NEM Online’s resources were developed based on their perceived needs and desires as well as those of manufacturers. Students were hired as Referral System representative to perform telephone interviews with assistance providers to create profiles for their organizations. The first calls took place in February of 1996, and the first Web pages appeared on the 20 Web in April that year. By January 1998, there were 296 profiles of assistance providing organizations in the Referral System. Communication and participation in NEM Onlinp While the Referral System was the most obvious connection between NEM Online and assistance providers, there were several other ways to participate in NEM Online. At the most basic level, assistance providers could use the NEM Online database as a resource for answering client questions and finding solutions to problems. This connection was to be facilitated by the training program designed by the common methods and training project team to teach assistance providers about the Internet, the Web, and NEM Online. In addition, NEM Online initiated eight pilot projects that directly encouraged assistance providers to contribute their expert knowledge to the NEM effort by suggesting criteria for evaluating resources, topics to search for on the Internet, and writing abstracts describing the contents and quality of information found at other Internet and Web locations. These assistance providers were expected to play pivotal roles on the second stage of diffusion, bringing NEM and its services to small and medium size manufacturers in Michigan. There were several NEM Online initiatives that were never completed. NEM Online originally planned to work with MERIT and other Internet access providers to provide assistance providers with electronic mail accounts and Internet access. This effort was halted in 1995 due to the number of commercially available Internet access providers that were suddenly available.6 Similarly, NEM Online efforts to develop a "one disk"7 software solution so that assistance providers could easily install and immediately 6 NEM did continue to work on building up the state’s telecommunications infrastructure in the northern lower peninsula and the upper peninsula of Michigan, where there were fewer Internet access nodes. 7 The "one disk" label was intended as a symbol of the ease with which the access could be activated; the number of disks actually involved was less important than the ease with which their contents could be installed. 21 use their new Internet access were halted when online service providers such as America Online began mass mailings of Internet connection software in 1996. NEM Online was initially successful with designing and hosting Web pages for any assistance provider who had manufacturing information to share with other assistance providers and manufacturers. Early takers included the Michigan Manufacturer's Association, the Michigan Economic Developers Association, the Michigan State University (MSU) Office of Intellectual Property, and the International Business Center at MSU. The page listing these resources originally included the NEM Online-created home page for the Michigan Jobs Commission as well, but political strife led to its replacement with an independent home page. NEM Online changed the page title from “NEM Online Mounted Home Pages” to “Home Pages of Supporting Organizations” but continued to list the Michigan Jobs Commission on the page. This change was solely to placate the Michigan Jobs Commission, as every other page listed was put up by NEM Online. However, by early 1997 the principal investigators of NEM Online decided that designing other organization’s Web sites was too much work for too little return. The most significant aspect of NEM Online’s communication with its intended audience is its passivity. The Referral System representatives were the only NEM Online staff who ever talked to assistance providers. NEM Online’s leaders attended many meetings, but they invariably met groups such as the top executives of associations, private database sales representatives, community college presidents, and the experts of partner organizations such as the Industrial Technology Institute. While some marketing was done at conferences such as the Society for Manufacturing Engineers and through presentations to associations such as the Michigan Economic Developers Association, there was no systematic plan. This was in large part due to the nature of the funding granted by the federal government; the marketing of the MIEP communications system was not funded, and the guidelines for expenditures are tightly connected to the tasks set forth in the funded proposals, which did not include marketing. 22 The options clients had for feedback were similarly limited. Interested assistance providers could contact the Referral System staff, or use the electronic mail buttons on a Web page to send electronic mail to a particular account (e.g., prestons@pilot.msu.edu) or use the feedback function on the NEM Online home page to communicate their evaluation of NEM Online. The feedback form asks users to note their name, their organization, the nature of the information they sought, how they heard about NEM Online, and evaluations of NEM Online as an information resource, both by itself and in comparison to other resources. While the feedback function was intended to improve the interactive capabilities of NEM Online by providing valuable information about who was interested and the advantages and disadvantages they perceived for NEM Online, responses were infrequent and often incomplete. NEM Online and virtual organization The goal for MIEP was to create a virtual organization of assistance providers at the local, regional and state levels to increase the global competitiveness of Michigan’s manufacturers. In contrast, NEM Online focused on building a resource that would make a much wider range of resources available to assistance providers. The virtual organization would not be a structure directed by NEM Online, but rather a doorway to a global community of resources across a range of forms. The focus of this dissertation is on the existing patterns of available resources for manufacturing assistance providers, as perceived by them, observed from the perspective of a NEM Online employee responsible for recruiting them for the Referral System. As virtual organization involves both cognitive and social structural elements, a theoretical approach to understanding how it takes place must integrate both. Chapter Two presents the elements of the theoretical perspective for this dissertation and poses theoretical research questions, while Chapter Three describes the methodology and presents several emergent research questions that must be answered to provide a context for answering the theoretical research questions. 23 Chapters Four and Five present detailed results of the investigation, with Chapter Four answering the emergent research questions and Chapter Five the theoretical research questions. Chapter Six applies the theoretical perspective to explain the results and evaluate the utility of the theory for understanding virtual organization, and concludes with applied lessons for NEM Online. 24 CHAPTER TWO THEORETICAL REVIEW The fundamental issue of this dissertation is how communication patterns are created and maintained. Chapter Two addresses the theoretical linkage of individual (micro) systems of behavioral production to each other in an evolving structure (macro), focusing on (1) individual cognition, (2) memory and (3) social structure. Two complementary issues require explication: How do people recognize other communication partners and resources that are not proximate as available for communication, and how do people organize their partners and resources into manageable patterns? The remainder of this chapter will establish a framework for exploring these issues and derive targeted research questions from several relevant theories. To address these issues, this dissertation will draw on several theoretical perspectives. First, a social constructionist approach necessary to accurately represent this model will be set forth. Second, a cognitive model of dyadic interaction adapted from cybernetic control theory (Carver and Scheier, 1990, 1982: Carver, 1979) will be used to demonstrate the micro processes determining individual level action. Third, the processes by which repeated patterns of interaction become proximate and virtual structures influencing future interactions will be described using structuration theory (Turner, 1988; Giddens, 1984, 1981, 1979). SOCIAL CONSTRUCTIONISM We may conceive of three realities. One is an external reality capable of stimulating us in simple ways but of which we know nothing other than our interpretations of those stimuli. The second is a reality formed of those interpretations of simple stimulation, an experiential reality 25 representing external reality so persuasively that we seldom realize our inability to verify it. The third is a universe of integrated interpretations, our rational reality. The second and third, of course, blend into each other. (Stake, 1995: 100) The aim of research is not to discover #1, for that is impossible, but to construct a clearer reality #2 and a more sophisticated reality #3, particularly ones that can withstand disciplined skepticism. (Stake, 1995: 101) Social constructionism is a qualitative perspective that emphasizes the importance of human creativity in the development and evolution of the second and third reality. While social constructionism does not reject the idea of the first reality independent of human observation, the second and third are considered the proper focus of social scientific research. For example, if television watching is the second reality, the first is the effect of certain patterns of colors, shapes and sound emanating from the screen on the human beings in its presence and the third is the programming preferences of different social groups of people, such as socioeconomic classes. The interesting phenomena occur in the second and third reality where concepts are recognized and tested; the first is concept-free, and thus difficult to discuss in terms that matter to people. In this perspective, the focus of empirical research is shifted from attempting to uncover the first reality to explaining the origins and maintenance of the second and third. This section will describe the social constructionist approach as it has been applied in three bodies of research, and demonstrate how the case study approach of Stake (1995, 1986) and others complements its assumptions. The case study approach takes a single person, organization, or initiative and emphasizes its unique characteristics in developing a narrative account of what happens over time to the subject of the case. Stake (1986) studied an evaluation project that used a case study methodology to raise awareness of how “it may be possible for a well- reasoned social science orientation to work against, even while also working for, the public interest.” (Stake, 1986: xiii) Stake does not make claims about all projects in the 26 public interest, nor does he argue that his concerns are important for all case studies. Instead, he uses the form of a case study to raise a complex question and provide an illuminating example. The approach is not just a method that can be applied to itself. A case study approach entails a philosophy that there is value in learning as much as possible about one case, and that numerous tools can be brought to bear for the same purpose so long as the researcher keeps in mind that it is never physical reality that is being studied, but what humans make of it. G. Sjoberg, Williams, Vaughan and AF. Sjoberg (1991) describe the advantages of a case study approach as follows: The advantage of case studies (as we perceive them) is that researchers who utilize them can deal with the reality behind appearances, with contradictions and the dialectical nature of social life, as well as with a whole that is greater than the sum of its parts. The case study approach that takes into account these kinds of assumptions can...provide us with fundamental sociological knowledge of human agents, communities, organizations, nation-states, empires, and civilizations. (G. Sjoberg et al., 1991: 39) While grandiose, the terms in which this claim is made - “the reality behind 99 ‘6 appearances, contradictions and the dialectical nature of social life” - suggest Stake’s (1995) distinctions between realities. More importantly, by making the claim in these terms, G. Sjoberg et al. (1991) call attention to the explication of complex relationships as the goal of the case study. While logico-deductive approaches also make such claims, Stake (1986) notes that cases are often more complex than the quantitative measures used to assess them. In his conclusion, Stake describes the reactions of the program staff being evaluated to the evaluators Stake himself was studying: They [the program staff] said the evaluators’ thinking was just not compatible with their work. They had acquiesced earlier on the matter of standards because they, too, thought CIS [the program] would accomplish them and because they had no sense of the technical difficulty in using common criteria with students having greatly uncommon problems. (italics in original; Stake, 1986: 159) 27 It is critical for any research effort to understand the specific situation in which data will be collected so appropriate measures can be developed. Stake’s (1986) insight is the recognition that some situations are best described by preserving their unique flavor rather than reducing it to elements when the whole is greater than the sum of its parts. Whenever the research focus involves issues at multiple levels of abstraction, a case study approach is justified. Stake (1995: 99) lists a social constructivist perspective‘ among his seven requirements for an effective case study. In doing so, he cites the primary advantage of a case study: Case studies provide insight into the world of the people being studied in a particular case through immersion in their world. This is different from traditional logico—deductive approaches, where the researcher is oriented towards uncovering the “truth” about a situation by taking careful account of subjective bias on the part of himself and the population being studied (Hempel, 1966). A case study is useful precisely because it rejects the ideal of a single truth and attends to the subjective experiences of the people involved in the case of interest (Stake, 1995: 103). That is, a case study allows researchers to explore the many “truths” held by different actors, including the researcher, without privileging one of them a priori. Keeping track of subjectivity remains an important goal, but instead of being parsed out of the data it is the data of greatest interest. This is the right lens through which to investigate virtual organization, because virtual organization involves issues of cognition, social structure, and perception that cross multiple levels of abstraction and are best viewed from many perspectives to best capture the whole. 1 Social constructivism and social constructionism are different in their ontologies but quite similar in their epistemologies. In essence, a constructivist argues that the individual creates meaning, while the constructionist argues that meaning is only social, with the individual as one of many participants in its creation (Baraldi, 1993). Both views focus our attention on the social nature of the reality we perceive and cope with everyday instead of attempting to discover the one “true” reality. 28 There are three bodies of literature relevant to this section. First, the social constructionist perspective explicated by Gergen (1985) focuses on how philosophers of science have rejected the ideal of scientific objectivity for the physical sciences, and makes the argument that social scientists should follow suit. Second, scholars of communication have rediscovered the need to focus on the relationship instead of the participants in communicative acts. Third, second cybemetics2 researchers have applied the recursive nature of cybernetic systems to those cybernetic systems, making the researcher part of the system being studied along with its other members. Together, these three perspectives point social science theory and methodology in a social constructionist direction. The following paragraphs will describe the four key assumptions of social constructionism. The first assumption rejects the Cartesian3 dualism between subject and object as misleading and of limited utility, as it falsely separates the world from our experience of it. Knowledge requires a knower with a theory in mind (Gergen, 1985: 266). That is, knowledge is shared between knowers, and thus only exists so long as it makes sense to a community of actors, whether they are defined by proximity, culture, or shared topical interest (e.g. invisible colleges; Crane, 1972; Price and Beaver, 1966). The philosophy of science literature supports this argument; consider Kuhn (1970) on scientific knowledge as paradigm, Hanson (1958/1985) on theory laden knowledge, and Feyerabend (1975/1984) on the excessive rigidity of the scientific method when applied universally. Under a social constructionist research paradigm, actors are situated in the world and approached as situated interpreters of their own everyday coping (Stewart, 1992: 340) using research methods that illustrate knowledge as a way of coping with, 2 Also known as the ecosystemic perspective (Fourie, 1995; Ingamells, 1993). 3 Rene Descartes (1637/1986) argued that mind is distinct from but connected to the body. Because humans have minds that command their bodies, people can be subjects within a world of objects. This qualitative difference between the subject who perceives and the object that she perceives is the foundation of natural and social sciences. 29 instead of representing, reality (Bochner and Ellis, 1992: 171). For this reason, Stake (1995: 99) argues that of all the roles a researcher can take, the role of interpreter, and gatherer of interpretations, is central. Researchers who take a social constructionist view acknowledge that they are a part of what they observe, and factor that into their accounts of behavior to create thick descriptions“ (Geertz, 1973) and explanations specific to the case. The second assumption of social constructionism is that relationships are more important than individuals or objects (Luckhurst, 1985). Baxter (1992: 335) points out that memory and identity are socially constructed via ongoing interaction, and calls for a focus on the process of fluid dynamics involved in the dialectic of interdependence and independence. Montgomery (1992: 470) argues that social phenomena are defined by the relations between characteristics rather than the characteristics themselves; no one is an island, and thus the web of social relationships in which interactants are embedded is crucial for understanding the meaning of an interaction. Bochner and Eisenberg (1987) emphasize the importance of family communication patterns to understanding interaction styles and the patterns of indirect influence that exist within groups. Atkinson and Heath (1990b: 166) argue that the great virtue of cybernetic thinking is its refusal to accept the certainty of any explanation and the resulting focus on interconnections in all aspects of life. Social constructionism requires a focus on social interaction and structures to explain the origins of meaning for individuals. That is, patterns of social relationships are important to explaining the creation of knowledge. As Gergen tells us: Communication requires the coordinated actions of at least two persons....What we take to be knowledge bearing propositions are not achievements of the individual mind, but social achievements. (Gergen, 1989: 472) 4 Le, a detailed account of actions taken from the perspective of the actor. 30 Gergen (1989) goes on to present a focused program of research that focuses on interaction instead of action, explains the construction of knowledge through attention to discourse, and changes the unit of analysis from the individual to the relationship.5 The third assumption derives from the second: Research must focus on the recursive nature of the system being studied. The reflexivity of social experience is a requirement for intelligible approaches to interpersonal communication, not an option (Montgomery, 1992; Baxter, 1992; Bochner and Ellis, 1992). McNamee (1988: 60) takes interconnectedness as a given, and emphasizes the need for research to explore the punctuation of cybernetic cycles. She recommends the use of circular questioning, asking the various individuals in a group context to interpret the statements of others so that speakers can see how their actions are interpreted by others. McNamee (1988: 60) states, “all punctuations are equally logical within the frames that different people use.” Lewis (1989) posits three needs for cybernetic research in family systems: (1) the need to draw boundaries for the self and the social systems, (2) establishing bridges between the individual and group systems that do not violate each one’s autonomy, and (3) modeling a process of change that is isomorphic to the different systems. These needs are based on his argument that family and self are conceptually distinct, yet mutually influence each other and have similar structures. The difference is in their levels of abstraction, with the implication that a researcher must choose which system he will limit his sight to (Lewis, 1989: 71). The fourth and final key assumption of social constructionism is the importance of the researcher’s own experience. Bochner and Ellis (1992) report on how they wrestled with the decision whether or not to have an abortion, and use that experience to identify the key components of their interaction that shaped their individual and joint experience 5 Misra and Gergen (1993a) apply this argument to cross-cultural research, arguing that culture should be the central focus on inquiry because it is “a historically situated, collective product, constituted by the values, beliefs, perceptions, symbols, and other humanly created artifacts which are transmitted across generations through language and other mediums.” (Misra and Gergen, 1993a: 226) 31 of the decision making process. In describing the absence of a widespread consensual view of the world, Stephen and Harrison (1992: 164) argue that the development of intimacy between individuals can only be understood if relational histories are approached anthropologically, by researchers using their own rooted perspective as an implicit foil for understanding the actors’ experiences. That is, the researcher’s own experiences as a person and as a researcher are explicitly stated and compared to those observed and elicited from respondents. These examples suggest a transformation of research methodology to respect the experience of subjects and researchers as a source of rich knowledge about human experience. Following Gergen’s 1985 paper, there have been many attacks on social constructionism.6 Two arguments in particular require a clarification: First, the argument that social constructionism cannot be proved, and is therefore not a valid theory, and second, that social constructionism entails a complete lack of consistency in actors’ use of symbols. First, the argument that social constructionism cannot be proved (as made by Nettler, 1986) is irrelevant. Social constructionism and behaviorism have different ontologies and epistemologies; they apply different values to the world, making them different paradigms, not just different theoretical perspectives (Kuhn, 1970). By definition, paradigms are incommensurable: Any attempt to judge one paradigm by the standards another betrays an inherent bias in favor of the paradigm whose values are presumed to hold for both. Thus, social constructionism cannot be evaluated by behaviorist standards, any more than behaviorism can be evaluated by social constructionism.7 It is important to note that social constructionism is not a replacement for behaviorism as a research paradigm (Gergen, 1986). Instead of trying to prove one at 6 For Gergen’s responses, see Gergen, 1986; Gergen, 1989; Gergen, Hepburn, and Fisher, 1986; and Misra and Gergen, 1993a and b. 7 In addition, empiricism cannot be applied to itself: How do you prove the world exists independent of your perceptions when all knowledge of the world is received through the physical senses (Hume, 1748/1977)? 32 the expense of the other, we should focus on the intelligibility of the two perspectives in trying to account for the production of scientific knowledge. A paradigm is not the truth, it is simply a system for interpreting the World that allows its proponents to answer the questions most important to them. Therefore, what is most important in deciding between the two paradigms is not which is true but which one can answer the questions being asked. This is Stake’s (1986) point: The wrong approach will fail to answer the questions in a satisfactory way. For some issues, one approach will be superior to others, but this does not mean those other approaches are inferior in general. The questions being asked about a specific situation dictates what methodology is appropriate, not the other way around. The second charge often leveled against social constructionism is that it grants actors complete freedom to use whatever symbols they want. The logical extension of this is a world where no two actors can understand each other, because the universe of potential symbols is too vast for any two actors to find common ground by accident. The error lies in assuming that social contructionism requires actors to seek common ground for every interaction. While social constructionism rejects the idea that there are a priori constraints on what symbols can serve as referents for what objects, there are still constraints on meaning that arise from continued interaction with similar others. Actors constrain the possible range of meanings voluntarily to make their world more predictable and thus better meet their social goals (Gergen, 1986). Thus, the question is not whether constraints exist (obviously they do) but rather their origin in social cognition and interaction. To answer this question, the remainder of this chapter will examine the connection between individual behaviors and social meaning, addressing individual cognitions and behavior production first and the social structures between individuals second. 33 COGNITIVE THEORY All cognitive theories posit the existence of a sentient actor whose behavior is governed by the meanings she gives to the environment and her own actions. Greene (1995, 1984) argues that all communication ultimately rests on the cognitions of an individual, as relevant to behavioral production (output) in a given situation. While it is easier to understand many phenomena at higher levels of abstraction (e.g., network models of social influence and ethnographic accounts of cultures), every act is performed by an individual who decides to enact a behavior based on her mental representations of events. Thus, cognitive theorists question the validity of any model that denies the importance of the actor's mind to communication and require that any model of communication must be framed at the individual level if it is to be generalizable to all levels (Greene, 1984). However, this is not to say that the actor is independent of her environment and other actors; the content of the actor’s mental representations is a social construction, and her actions are granted meaning by others independent of the original intent. An actor's interactions with the environment are moderated by mental operations performed on the actor's cognitive representations of the environment (Hewes, 1995). But it is not the mental decision of the actor that determines what others perceive, but the physical stimuli they must decode for meaning. The actor's cognitive representations and processes are driven by the interaction of what Hewes (1995) calls actor and design goals. Actor goals are the conscious or unconscious wishes of the actor, while design goals are the genetic legacy of evolution; design goals include biological impulses to seek a mate, protect one’s children, and pursue status in a social group. This brings us to the requirements for a cognitive theory. Greene (1984) names three: structure, process, and content. Because goals differ in the import they hold for actors, structure is necessary to distinguish between them: the goal of satisfying one’s hunger cannot be directly compared to the goal of a promotion. The 34 production of behavior is not instantaneous or free of inputs, but is a process. Without content, the structure and process of the mind would be empty and void of meaning. Cybernetic control theory (Carver and Scheier, 1990, 1982; Carver, 1979) provides us with the structure and process we need to understand the interaction of goals and behavior. The structure is termed a hierarchy of control; the process is termed a negative feedback loop. Together, they set the foundation for understanding cognitive organization. Carver (1979) initially proposed the negative feedback loop as the process by which people decide whether and how to act, and used it to explain the direction of human behavior and regulation of behavioral intensity. _Inte_nsity is the collection of muscle tensions whose aggregate is perceived as behavior (Powers, 1973). As the outcome of this loop is behavior, Carver and Scheier (1990) termed it an action loop. Figure 2.1 displays the component steps in the action loop. The action loop begins when a person (call her Emma) focuses her attention on her behaviors, a state known as self- focused attention.8 First, Emma perceives the current state of the environment; for example, Emma is concerned about a vote on campaign finance reform in the US. Senate that took place earlier in the day, but she just arrived home from work and missed the evening news. Second, this perception is compared to a referent value that marks how Emma wants the world to be via a comparative function, such as staying informed about campaign finance reform. If there is a difference between the desired situation and the perception of reality, then Emma will select a behavior to enact to reduce the discrepancy. Emma wants to know what happened, so she can call a friend and ask if he heard the vote’s outcome, log on to the World-Wide Web and search for a record of the vote, or turn on the television and watch an all news channel until they mention the vote’s outcome. The feedback loop is negative because people act to reduce discrepancies, not 8 This state is induced by placing a mirror in front of the subject during the experiment (Carver, 1979). 35 to enhance them (which would entail a positive loop). If there is no difference, then Emma will have no basis for action, and so will not act to change that perception; of course, there may be other perceptions that motivate her to act. If Emma heard on the radio driving home that the reform bill was voted down, then she doesn’t need to seek out the answer, but might be angered enough to write a letter to her representative. Third, when a behavior is enacted, it has an impact on the environment. This impact is mediated by external forces such as others' actions and situational variables. The loop then comes full circle, as Emma assesses the consequences of her action by again focusing attention on the current situation. If the cable news program doesn’t mention the vote for an hour, Emma will put off dinner and not practice her French for an upcoming trip to Montreal. The action loop ends when Emma’s perception matches her referent value or when she becomes frustrated with the lack of progress in reducing the discrepancy between her actual and desired state. After a half hour of news she doesn’t care about, Emma may turn off the television and log on the World-Wide Web to search for the outcome of the vote. Once she finds out that the vote was against the reform bill, her discrepancy is satisfied; of course, she may well be motivated to act as a more significant discrepancy, e.g., the gap between the power of special interests and the average citizen to gain the attention of politicians for their concerns, is not being reduced. INSERT FIGURE 2.1 HERE The action loop just described is a simple model for cognitive processes. Yet the same idea holds across many levels of abstraction; thus, it is an example of what Greene (1984) terms an elementaLlconstruct. terms that can be combined with others to form representations of more complex phenomena. The goals actors strive to obtain are not identical in either their importance to, or the expectations they invoke for, actors. Carver and Scheier (1982) explain this by organizing actors' negative feedback loops in a 36 Referent Value Comparator Perception of Environment /\ Impact on Environment Behavior to Reduce Discrepancy fl External Forces Figure 2.1 Negative feedback loop (from Carver and Scheier, 1990: 20). 37 hierarchy of control. Building on Powers (1973), Carver and Scheier (1982) posit a nine level hierarchy of self-focused attention. INSERT FIGURE 2.2 HERE The most concrete action loops involve the muscle tensions (intensity) whose aggregate is the observed behavior, while the most abstract involve concepts about one's self as a moral and social entity (system concepts) that provide an overall framework for assessing one's actions. Each level of this hierarchy is distinguished by its time scale for discrepancy reduction and its emergent properties, most importantly the reference values for the levels immediately below the current level of focus. As the actor's focus becomes more abstract, discrepancy reduction takes more time as greater numbers of concrete behaviors (the lower levels of the hierarchy) are required to complete the loop. Ascending the hierarchy, more emergent properties can be recognized, such as contextual attributes for alternative behaviors when the stimulus indicating a discrepancy is from a particular person. Thus, the highest levels of the hierarchy are most concerned with essential features of self, which define the person's referent values for the more concrete levels and so set the referent values for the levels whose goals must be met if the higher level goals are to be achieved.9 The content of human memory organized in the hierarchy of referent values is drawn from each person’s previous experience. When we act, we are always minimally operating at the level of muscle tensions and evaluating their effects on our environment. But this is not the level at which we focus our attention. Having mastered the behaviors at the lowest levels of abstraction, we only attend to their enactment when our progress is impeded. However, we are also not always attending to the highest levels of control, 9 This is due in part to the faster cycles of the more concrete action loops; many cycles are necessary at the intensity level to create a perceptible effect at the program level. 38 System concept 5..) .3... K9 K9 K9 fl a r) 1 V Program V Relationship 1 Sequence 1 Transition 1 Configuration 1 Sensation <- _ _ .