Part I: A Literature Review

Writing a “history of communication theory” and then forecasting where the field is headed is a daunting endeavor. Massive volumes of information have been written over the years, filling library shelves and electronic storehouses. This attempt will, most likely, add nothing new, except my personal perspective in telling the story. Part 1 of this project will serve as a backwards glance, first at the history of communication research, followed by a review of pertinent theoretical development. Looking back to understand the maturing of the field will enable a forward look as well, and Part 2 will be a crystal ball gaze into the future of the discipline. To accomplish this goal while maintaining a reasonable contribution:length ratio requires that some areas receive only broad brushstrokes in comparison to others. Whenever possible, the reader will be directed to review articles or the original work for more comprehensive information. As the journey begins, let’s start with the following admonition, offered under similar circumstances during the 2005 Presidential Address to the International Communication Association:

Historical perspective

The study of communication has been marked, throughout its history, by a self-proclaimed “ferment in the field.” While the ferment has varied according to the emergent themes of the times, researchers have struggled, from the origination of the discipline, to define communication in a meaningful way. This inability is perhaps due to the breathtaking expansiveness of the term “communication.” Nilsen states that “The meaning of the word ‘communication’ is at once both clear and obscure” (1970, p. 295). Often, how the researcher defines communication reflects either a scientific or humanistic approach. Griffin highlights this divergence in his text, A First Look at Communication, metaphorically describing the twin rivers of science and humanism converging to form the full stream of communication research (Griffin, 1997). In support of this picture, he points to the fact that one invited chapter contributor, engineer Claude Shannon, defines communication as the “transmission and reception of information,” a rather mechanistic perspective. Griffin then contrasts this approach with that of I.A. Richards, who defines communication from a more philosophical perspective as “the generation of meaning.” So, what is communication? If we can’t define it, can we study it? And if we study it, will we ever agree on outcomes if there is not agreement about this common point of reference? These basic questions have provided a contentious “point-counterpoint” throughout the years, so let’s see what the “ferment” is all about by looking at a small sample of articles that embody some of the challenging issues.

An early attempt to characterize the study of communication illustrates the scope of the controversy:

He repeatedly uses the term “ambiguous” as the best descriptor of communication throughout the article and suggests that part of the problem is that, “There are probably more theories of communication afloat today than there are theories of anything else whatsoever,” (Clevenger, 1960, p. 169) referring to the mass/interpersonal division as a case in point. A polite but firm commentary attached by the associate editor as a postscript to the article takes exception to the repeated use of “ambiguous” and “fragmented,” a harbinger of the then and future debate. One decade later, Dance (1970) was still searching for the communication headwaters as he examined the “multitudinous definitions of communication.” His content analysis of pertinent literature produced a master list, comprised of 4560 words/tokens categorized into 2,612 types. From this pool, he classified thirty different terms, providing the basis for ultimately 15 “distinct conceptual components” (Dance, 1970, p. 204) He parses these 15 concepts into a final triad of competing definitions which, by nature of their “conceptual cleavage” (1970, p. 208) he claims both directs and complicates theoretical development. Wiemann et al. (1988), at the passing of yet another decade, are still asking the elemental question “Who are we?” to the far from nascent field of communication researchers. The authors conclude that ongoing fragmentation results from ill-defined distinctions between mass and interpersonal communication scholars, suggesting that “A reasonable reading of the history of Communication is that it was, at its outset, a discipline that did not have the institutional trappings necessary to bring its adherents together” (Wiemann et al., 1988). A few years later, Craig (1993) asked Why are there so many communication theories? in his struggle to understand the lack of coherence within the field. He answered, at least in part, when he later concluded that it is not an “identifiable field of study” (Craig, 1999, p. 119), arguing that multidisciplinary roots are a main cause of the incoherence, and suggesting that survival requires a major overhaul. He notes that unlike most other disciplines, communication theory has developed out of seven different traditions, and is thus without a universal or unifying foundation. He identifies these traditions as rhetorical, semiotic, phenomenological, cybernetic, sociopsychological, sociocultural, and critical, emphasizing that each one views communication in theoretically different ways (Craig, 1999). The historically unresolved “theory wars” make more sense in light of these facts, as does the often hostile reception to attempts at interdisciplinary work (Berger & Chaffee, 1988; Krippendorff, 1993; Rogers, 1999; Wiemann et al., 1988).

Whether the common goal is believed to be cohesion or coherence, it is difficult to trace the path of communication research, in part because “communication happens” in disciplines as diverse as sociology, psychology, literature, mathematics and engineering (Craig, 1999). Further complicating the matter is the fact that a number of content surveys (e.g. DeFleur, 1998; So, 1988) indicate that much of the research consists of “one-shot studies without a theoretical trail” (Kamhawi & Weaver, 2003). Bryant and Miron (2004) capture the debate in a systematic content analysis covering the years 1956-2000, which they metaphorically describe as akin to doing live-feed weather reports on television immediately after the storm passes. They acknowledge that rapid changes in mass communication research seem to prevent an accurate chronicling:

The preceding articles are representative snapshots of the critical introspection to which the field of communication study has subjected itself, literally from its inception. The roots of this continued scrutiny might be traced back to the four predominant “schools of thought” credited with impacting the “birth” of the young field of communication theory. Each of the schools is marked by a different theoretical bent which impacts both interpersonal and mass communication research today, in ways both overt and subtle. I will first briefly consider the schools before moving on to the “massive body of knowledge” (Bryant & Miron, 2004, p. 663) they spawned.

Historical schools of thought

The Chicago School: Noted for humanism and pragmatism, the Chicago School evolved from the first department of sociology in the United States in 1892. The humanistic perspective of the Chicago School combined social investigation with social reform. The locating of social facts within contextual layers of time and space marked the Chicago view (Abbott, 1997). The pragmatists were headed by John Dewey and included an impressive list of scientists, including the philosopher George H. Mead, who made significant contributions to social psychology, planting the seeds of what would later flourish as symbolic interactionism (Abbott, 1997). C.W. Morris, most noted for advancing semiotics, is an other such scholar whose work reflects the influence of immigrant German scientists and their practice of logical empiricism (Bryant & Miron, 2004).

The Vienna Circle: Started in Vienna in 1922 as a group of scientists and philosophers attempting to objectify a sound basis for scientific inquiry, they developed a series of ideas that were later collectively termed logical positivism, or empiricism (Passmore, 1967). Logical positivism is a complex blending of philosophies that purports true knowledge comes only from logical reasoning and empirical experience. The “verifiability principle” of Wittgenstein was foundational in the movement’s attempt to reconcile the disciplines of science and philosophy through the use of symbolic logic (Hunt, 1991). Any doctrine or theory was considered to be “cognitively meaningful” only if there was a verifiable procedure for conclusively determining whether it was true or false. While many argue that logical positivism is all but dead (e.g.Agger, 1991; Demers, 1998), it has “left a legacy behind” (Passmore, 1967, p. 54) that continues today in the analytic sciences, despite being treated as a corpse within much of the humanities.

The Frankfurt School: Started in Germany in the 1930’s as the Institute for Social Research, the Frankfurt School is associated with the development of critical theory. Based on a revised Marxist philosophy made congruent with the emerging capitalist system of the twentieth century (Agger, 1991), this group focused on culture and technology (Kellner, 2007). Themes of ending domination and authoritarianism are the cornerstone of a theoretical legacy that vehemently fought against positivists (Scheepers, Felling, & Peters, 1990), whom they perceived as purveyors of “the most effective new form of capitalist ideology” (Agger, 1991, p. 109). The Frankfurt thinkers sought deconstruction of a system which they believed was inappropriate for and incompatible with the complex study of human behavior, as it varies within changing socioeconomic and political realities (Hardt, 1986; Melody & Mansell, 1983). While they were located for a time in the U.S., returning the Institute to Frankfurt in 1950 did not diminish the lasting impact of these scientists, as the persistent and spirited debate between positivists and critical theorists in most of the social sciences demonstrates (e.g. Agger, 1991; Bryant & Miron, 2004; Culbertson, 1983; Hardt, 1986; Hunt, 1991).

Birmingham School (British Cultural Studies): Coincident with the second generation of Frankfurt School scholars in the early 1960’s, the Birmingham School developed a similar ideology, replicating many of the Frankfurt social theories, political perspectives and methodologies (Bryant & Miron, 2004; Kellner, 2007). Distinct to this school are the metatheoretical amalgamations produced by combining Marxism and a wide range of sociological and philosophical perspectives, focusing on the ability of various subcultures to resist hegemonic domination (Bryant & Miron, 2004; Kellner, 2007). Like the Frankfurt School, it is easy to see how this group has greatly impacted the field of communication theory, as both groups were among the first to study the effects and interpretation of various forms of media and popular culture on audiences (Kellner, 2007).

Public opinion research: an illustration of the historical divide

One way to highlight the historical contributions, both positive and negative (depending on your world view) is to consider the so-called “golden age” of research on mass communication. DeFleur points out in his plaintive work Where have all the milestones gone? that the 1930’s launched a fertile period where psychologists and sociologists led the way to a belief in the “magic bullet” theory of consistent, immediate and powerful effects (1998). The next surge of research resulted in a belief in selective or limited effects theory, which recognized the complexity of delineating just how the media shaped or changed beliefs, opinions, attitudes, and ultimately, behavior. There was a wealth of research that evolved from the initial development of a two-step flow hypothesis of communication and the associated construct of personal influence, such as uses and gratifications, social cognitive theory, and adoption of innovation theory, all of which will be described later. Ultimately, recognition of a middle ground of “moderate effects” has balanced the field, which now recognizes that more powerful effects might occur conditionally, under specific circumstances. For now, I will examine a single slice of early communication research, described by one writer as “the oldest of topics with which communication research has been concerned” (Comstock, 1983) by examining the 20^th anniversary issue of Public Opinion Quarterly (POQ).

The 1957 POQ issue was a retrospective look at the first 20 years of public opinion research. The issue offers a unique view of the times, seen through the eyes of a number of influential researchers, including most of the acknowledged “founding fathers” of communication research (Griffin, 1997; Schramm, 1963). Further, it is instructive as to the state of the already fragmented field that Albig, who, had for years criticized the “tangled, matted field of opinion theory” (1939), was invited to contribute. His content analysis includes scathing indictments of what he sees as a lack of theoretical integration:

Lazarsfeld is a notable exception to Albigs’ disparaging remarks. He bravely wades into the tension of the battles between those scholars emphasizing the classical tradition versus the empirical approach, amplified by the divide between mass and interpersonal communication. His article is an attempt to reconcile, or at least defuse, what he terms these “petulant antagonists” (Lazarsfeld, 1957, p. 40). Lazarsfeld points out that this conflict is common to other disciplines and believes there is a need for synthesizing the two traditions, producing “a careful analysis in the classical tradition supported by modern empirical data” (1957). Hyman (1957) agrees that this divide is real but workable, but slips in the additional wedge of the competing views of the relative importance of studying individual (interpersonal) as opposed to public (mass) opinion as a driving force in the debate. Lasswell, focusing on the political impact of public opinion work, complains that no one has developed any fundamental theories that have added to the existing body of knowledge since the French Revolution. However, he does state that changes in instruments, occurring at a “lower level of theory” (Lasswell, 1957, p. 34) have facilitated data gathering and processing through new survey techniques, which will break down the walls between “pure” and “applied” research. Gallup (1957) writes of the history of one of these new instruments in a brief but interesting history of public opinion polls, and defends using the word “science,” without quotation marks, when talking about them. He notes that reporting on the present-day social, political and economic issues is the greatest contribution of polls, not simply forecasting election results, and encourages this broader view of their value.

In an invited critique of the issue as a whole, de Sola Pool (1957) notes the restructuring of survey research as a major scientific advance, reflected by the fact that 18 of the 22 articles begin with a discussion of the role and force of opinion in the process of public decision making, but quickly shift to talking almost exclusively about polls to do so. He also remarks on the fact that “we who use it are now our own sharpest critics,” reveling in a “spirit of self-critical triumph” that pervades the issue (de Sola Pool, 1957, pp. 191-192). He does not minimize, however, the persistent tensions that prevail, in this case “counterposing polling to theory” (de Sola Pool, 1957, p. 191) He, too, enumerates the challenges of fostering a dynamic view of public opinion that encompasses a blending the interpersonal and mass approach from both an empirical and critical perspective. He claims that “the study of opinion is the study of communication” (de Sola Pool, 1957, p. 194) and applauds the trend towards quantitative research, pointing us to Schramm’s article titled “Twenty years of journalism research” which defines journalism, quite broadly, as “research on the printed media of mass communication” (Schramm, 1957, p. 91).

Schramm explains that the initial emphasis on content analysis as the primary research tool was a practical matter rather than an academic decision; incorporating quantitative research methods became possible with newly developed tools and other “facilitating developments.” The centerpiece of his article is identification of “three great currents, running parallel, sometimes spilling over into one another, but nevertheless distinct in purpose, making use of different methods, and for the most part attracting different scholars” (Schramm, 1957, p. 99). Naming Mott, Lasswell and Lazarsfeld as the “starters” of research in the three streams (history and biography, the press and society, and the press as communication), Schramm suggests that the preceding 20 year period of studying the process and effects is the “great central strand of behavioral research on the media” (1957, p. 106). His conclusion is an optimistic assessment of the next 20 years, an opinion supported by the fascinating “up-to-date report” on the hypothesis of the two-step flow of communication, written by Katz to analyze the growth, refinement, and reformulation of the initial work published in The People’s Choice (Katz, 1957).

In his article, Katz (1957) identifies the three sets of findings included in the initial 1940 voting study: the impact of personal influence, the flow of personal influence, and how opinion leaders relate to the mass media. He then details three subsequent studies (the Rovere Study, the Decatur Study and the Drug Study), each of which attempted to extend or improve the design of their earlier work. Katz summarizes the results of these initial findings, reporting that there is homogeneity among opinion leaders and those influenced, but that the boundaries separating them are relatively fluid and variable according to the examined sphere of influence. Opinion leaders are more exposed to media but are most affected by other people, supporting the role of interpersonal relations as primary communication channels and means of social support and conformity. Concluding his analysis, Katz recognizes, once again, the dilemma of the “interpersonal versus mass” puzzle: “The central methodological problem in each of the studies reviewed has been how to take account of interpersonal relations and still preserve the economy and representativeness which the random, cross-sectional sample affords” (Katz, 1957, p. 77) and predicts that future researchers “will probably find themselves somewhere in between” (1957, p. 78).

Ferment in the field: the ongoing divide

That “in between” stage is where we find ourselves today, 50 years later. Even more telling is the fact that we could change the author names and tweak a few of the key points and delivery systems, but the content of the 1957 POQ issue is just as vital today as if it were hot off the press. We say “the more things change, the more they stay the same,” and that axiom seems true as well for the ongoing impact of the historical schools of thought fueling the smoldering ferment in the field. Before turning to the main discussion of the various theoretical paths that model the story of mass communication theory, I will further examine the foundational fault lines which generate the ongoing rumblings and occasional flares.

Thayer (1963), writing about the same time as the POQ retrospective, identified what he termed in his title “some conceptual problems” with theory building in communication. He addresses what he feels is the most basic problem: the semantic challenge of what constitutes “communication.” His cursory examination uncovered 25 “conceptually different referents for this term” (Thayer, 1963, p. 219), a vast and unwieldy collection, similar to the findings of other colleagues mentioned previously, who were also searching, unsuccessfully, for a clear and universal definition. In his down-to-earth attempt to explain what was going on “out there” in the “real world,” he concludes that one’s world view guides and directs the kind of theory one might construct/articulate, and that this fact was central to present challenges. While this intuitive statement may seem self-evident, many of the recurrent discussions about “how” to do research or what communication “is” seem to neglect its simple premise. Editors of the 1983 Journal of Communication (JOC) Ferment in the Field issue, as well as the subsequent Future of the Field issues in 1993, wisely invited the participants in the battle to examine the issues which were continuing to define communication research. The result was a prodigious amount of writing by a broad cross-section of academicians, all reflecting a widespread desire to examine the scholarly roots of communication research as well as to defend and strengthen individual positions.

Bryant and Miron (2004) capture the ferment in a backhand way with their extensive content analysis, criticizing mass communication scholars for not acknowledging the epistemology of their scientific philosophy. They found that nearly three-quarters of the use of references to theories, scientific paradigms, and schools of thought in key journals consisted of “mere reference” (48%) or simply as a “theoretical framework” for the study (26%). There were only 54 total references to schools of thought during this 54 year period, an average of just one per year. The Frankfurt School (27 references) and Vienna Circle (17 references) received the majority of attention. It is interesting to note that this poor showing occurred in the same time period that yielded a count of 600 theories or paradigms that were cited a total of 1,339 times. (These values are somewhat inflated by the infamous Ferment issue.) Bryant and Miron (2004) graph the almost non-existent distribution of references to formational schools of thought, plotting a miniscule sample for any of the schools outside of the 1983 “blip” that was the Ferment issue. It appears that the odds of surviving the ferment seemed better if the adversaries could rely on and refer to the epistemological and metatheoretical traditions providing the foundation for their competing points of view.

Even a casual reading of these articles makes clear that the prevailing opinions were decidedly supportive of a critical approach to the study of communication (e.g.Curry Jansen, 1983; de Sola Pool, 1983; Ewen, 1983; Gerbner, 1983; Haight, 1983; Hamelink, 1983; Lang & Lang, 1983; Melody & Mansell, 1983; Miller, 1983; Noelle-Neumann, 1983; Rosengren, 1983; Schiller, 1983; Smythe & Van Dinh, 1983; Stappers, 1983; Stevenson, 1983; Tunstall, 1983). One representative article applauds a diversity of critical approaches that vary according to how they “fracture the simple linear causal model” and “redefine the nature of the communication process” (Slack & Allor, 1983, p. 213). Schramm (1983) brings some balance and gravitas to the issue, recalling his humorous jab that the corpse seemed “extraordinarily lively” in response to the oft-quoted Berelson quote regarding the “withering away” of the field (Berelson, 1959, p. 1). Schramm claims that “Communication study has always stimulated a certain amount of ferment” (1983, p. 6), and, unlike most of the writers in this issue, does not seem unduly concerned. He contends that “communication is always part of something” (Schramm, 1983, p. 16), and closes his paper by calling for it to be a

One decade later, the JOC published an issue styling the field as somewhere between fragmentation and cohesion, a recurrent theme testifying to the truth of the maxim that there truly is “nothing new under the sun.” The authors agree that not only is there still ferment in the field, but that major questions remain unanswered, reflected in article titles such as What are media? (Meyrowitz, 1993) or Why are there so many communication theories? (Craig, 1993) and Is ours a scholarly discipline? (Babrow, 1993). Still others saw the question as the same philosophical ones that estranged researchers in 1983, and the contempt and disdain of one camp for the other drip from the pages. For example, one author dramatically writing that the movement towards a critical approach lacks sufficient speed and breadth, seems bewildered that anyone could not see that “critical studies in communication had torn asunder the guise of scientific neutrality that had masked the ideological premises of dominant communication research for decades” (McChesney, 1993). Many of the authors in this issue also point out the practical issues that have helped drive some of the divisions, most identifying the often highly political decision of where a department might reside within a given university as the most common cause of intellectual separation. Most heartening, though, are the expressed hopes for some measure of harmony (Meyrowitz, 1993), an acceptance of the value of “promoting theoretical and methodological tolerance and disciplinary cohesion” (O'Keefe, 1993) and the now and not-yet “Golden Future Time” (Swanson, 1993, p. 191) that will be both multi-disciplinary and substantive.

The “substantive multi-disciplinary approach” remains elusive. Clearly, “how” one studies communication is based on implicit assumptions derived from that individuals world view, i.e. choosing to employ a critical rather than an empirical approach. “What” one studies and calls “communication” flows, as well, from a belief system that defines communication in a particular fashion. Perceived differences in just “what” communication “is” have led to a significant lack of interaction between the proponents of interpersonal and mass communication studies. Rogers points to the “sub disciplines” of mass and interpersonal communication, describing them vividly as an “intellectual canyon” which slows theoretical development, decreases scholarly coherence and ultimately “violates the holistic nature of the human communication process” (1999, p. 618). The exponential growth of methods and forms of communication have further accelerated the development of restricted models, methods and meanings, often serving to further complicate research questions, rather than bringing clarity. Thus, while there appears to be great awareness of the division and the historical underpinnings that have caused it, “bridging” the gap seems to be a nearly insurmountable challenge.

The Journal of Human Communication Research (HCR) addressed this concern in two 1988 publications devoted to exploring this gulf between mass and interpersonal communication. A number of invited authors wrote with significant agreement about the need to move towards greater synthesis, using a variety of methods to show how the current dysfunctional state affects the growth of the discipline (e.g. Berger & Chaffee, 1988; Reardon & Rogers, 1988; Rice, Borgman, & Reeves, 1988; So, 1988; Wiemann et al., 1988). Three of the papers detail the development of the subdisciplines, noting that their different origins help, in large part, to explain their ongoing separation (Berger & Chaffee, 1988; Reardon & Rogers, 1988; Wiemann et al., 1988). For instance, they agree that the emergence of mass communication studies in the early 1900’s coincided with the interest of sociologists and political scientists in the potential influence of new technology (television, radio, film) on mass audiences. It was psychologists and social psychologists at about that same time, however, who began exploring one-on-one interactions, from a predominantly behavioral perspective. Thus, despite a common goal, the two areas were initiated, and have remained, for the most part, autonomous and distinct.

Two citation analyses in the 1988 HCR issue point to the small volume of subdisciplinary cross-citation as evidence of a lack of communication and intellectual separation (Rice et al., 1988; So, 1988); other reviewers have found similar patterns in their analyses (e.g. Paisley, 1989; Reeves & Borgman, 1983). Wiemann et al. (1988) attribute the split, in part, to these theoretical differences. For example, they suggest that mass communication research focuses first on behavior and then studies outcomes, while interpersonal research most often considers communication behavior as the outcome of some causal antecedent. Reardon and Rogers (1988) argue that the dichotomy is a false one and exists only because of perceptual differences that are more historical and political than real. They acknowledge, however, that the discipline functions dichotomously, leading to a lack of integration in communication theory, an incomplete understanding of the communication process and predict a future struggle when attempting to study newer, more interactive forms of communication. As an attempted solution, these authors suggest one way around the dilemma is to classify by research specialty across the chasm of the mass and interpersonal divide. Creating scientific categories, such as agenda setting or persuasion, for example, would offer a flexible frame to accommodate shifting emphases within the field (Reardon & Rogers, 1988).

While these ideas seem both practical and manageable, implementation is another matter. Within this same 1988 issue, Berger and Chaffee write of their attempt to “force the field together” (1988, p. 312) and the challenges that go beyond the history and politics which other writers purport to be the root issues. They agree that effects of theoretical differences are substantial, but also argue that critics of mass communication go well beyond the “descendants of the Frankfurt school” (1988, p. 315), including just about any and everyone who is involved in the “business” of communication, which is just about any and everyone! One additional distinction they make between the mass and interpersonal scholars is the social policy involvement of mass communication that is not an integral part of the interpersonal research world (Berger & Chaffee, 1988). Thus, while most of the authors agree that the bridge must be built, how to begin construction and who will be the architects is a persistent challenge.

In a more recent article, Rogers (1999) provides an elegant overview of the issue from both a historical and practical perspective, along with supportive evidence that the divide persists. He concludes that the outcome of this ongoing subdivision has three primary consequences and suggests ways and means to overcome these deficits. First is the lack of integration in communication theory: he points to persuasion and attitude change as a place where interpersonal theories could cross-over, as in health communication campaigns. Second is the inability of one discipline to encompass the complexity of communication, and thinks a merger is in order. He refers to the term “intermedia” coined by Gumpert and Cathcart, and the “inherent interconnectedness” that implies (Rogers, 1999, p. 627). Third is facing up to the fact that understanding new, interactive technologies will need the strengths of both disciplines. He maintains that now is the time to integrate, simply because of the changing nature of the medium.

It is important to point out that almost all of the authors writing of this fragmentation at least mention, if only in passing, the fact that it has not always been this way. The most common example cited is the work of Katz and Lazarsfeld (1955) who extensively covered the pivotal role of role of interpersonal communication in mediating mass communication effects. In his narrative defense of the generations of communications research which followed this early study, Katz writes in support of Lazarsfeld and his bridge-building ways, not only between mass and interpersonal communication researchers, but to the elemental distinction between schools of thought:

Artistotle said “If you would understand anything, observe its beginning and its development.” As demonstrated, understanding the history of communication research is essential to understanding the subsequent theoretical development. I have examined the beginning – and turn next to the development of communication theories.

Communication Research: Theoretical development

The earliest years of communication research have been described as the “rise of rhetoric” (Griffin, 1997), reflecting the roots of persuasive communication in the discipline of philosophy. Most texts cite the impact of the writings of early philosophers, such as Plato, Socrates or Aristotle, as the foundation of this rhetorical emphasis on the “art” of public address. McCroskey and Richmond claim (as cited in Bryant, 2004) that “the study of communication is older than any area of current academic interest,” moving the timeline back to about 3,000 B.C. in Egypt. Others (as cited in Bryant, 2004) attribute the practical beginnings to Corax, a Greek scholar in the 5^th century B.C., who developed a method to train average citizens to publicly defend their rights in order to reclaim confiscated property. This attempt may be the earliest record of a systematic attempt to teach others to organize information and logically develop arguments. It is from this heritage that Plato, Socrates and Aristotle created theories of rhetoric that influenced the development of communication research centuries later.

Griffin (1997) suggests that early recognition and acceptance of rhetorical study as the main theoretical direction of the field developed out of the influence of a 1925 paper focusing on the “effect” of public speaking, distancing the field from the study of literature. Thus, the art of speech criticism, using the neo-Aristotelian categories of logical, emotional and ethical appeals, became the established standard, although it was considered by its practitioners to be more “art” than “science.” Because of the humanistic nature of these beginnings and the primarily qualitative approach to the study of rhetoric, the first communication scholars discounted the value of quantitative measures. As the years passed and the technology of the world changed, rhetoricians tended to dismiss new communication techniques as frivolous entertainment, and, rather than study them, most often ignored them. The separation this caused, as others turned to study the effects of media, is evident in the stand-alone speech departments housed on many university campuses today.

The field of communication research changed significantly during the World Wars in that many of the behavioral scientists who became involved in the effort to understand the effects of messages broadcast to listeners came, not from speech departments, but from the fields of sociology, psychology, journalism, and political science. The study of media effects dominated the field from about 1930-1960 and represents the original “milestone” era of seminal research and theoretical development (DeFleur, 1998). Most historical accounts paint Wilbur Schramm, as the “driving force behind the creation of communication study” (Rogers, 1994, pp. 1,7), and he is further credited with training the first generation of communication researchers to use quantitative, empirical models. Wahl-Jorgensen (2004) makes a compelling case that the origins of mass communication are both more contextually and conceptually rich than the product of a single man, proposing that the inter-disciplinary Chicago programs made vast and diverse contributions prior to fading away due to lack of institutional support. Regardless of where one assigns credit, it was this small group of behavioral scientists who, by separating themselves from the humanities and rhetorical arts, expanded and developed what many recognize as Schramm’s institutional and empirical vision of communication research. These men were the trailblazers who started down the road towards what we would, today, recognize as communication research. On that journey, the individual paths each took led to new insights and greater understanding. At the same time, they were also responsible for carving some of the first divisive ruts in that road that remain visible even today.

The study of groups

One of these key researchers was social psychologist Kurt Lewin, who came to America to escape the Nazi holocaust. Much of his interest in groups grew out of his concerns about authoritarian leadership. His empirically based research reflects his belief that social science should follow an engineering model, combining a study of the general laws that govern the universe, with a situation-specific diagnosis in harmony with those laws (Lewin, 1945). He describes his approach as one that integrates “divergent physiological, psychological and sociological facts on the basis of their interdependence” (Lewin, 1939, p. 126) in his development of a field theory in social psychology. Lewin chronicles his concerns and beliefs in an article detailing his integrative approach to social science in serving modern society:

Lewin’s notable work in group dynamics, his development of the concept of “gatekeepers” and his much-quoted statement that “there is nothing so practical as a good theory” follows others who made landmark contributions in the areas of group norms and pressure. Sherif’s early autokinetic light experiments demonstrated that individuals rely on others for guidance, and are influenced by them, even in their absence, particularly in uncertain situations. Asch investigated group pressure in unambiguous settings, finding high levels of conformity, even when individuals “knew” they were right, yielding instead to the majority who disagreed. Both of these researchers measured these effects with casual groups. The clear implication is that the effects would potentially be much greater in closer relational groupings, such as family or co-workers (Severin & Tankard, 2001). In the area of mass communication research, much of this work contributed significantly to the development of a limited-effects model. Klapper, for example, detailed five generalizations about the known effects (and their limitations,) in his book The Effects of Mass Communication, and the first two focused on what he termed “mediating factors” (as cited in Severin & Tankard, 2001, p. 263). These “mediating factors” included group processes and norms, opinion leadership and the selective nature of perception, exposure and retention.

In an extensive review of newly emerging trends in the study of “intergroup relations,” Tajfel acknowledges that while they had value, the cognitive basis of these early findings in this area “do not seem able at present to provide generalizations to interactions between groups which are powerfully determined by conflicts and value-laden social differentiations” (1982, p. 31). He concludes that the shift towards an understanding of group behavior which considers group membership and the attendant “umbrella concept,” the function of social stereotypes, processes of group discrimination and the development of social identity/comparison as more fertile ground for ongoing study. Examples of research that has evolved out of both early and subsequent work on groups includes studies on the formation of norms (Bettenhausen & Murnighan, 1985), the fear of social sanction that imbues normative behavior with power (Reno, Cialdini, & Kallgren, 1993) and the extreme form of group pressure exerted through cults (Melton, 1986).

Consistency Theory: Cognitive Dissonance

One of Lewin’s students was Leon Festinger, who developed a theory of cognitive dissonance that has endured and generated significant amounts of related research in a wide range of disciplines. Fundamentally, cognitive dissonance involves motivation: inconsistency among cognitions leads to the negative intrapersonal state of dissonance, thus motivating the individual to find and then implement a strategy to alleviate that dissonance to achieve consonance. In a review of Festinger’s classic work A Theory of Cognitive Dissonance, Aranson (1997) notes that the theory was quite powerful at the time, primarily because it mounted the first successful challenge to reinforcement theory. The behavioral approach was one of stimulus-response, and it had been used to explain, albeit not effectively, just about every social interaction, including Lewin’s work on democratic and autocratic leadership and Asch’s conformity experiments (Aronson, 1997). The classic 1959 dissonance experiment showed that people believe lies only when insufficiently rewarded, in direct opposition to reinforcement theory, which alleges that rewards would actually change someone’s beliefs and responses. From these humble beginnings, dissonance theory began to confront reinforcement theory at every level. Examples of areas where the theory of cognitive dissonance is used to explain attitude change and behaviors include decision making, buyer’s remorse, forced compliance, contradictory information and selective exposure/attention, to name just a few. Dissonance Theory, in contrast with Stimulus-Response models, provided a compelling structure to explain reactive behavioral change in a variety of situations, even allowing researchers to predict when/how dissonance might occur and the resultant response.

While it is hard today to comprehend a view of people functioning as “reinforcement machines” rather than thinking beings, the shift was revolutionary at the time, and affected multiple areas of research. The inability of reinforcement theory to explain or account for human behavior in a number of different settings, such as initiation to a group, tended to be glossed over, as it was the only theory in play at the time. Despite the robustness of dissonance theory and its successful application in numerous settings, Aronson notes that work on cognitive dissonance virtually disappeared starting in the 1970’s. He claims that although dissonance theory “revitalized” social psychology (Aronson, 1997, p. 129), societal changes, which rendered experimental deception unacceptable, were influential in the decline of its usage. The primary cause of the drop-off, however, was the move away from emotional/motivational constructs in favor of purely cognitive judgments and attributions. Renewed interest in dissonance theory is exemplified by the recent increase in “self” oriented research, which he contends falls under the “big tent” of dissonance theory. Aronson makes a strong case for the benefits of having a single, more universal theory rather than numerous mini-theories that are limited in both scope and application domain. He points to the proliferation of narrowly focused research in self-verification theory, self-evaluation maintenance theory, self-affirmation theory, symbolic self-completion, self-discrepancy theory, and action identification theory as evidence, and illustrates how each is connected to dissonance theory (Aronson, 1997). While he does not question that these other theories add some value, he does wonder if science is advanced when “we have a half dozen or more theories doing the work of one” (Aronson, 1997, p. 133), blaming this tendency on an overly-zealous embrace of analysis over synthesis and ignoring older theories as non-relevant “ancient history” (1997, p. 134). Dissonance theory is generally recognized as the most universal of what are termed “consistency theories,” all of which attempt to explain how individuals deal with the psychological tension accompanying discrepant and/or inconsistent information.

Consistency/Balance Theory

Consistency theories strive to incorporate the physiological drive towards homeostasis into the psychological domain, and a brief description of three of the major ones follows. The first is Heider’s Balance Theory. His theory looks at how individuals organize attitudes (towards people and objects) in relation to one another: being out of balance drives tension and generates mechanisms to restore balance (Opp, 1984). Second is Newcomb’s Symmetry Theory (as cited in Severin & Tankard, 2001), which extends the drive for balance beyond the individual to communicative interactions with others. The strength of the relationship and the level of agreement within the triangle of the two individuals and the object of their communication affect the accompanying degree of tension, which in turn yields varying degrees of “symmetry.” Implicit in the model, as well, is the potential resultant asymmetry in the triangular connectedness of individuals who “agree to disagree” about something. The third consistency theory is Osgood’s Congruity Theory, described by Zajonc as a “special case of balance” (1960, p. 286). The congruity model involves the attitudes an individual has towards a communication source (person) who is providing information about an object: when change occurs, it always moves toward greater, not less, congruity. Examinations of congruity theory, particularly in light of incongruous media events, identified moderating and/or mediating variables, including selective perception, selective attention, source credibility, denial and selective retention (Severin & Tankard, 2001).

As mentioned previously, consistency theory disappeared from the research landscape for several decades. Greenwald et al. (2002) maintain that this departure was due, in part, to theoretical competition that did not appreciate and develop their common ground. These authors present a “unified” theory, which they allege blends “social psychology’s most important cognitive constructs (stereotype and self-concept) with its most important affective constructs (attitude and self-esteem)” (Greenwald et al., 2002, p. 21). Hummon and Doreian (2003) suggest yet another revised balance model: both of these examples reflect a recent return to some of the earlier processing models with strong mathematical components. Once again, the historical perspective is enlightening, when we consider that Zajonc wrote years ago of the connectedness between the concepts of balance, congruity, and dissonance, all relying on the underlying notion that “thoughts, beliefs, attitudes and behavior tend to organize themselves in meaningful and sensible ways” (1960, p. 280).

Mathematical/Information Theory

The balance models developed, for the most part, out the work of Shannon’s mathematical theory of communication, initially praised for the clarity of a scheme which does not confuse information with “meaning” and focuses solely on the transmission of a signal. Severin and Tankard (2001) declare that Shannon’s theory of signal transmission, or Information Theory, is

Originating from the book The Mathematical Theory of Communication (Shannon & Weaver, 1949), the authors utilized the technology of the early telecommunication systems, and borrowed the concept of feedback and the statistical foundation of communication from the book Cybernetics (Wiener, 1948). This model depicts communication as the following flow, noting that efficient processing requires a balance between entropy and redundancy while minimizing noise:

Information source Chosen message Transmitter converts Channel-appropriate signal Receiver converts received signal Message arrives at destination

Because of the mathematical and statistical nature of the model, it is described by it’s founders as “exceedingly general in its scope, fundamental in the problems it treats, and of classic simplicity and power in the results it reaches” (as cited in Severin & Tankard, 2001). At the time, many researchers also believed it was a choice example of a theory with great potential for multi-disciplinary integration. The technical nature of information theory is appealing in that it describes the transmission of a signal, reducing communication to a relatively simple process. That simplicity, however, presents a great challenge, as Ritchie points out in his critique of multiple inappropriate applications of Information Theory (1986). Others have noted that Shannon is describing a structural system, while humans are functional systems, who change and grow and learn, potentially upsetting the probabilistic balance of the model (Severin & Tankard, 2001).

Other linear “transmission” models were developed as well, known to every student of communication as Lasswell’s basic “Who Says What In Which Channel to Whom With What Effect, or Berlo’s Source Message Channel Receiver. Both of these models are general and useful in practice, yet most academicians question their theoretical value and ability to incorporate meaning or mutual participation in the process. Gerbner (as cited in Severin & Tankard, 2001) furthers Laswell’s model, developing a more dynamic verbal model of human communication. A self-described “general” model of communication, this work identifies ten aspects of communication which he links to specific areas of study as described below:

Someone perceives an event and reacts in a situation through some means to make available materials in some form and context conveying content of some consequence.

The Osgood and Schramm Models (as cited in Severin & Tankard, 2001) break away from this linear approach to account for the more circuitous nature of communication, attempting to account for the multiple participatory roles of both “encoding” and “decoding.” In addition, these models incorporate “loops” for feedback, accumulated experience and interpretation of meaning lacking in the one-way transmission models. Westley and McLean (as cited in Fawkes & Gregory, 2000) were also influenced by Shannon, developed a sequential, non-linear model, extending Newcomb’s Symmetry Theory to incorporate mass communication, and introducing the “gatekeeping” function of the “channel,” which had previously only been perceived as a conduit. The Maletzke model (as cited in Fawkes & Gregory, 2000), unlike many of the previous models, focuses on the complexities at the communicator and receiver ends (i.e. self-image or personality.) A final model of information processing is the Schema Model developed by Axelrod (1973). His process model is relatively complex and diagrammatically reminiscent of the old “snakes and ladders” game where entry and exit in the communications flow is expedited or shunted depending on a large number of variables.

The seeming incongruity of using mathematical models to describe communication can be a rather disquieting conundrum, since “meaning” is intuitively perceived as the heart of human communication. In a 1961 article written as a contemporary review of the general models of communication research of the preceding decade, Johnson and Klare foresaw the power and reach of Shannon’s mathematical formulations, declaring that “Of all single contributions to the widespread interest in models today, Shannon’s is the most important” (1961, p. 15). It is only fitting, then, that another statistician, describing the limitations of statistics, has pointed out “The real world is limitless, however, too expansive and complex to be observed fully. Observation, then, is necessarily incomplete. In addition, perception is selective: it filters, distills and organizes” (Aneshensel, 2002, p. 24).

The role of perception

How does perception function in communication? Johnson and Klare describe communication occurring at both the interpersonal and mass level as either “face to face” or “place to place” (1961, pp. 13-14), and note that perception is influential in both. It seems intuitively obvious that one’s perceptions are integral to effective communication, but that does not make it a simple process. Severin and Tankard (2001) demonstrate that the process of perception is extremely complex. They state that assumptions, cultural expectations, moods, needs and attitudes, all impact perception, and operate as either single variables (structural/ physical or functional/psychological,) or some combination of both. It is difficult to separate perception from the raw processing of information: the mathematical models described above delineate a logical, sequential process of various cognitive steps. Schema Theory is one example of how perception can work in tandem with information processing in a mass communication setting. Graber (1988) evaluated the processing of news information, discovering that people use multiple strategies, such as the use of inferences or matching information to existing schemas for perceptual interpretation and handling information overload. She suggests that an understanding of processing information via audience schemata was neglected, originally, because of the early belief that media effects were “hypodermic” in nature, internalized in exactly the form the sender intended, which she likened to Humpty Dumpty’s disdainful response to Alice: “When I use a word, it means just what I choose it to mean, neither more nor less” (Graber, 1989, p. 151). Alice has a better understanding of media effects: there must be a signal and receiver, but meaningful communication is only possible with shared meanings, which are, in turn, affected by idiosyncratic responses (i.e. selective perception, retention or memory.) The fact that we now know that “…the meanings extracted from content depend very heavily on what viewers bring to the situation, including their biophysiological and psychological limitations” (Graber, 1989, p. 148), brings us to a discussion of the changing knowledge of media effects.

Some might liken changing beliefs about media effects to a wide pendulum arc, swinging slow and wide from the powerful effects side to that of limited effects and resting, ultimately, somewhere in the middle, to signify the current understanding of situational effects that may be relatively small or large depending upon circumstances. The initial idea that mass communication was a powerful tool is exemplified by the inert receiver at the last step of the process models: hit the target and achieve the desired effect. Often called the “bullet” or “hypodermic needle” theory, this view is based on a stimulus-response relationship predicting a strong and universal response to a message. Anecdotal observation, rather than empirical evidence, is most often cited to explain the development and widespread acceptance of this view, such as the mass hysteria following Orson Wells’ War of the Worlds in 1938. The effectiveness of Nazi propaganda and the ensuing fear it generated led to the birth of the Institute for Propaganda Analysis to identify tools used in propaganda. This group did not study the effects of propaganda: the impact was assumed to be both direct and powerful. One of the main things the Institute did was to identify the seven key propaganda devices of name calling, glittering generality, transfer, testimonial, plain folks, card stacking and bandwagon, and then instigated a public education/awareness program in hopes of inoculating the American public against their dangerous effects (Severin & Tankard, 2001). These techniques have subsequently been evaluated and found to be selectively effective, but this primitive work on the effects of propaganda stimulated further work on attitudes and behavior change, and pointed the way towards an understanding of a more limited-effects model.

Persuasion and attitude change

One of the first major programs to examine attitude change was that of Hovland and associates during World War II. The controlled experiments conducted by this group working for the U.S. Army’s Information and Education Division demonstrated that while military films could inform and mildly affect opinions about the conduct of the air war, they did not alter motivation to serve or increase resentment of the enemy. In other words, they found that the ability to use persuasive messaging to change attitudes or behavior was much more limited than supposed. Their work was instrumental in slowing the inexorable movement of the pendulum towards the powerful effects side. Other work supportive of this limited effects approach was described previously, including that of Lazarsfeld and associates, in the political arena, and Klapper’s influential book describing limiting mediating factors.

Attitude change: One-sided and two-sided messages

Concern about the effects of propaganda led Hovland’s group to study the use of “card stacking” through an evaluation of one-sided and two-sided messages, specifically to answer the question of how best to convince the troops there was still an ongoing battle in the Pacific, despite the German surrender. Using a creative experimental design, they determined that there was no difference between a one- versus two-sided version of a presentation on audience members, regardless of whether they were initially opposed to or sympathetic towards the message. Further evaluation of their results, however, indicated the complexity of attitude change, as they discovered the impact of interactive variables. They found that a one-sided message is most effective with those who are initially favorable, while a two-sided message appeals to those who are opposed. In addition, education level was a moderating variable: one-sided messages were more effective with those less-educated, two-sided with those more highly educated. At this time, the pendulums’ swing towards the more limited effects side was propelled by a growing recognition that propaganda is not invariably effective, but its impact more often reflects situational contingencies (e.g. Brown, 1958; Jarrett & Sherriffs, 1953).

Hovland realized that changing attitudes and behavior through persuasion has much broader application than the study of propaganda. He continued his systematic research at Yale University after the war ended in order to attempt to identify the variables influencing this process. Larson (1998) categorizes the resultant work as “single-shot attitude change theory” since the underlying belief was that a change in attitude would result in a change in behavior. He diagrams what this group considered an ideal persuasive attempt as moving someone up a behavioral ladder from the bottom rung of attention to comprehension, and once gaining the step of acceptance, retaining and acting upon the information. While each step matters, he states that researchers in the Yale tradition hold the “acceptance” stage to be vital, thus their focus on the factors affecting that decision. Determining why someone accepts or rejects a persuasive appeal directed them to consider four areas, described earlier by the linear processing models as the source, message, channel and receiver components.

Attitude change: Source

The first work on the issue of how perceived credibility affects one’s behavioral response was a result of the amazing success of Kate Smith, an entertainer who raised $39 million for war bonds in a scant 18 hours (Severin & Tankard, 2001). Perceived sincerity and trustworthiness were identified as the key to her success, and Hovland and Weiss (1951) devised a plan to study subject response to both a high and low credibility source to tease out the isolated effects of the source variable. Their findings supported a high-credibility source as more likely to produce opinion change on a variety of topics, identified the importance of expertness and trustworthiness, and offered a clarification of what had been dubbed a “sleeper effect.” Most of the subsequent research in this area has focused on identifying the dimensions of source credibility.

In an excellent article written from a practitioner’s point of view, Erdogan et al.(2001) provide an historical perspective, starting with Hovland’s work and his development of the Source Credibility Model. Additional research in this area includes the Source Attractiveness Model (McGuire, 2000), the Product Match-Up Hypothesis (Kahle & Homer, 1985; Kamins & Gupta, 1994) and the Meaning Transfer Model (McCracken, 1989). Much of this body of work incorporates Social Influence Theory, which has demonstrated that message receptivity is influenced by salient characteristics of the communication source (Erdogan et al., 2001). Social Cognitive Theory is a related concept in this theoretical grouping of the study of “source effects,” and will be covered in the Diffusion Theory section of this paper.

Social Identification Theory

Kelman proposed years ago (1961) that there are three distinct processes which describe social influence, each with unique antecedents and consequences. Compliance is defined as an individual accepting influence from another in hopes of receiving a favorable reaction. This behavior does not indicate adoption, but rather mimicry of an expected response in given situations. Kelman (1961) suggests that when one is concerned about the social effect of a behavior, the result is compliance, noting that one tends to only perform that behavior when observed by significant others. Identification involves adoption of a behavior out of a desire to be related, in a “satisfying, self-defining” way to another person or group. Identification takes a number of different forms in social behaviors, but several points are essential to understanding the construct. First is that classical identification involves an attempt to be like, or even “be” the other person, because the emulated one occupies a highly desirable role or exhibits attractive characteristics. This level of identification could be particularly important in areas such as encouraging new health patterns, as the goal is to help viewers to be willing to alter current behaviors in order to “be like” the spokesperson. Another form of identification occurs with the fulfillment of particular “roles,” i.e. doctor and patient, with each behaving according to expectations. In this case, the individual believes he “is” who he is expected to be, and acts accordingly. Lastly, identification may provide a means of self-definition within a group, requiring one to meet certain behavioral expectations to remain accepted. With identification, unlike compliance, the individual, both publicly and privately, believes the opinions and actions he has chosen to adopt. However, these beliefs are not fully integrated, as the behavioral response in a given situation is more dictated by external expectations and a need to play the expected role in order to remain “socially anchored.” Internalization occurs when the reason for and value assigned to the behavior are determined by the individual without external influence, requiring primarily congruence with one’s own beliefs for acceptance. Internalization also reflects a belief that the content of the induced behavior is perceived as intrinsically rewarding. One example might be accepting and acting on the recommendations of an expert because they are both relevant to a personal situation and in agreement with personal values. Adoption is often not identical to the externally proposed solution, but modified to “fit” by the individual. Kelman suggests that, compared to the viewer, the source derives much of its persuasive power base from meaningful content and credibility.

Ongoing research in the area of persuasive messaging has attempted to determine whether a high credibility source is more persuasive than one with low credibility in changing the attitudes, beliefs and/or behaviors of the viewers. To date, there is equivocal empirical evidence that offers somewhat conflicting results about the role and potential impact of high and/or low credibility sources. (Pornpitakpan, 2004). In general, sources with higher credibility seem to be more persuasive for both attitude and behavioral measures, and most studies identify the dimensions of source credibility as expertise and trustworthiness, just as Hovland found. Expertise is the perception that the spokesperson is capable of providing accurate information and encompasses the perceived knowledge, experience and skills of the endorser. Whether this is, in fact, true of the spokesperson is another matter entirely: it is the perception of the viewer that bestows the label of “expert.” Trustworthiness, the other primary dimension of credibility, is the perception that the stated assertions are valid. Like expertise, it also depends on the perceptions of the viewer and has been found to strongly correlate with “likeability” (Friedman, 1991). McGuire (2000) contends that similarity, familiarity and liking for the endorser determine message effectiveness. Similarity is the perceived resemblance between the source and recipient, familiarity is knowledge of the source, and likeability is positive affect stemming from the physical appearance and behavior of the source. Deshpande and Stayman (1994) confirmed that the ethnicity of the spokesperson affects perceived trustworthiness, suggesting that similarity also significantly enhances trust. While the interaction between source credibility and other variables such as physical attractiveness and gender have been studied, there is no clear consensus as to consistent interaction effects or degree of interaction. There are multiple opinions as to what factors comprise source credibility, their relative value, and which factors are more or less important in different situations. Scales that measure source credibility based on attractiveness, trustworthiness and expertise are available (Ohanian, 1990), even these empirically driven ratings are criticized as being unable to handle the multidimensionality of source effects (DeSarbo & Harshman, 1985).

Attitude change: message

The one-sided versus two-sided message research reviewed earlier is an example of how the message design affects attitude change. Hovland’s study of fear appeals also focuses on the delivery of a persuasive message, and he predicted that a stronger fear appeal would lead to increased attitude change, and, as a result of increased arousal, increase attention, comprehension, and acceptance of the message. Conversely, he expected that a fear appeal might have an adverse effect, caused by a protective defensive response and distortion of meaning. Their initial study (Janis & Feshbach, 1953) concluded that strong fear produces less attitude change, and is considered the gold standard. Unlike other studies, they measured actual behavioral change rather than relying on a self-reported attitude and performed a one year follow-up, finding lasting differences between experimental and control groups. Leventhal and Niles (1964) challenged this finding with a fear scenario (stopping smoking and lung cancer) that was perhaps more believable than that of Janis and Feshback (1953), whose highly effective minimal fear appeal may have reflected an inherent disbelief of the direct cause-effect relationship between tooth brushing and extreme gum and tooth disease. Successive studies suggested a potentially curvilinear association between fear appeal and attitude change, altering the view of fear appeals to an inverted U relationship. Extending previous work, Rogers’s protection motivation theory has become the dominant model of “fear effects” due to it’s robust application across a wide range of issues, ranging from earthquake preparedness (Duval & Mulilis, 1999) to prevention of AIDS (Tanner, Hunt, & Eppright, 1991) and smoking cessation (Keller & Block, 1996). There is some concern about the relative risk of fear appeals in that they may drive other emotional responses, yielding a variety of unintended consequences, such as a boomerang effect where the message is rejected because it drives too much fear (Larson, 1998). A number of authors (Keller & Block, 1996; Olson & Zanna, 1993) suggest that protection motivation theory is linked with general dual-process models of persuasion through varying degrees of message elaboration, a concept which will be discussed later in this paper.

Attitude change: channel

A succinct description of the work on the delivery channel for the persuasive message is that complexity determines the degree of response. For example, a simple message, presented through an audio-visual medium, will produce greater change than an audio-only message; the audio channel alone will produce more attitude change than printed versions of the same message. The written channel becomes highly (perhaps most) effective as the level of message complexity increases. The issue of noise was mentioned previously with Shannon’s mathematical model, which required that noise be minimized for effective communication. Research evaluating this effect most typically attempts to distract the receiver, often resulting in short-term attitude change which quickly dissipates, and is dependent upon how the receiver views the message (resonance versus discordance) (Larson, 1998). Much current research is directed at understanding the multiplicity of channels developed within the context of “new” media: Part 2 of this project will address this area more fully.

Attitude change: receiver

There is a wealth of research surrounding the characteristics of the recipient and how both chronic and situational dispositions may affect or alter the impact of a persuasive message. Receiver characteristics are generally divided into two areas: personality factors and the degree of personal involvement with the issue at hand. Personality involves elements such as self-esteem: those with high self-esteem, not surprisingly, are less vulnerable to persuasive efforts, as their confidence is not threatened by the presentation of conflicting views. Researchers who have studied the motivational and emotional aspects of the “self” also deal with personality issues, typically proceeding along two separate paths, described as “hot” or “cold” (Leary, 2007). The “cold” aspects were cognitive views of the self, such as self-categorization or self-reference efforts. “Hot” processes included variables like self-esteem or self-conscious emotions. Although it can seem that there are more “self” theories than any other kind, and they imply an individual approach, there is actually a cohesiveness to many of these theories that links mass and interpersonal communication research. Leary maintains “motives and emotions do not operate to maintain certain states of the self, as some have suggested, but rather to facilitate people’s social interactions and relationships” (2007, p. 317) because the “outcomes towards which motives and emotions are pointed are situated in the individual’s social and physical environment” (2007, p. 334). Aspects of motivation and emotion research often served as a bridge between the cognitive, psychological approach and the sociological, functional approach.

The degree of personal involvement in an issue also greatly affects the degree and magnitude of response: the higher the involvement, the more one is able and ready to engage in processing and responding to persuasive information. One specific model designed to eliminate the passivity often assumed in the role of the receiver would be Rank’s model of persuasion (as cited in Larson, 1998). Rank’s model depicts an active response to persuasive communication, teaching people to be critical receivers. He describes the movement through the model as an intensify/downplay schema designed to increase awareness of potential manipulation.

A recent review article on attitudes and persuasion led the authors to write of being “overwhelmed” by the volume of research generated in just a three year period, supporting their claim that this work is the “most distinctive and indispensable concept” of social psychology, just as Allport claimed 70 years ago (Crano & Prislin, 2006, p. 346). After synthesizing the current literature, they identified trends such as a renewed interest in attitude formation, development of integrative models linking attitude and behavior, the focus on dual-process models and attention to message reception processes in persuasion. They summarize the present understanding as follows:

Inoculation

Not all persuasive communication is intended to bring about attitudinal change. At times, the goal might be to develop resistance to change, a situation termed “inoculation” and first examined by Lumsdaine and Janis (as cited in Severin & Tankard, 2001), colleagues of Hovland. Targeting differences in opinion during the Cold War about Russian ability to rapidly develop nuclear bombs, this study presented both propaganda and counter-propaganda, similar to their previous war-related work. The ultimate question was whether Russia would be able to produce atomic bombs within a five year period of time, and the answer was dependent upon the type of message received: without counter-propaganda, the messages were equally effective, replicating their earlier findings. The groups receiving counter-propaganda, however, responded very differently, and the results suggest that the most effective strategy for building resistance to ongoing persuasive efforts is the use of two-sided messages (Severin & Tankard, 2001).

McGuire and Papageorgis (1961) further developed inoculation theory, as they extended the medical analogy, arguing that immunization is best accomplished by inoculation. They studied “cultural truisms” (e.g. “Most forms of mental illness are not contagious,”) finding that immunizing material that is refutational rather than supportive is a more powerful inoculator. In addition, they reported that passive conditions (i.e. reading), were more effective in developing resistance than active (i.e. writing) conditions. Further research (Papageorgis & McGuire, 1961) demonstrated that inoculation also leads to long-term immunity, in much the same way a measles vaccine protects against the virus for many years and multiple exposures. Fittingly, Inoculation Theory is often used in public health campaigns to help develop resistance during a vulnerable period. One example would be Pfau’s work (as cited in Severin & Tankard, 2001) using communications aimed at students entering middle school, designed to strengthen their immunity against escalating pressure to smoke.

The functions of attitudes

The attitude construct has long been considered to consist of affective, behavioral and cognitive components. Persuasion, as one of the most basic forms of communication, was defined simply as “attitude change resulting from exposure to information from others” in a 1993 review titled Attitudes and attitude change (Olson & Zanna, 1993). Understanding persuasive communication necessitates an understanding of the fundamental attitude construct, which, according to Crano and Prislin, the field has been trying to define since the time of Thurstone (2006). The approaches to attitude change presented previously (both the “transmission/signal” and consistency models) peacefully co-existed for quite some time despite their alternative views of attitude change. The functional approach developed out of a desire to reconcile two competing views of human behavior. Irrational models of human behavior view people as malleable, non-thinking, reactive creatures. Rational models, in contrast, consider humans as being capable of developing both wisdom and intellect with adequate information. Katz was one of the first strong proponents of this approach, stating that “unless we know the psychological need which is met by the holding of an attitude we are in a poor position to predict when and how it will change” (1960, p. 170).

The Functional approach

Katz and others supporting this functional approach contended that humans are both rational and irrational, and his tongue-in-cheek description of the dilemma conveys his frustration that what he perceives to be a common sense approach has not prevailed:

That elusive middle ground is what Katz maintained the functional approach could capture. He argues that much of the prior study of communication effects had measured only constructs, such as exposure to a movie, without determining what function that exposure served in a given individual. Katz maintained that understanding and predicting attitude change required knowing the individual motivational basis of that change. He outlined four major functions which attitudes serve for the personality, described in his words as

In this same paper, Katz details the role these functions play in attitude formation, delineating the origin of the function, and the associated arousal and attitude change requirements of each. Wright (1960), in a paper detailing theoretical and methodological issues within the origins of functional research, maintains that functional analysis is not restricted to the study of useful consequences. He broadens the perspective to include “consequences of social phenomena which affect the normal operation, adaptation, or adjustment of a given system: individuals, subgroups, social and cultural systems” (Wright, 1960, p. 606), indicating that functional theory is appropriately applied to all forms of communication research. At the heart of this functional approach is the belief that persuasive messages should be fashioned according to the motivational base that supports a particular attitude, leading to a new direction in the study of persuasion and attitude change that asked a basic question: does a meaningful change in behavior accompany a change in attitude?

Initial thinking about why behavior change does not always follow attitude change targeted the complicated relationship between attitude and behavior, and the possibility that researchers may have not have actually identified the variables that impact human behavior to bring about change. Azjen and Fishbein (1970) developed a modeling equation, which states that a person’s behavior or intention to perform that behavior is equal to the attitude towards the behavior plus motivation to conform to those beliefs. This model has been used to demonstrate behavior change resulting from persuasive communication, and its usage has resulted in the addition of behavioral measures in most attitude-change research (Azjen, 1971).

Greenwald (as cited in Severin & Tankard, 2001) attempted to identify the thought processes involved in attitude change through a cognitive response model developed to address the limitations of the classic single-process models of attitude change. The cognitive response model suggests message retention and acceptance (as measured by changed behavior) are separate and distinct processes. The cognitive response model specifies a persuasion scenario that portrays the receiver reflecting upon a message and then relating it to existing attitudes, knowledge and emotions. Additional process models developed from this cognitive approach, include McGuire’s information-processing theory and the dual-process models, described as “today’s most influential persuasion paradigms” (Crano & Prislin, 2006, p. 348).

Information processing theory

McGuire’s information processing theory initially involved six steps deemed necessary for attitude change; he later expanded this to a grouping of eight and then finally twelve categories, each permutation adding more dependent variables to the persuasion process (as cited in Severin & Tankard, 2001). The basic premise of information processing, however, remained consistent, involving a movement from message exposure towards ultimate behavioral change. His model depicts an increasingly intricate web of dependent and independent variables seeking to bring about attitude change, clearly a more challenging proposition than previously thought. Current theories which describe a single cognitive process accounting for the effects of source and message in persuasion (termed “unimodels”) have provoked fruitful discussion, particularly in the impact of mood states, but lack the “comprehensive database” of the dual-process models (Crano & Prislin, 2006, p. 349).

Heuristic-systematic and elaboration likelihood models

The two primary dual-process models include the heuristic-systematic model and the elaboration likelihood model. Both describe how individuals respond to a persuasive communication, evaluating both the source and the message, which, in combination with motivation and ability to process information determine its final degree of persuasiveness. The heuristic-systematic model entails two separate means of processing messages, one described as “relatively effortless,” the other as more “cognitively demanding” (Chen, Shechter, & Chaiken, 1996, p. 262). Systematic processing necessitates a detailed, effortful examination of a message, and individuals must be both motivated to process in this manner and free from potentially adverse influences, such as time pressure. Heuristic processing relies on inferential rules (schemas) to make decisions or form broad judgments, and has been used to describe numerous situations, such as depth of processing (Griffin, Neuwirth, Giese, & Dunwoody, 2002), risk judgment (Trumbo, 1999) and minority versus majority influence (Martin & Hewstone, 2003). The heuristic and systematic processes are not (necessarily) mutually exclusive, but reflect the need to balance between “minimizing cognitive effort while maximizing judgmental confidence”(Chen et al., 1996, p. 262).

The elaboration likelihood model (ELM) also acknowledges that individuals process messages differently: some require great attention to detail and others only a more superficial treatment. The ELM identifies these two routes towards attitude change as central (active information processing) and peripheral (minimal cognitive energy expended.) Central route attitude changes are those that require extensive and effortful processing to scrutinize and evaluate the merits of an argument. Peripheral route attitude changes are based on various processes requiring less effort, but which are qualitatively different. Peripheral options include the classical conditioning process, self-perception, or the heuristics of credibility, liking, and consensus (Severin & Tankard, 2001). The ELM construct represents an elaboration continuum that has individuals at the high end assessing relevant information to arrive at a thoughtful (but not necessarily unbiased) attitude they can both support and articulate. Attitude change is still possible at the opposite end of the ELM spectrum, but it occurs with less effortful thought through a peripheral process that places a low demand on cognitive resources. Attitude changes that occur as the result of minimal object-relevant processing are thought to be weaker than those derived from high-elaboration processing. The ELM proposes that individuals use both routes to deal with the large volumes of information encountered in everyday life, and entails certain formal postulates described in the initial work of Petty and Cacioppo (as cited in Petty, Cacioppo, & Schumann, 1983). The ELM is thus considered a dual route but multi-process theory.

Level of involvement

The moderating role of involvement in the choice of route (central or peripheral), is generally accepted and has been recognized for some time, although measuring the construct is challenging (Zaichkowsky, 1985). One of the earliest papers to discuss the topic considered the issue of involvement in television advertising. Krugman writes of “un-anchored learning,” also referred to as “learning of the nonsensical and the unimportant,” which share the common ground of a lack of involvement (1965, p. 352). His prescient observations about high and low levels of involvement were fundamental to the development of the ELM. They are worth quoting at length, as he asserts that there are:

There is not consensus regarding the theoretical mechanisms that drive involvement, and Greenwald and Leavitt (1984) list a variety of concepts that have been considered over the years, including personal connections, linkage to central values, complexity of decision-making, peripheral vs. central cognitive processes, and arousal level. This theoretical uncertainty shows up during interpretation and application, as one group would predict greater cognitive change, others greater resistance to change, and still others high involvement, in the exact same situation. In general, high involvement equals personal relevance or importance, and Greenwald and Leavitt provide this helpful summary:

Petty et al. (1983) focused on the differences between the central and peripheral routes in attempting to uncover the moderating role of involvement. In general, they suggest that peripheral cues are more salient than issue-relevant arguments under low involvement conditions, and the reverse is true for high involvement. Personal relevance matters, and they reported a continuum of effects for attitudes that would be anchored by a peripheral and central option, which they anticipate would have very different antecedents and consequents. Others have studied these antecedents, consequences and processes of involvement, and the model has been applied in multiple research domains. For example, Greewald and Leavitt (1984) linked audience involvement to several psychological concepts and evaluated their connection to the antecedents and consequences of involvement, and found relationships with both. Antecedents are related to involvement through what they term “bottom-up” and “top-down” (Greenwald & Leavitt, 1984, p. 591) processing, as well as competence and capacity limitation. Involvement was related to consequences through both cognition and attitude. The authors claim that linkage to the constructs of variable attentional capacity, levels of processing, qualitatively different representational systems, and arousal, create a bridge that has been missing between the psychological and cognitive approaches.

Diffusion theory

The study of persuasive communication and attitude change within groups was discussed previously, using both propaganda and social identification research (source credibility) to demonstrate both mass and interpersonal theoretical development. Each of these areas contributed to shifting beliefs away from the direct effects of persuasive communication towards the more limited effects theories, as did the work of Katz and Lazarsfeld. Their investigations into the process of changing political attitudes at both the group and personal level is another area where the study of persuasive communication has flourished, as detailed earlier in this paper using the 1957 issue of POQ as representative of the original work. The development of the two-step flow model of communication, while criticized mainly because it lacked explanatory power, gradually developed into a multistep flow model used to study the social processes of spreading new information, called diffusion. The book Diffusion of Innovations (1995), is an examination of close to 4,000 diffusion publications, which revised previous theory and authoritatively defined innovation as “an idea, practice, or object that is perceived as new by an individual or other unit of adoption” (Rogers, 1995, p. 11). This updated material clarified the process of what has been described as a multi-directional exchange of information among participants (rather than a one-directional flow), simultaneously generating a level of uncertainty by providing alternatives to the status quo. Rogers used the mathematical model of Shannon and Weaver (1949) to explain how individuals resolve this issue of uncertainty, and identified three useful categories, each containing five characteristics involved in the innovation process. The first involves characteristics of innovations that affect their rate of adoption, including the concepts of relative advantage, compatibility, complexity, trialability and observability. Rogers also listed the stages of the innovation decision process as moving through the following progression: knowledge, persuasion, decision, implementation, and confirmation. Lastly, he names five types of adopters, including innovators, early adopters, early majority, late majority and laggards.

Opinion leaders and Diffusion Theory

The foundation of classic Diffusion Theory rests on a 1943 article by Ryan and Gross (1943) who found that social contacts, social interaction, and interpersonal communication were key influencers of new behavior adoption. While a number of factors affect the diffusion of innovations, these early findings have been consistently reproduced in a variety of settings (Robinson, 2006). Social Influence Theory posits that the behavior of one individual affects others (Zimbano, 1991) and that behavior is directed by customs, habits, group norms and common social meanings (Mittman, 1992). Diffusion Theory relies on these interpersonal contacts to affect the adoption of new behaviors, often using network analysis to determine which individuals are central to and influential within a community, and thus potentially able to accelerate the process of diffusion (Katz, 1957; Katz, 1963; Valente, 1996). Research labels these individuals “opinion leaders” and their involvement in and support of an intervention affects successful implementation (e.g.Chan & Misra, 1990; Feathers, Kieffer, Palmisano, Anderson, Sinco, Janz, Heisler, Spencer, Guzman, Thompson, Wisdom, & James, 2005; Grimshaw, Shirran, Thomas, Mowatt, Fraser, Bero, Grilli, Harvey, Oxman, & O'Brien; Oxman, 1995; Rogers, 1995; Thomson O'Brien, 2000; Valente & Davis, 1999; Venkatraman, 1989).

Opinion leaders function by altering group norms and shifting acceptance of what constitutes acceptable behavior (e.g. Cosens, Ibbotson, & Grimshaw, 2000; Feathers et al., 2005). There is not, however, a universally accepted description of the role and/or function of the opinion leader. This inconsistency makes it challenging to compare interventions and contributes to construct ambiguity (Feathers et al., 2005). The label of opinion leader is often interchangeable with that of “champion,” although much of the social influence research is connected to “product champions” within business management literature (e.g. Chakrabarti, 1974; Howell & Higgins, 1990; Markham, 1998; Markham & Aiman-Smith, 2001; Schon, 1963). In contrast, the healthcare literature has focused on professional medical practices and those who are “educationally influential” (e.g. Clark, Hiss, Chin, Marrero, Davis, Walker, Fisher, & Wylie-Rosett, 2001; Fairhurst, 1998; Locock, Dopson, Chambers, & Gabbay, 2001; Mittman, 1992). Other research uses mass media in conjunction with interpersonal communication to drive desired behavioral changes (Valente, Saba, W.), while others use peer networks (Jemmott, Jemmott, & Fong, 1998; Neaigus, 1998; Valente, Saba, W.). In general, “opinion leaders” accelerate diffusion and serve as champions for interventions (Valente & Davis, 1999) in a variety of ways. Social Cognitive Theory provides some common ground for understanding and interpreting the effects.

Social Cognitive Theory

Social Cognitive Theory (SCT) is a dynamic process describing the interaction between personal factors (including cognitive, affective and biological events), environmental factors, and human behavior, often referred to as “triadic factors”(Bandura, 1986). Evolved from a theory of Social Learning (Miller & Dollard, 1941) that was expanded to include self-efficacy (Bandura, 1977), SCT suggests that humans adapt and change in response to cognitive, self-regulatory, vicarious and self-reflective processes (Bandura, 1986). Self-efficacy was considered by Bandura to be the most important personal factor in changing behavior. It is defined as confidence in one’s ability to take action and overcome barriers, even when faced with obstacles (Bandura, 1986). Self-efficacy beliefs influence behavior in numerous ways, affecting personal choices, effort and persistence expended, thought patterns and emotional reactions (Bandura, 1986; Bandura, 1999; Pajares, 2002). Self-efficacy beliefs develop from previous performance, vicarious experience, social persuasions and somatic states. In addition, one forms efficacy judgments based on the integration, interpretation and recall of pertinent information and circumstances. This process occurs within the triadic factors: interactions among these will reflect individual differences, the particular behavior examined, and the situation in which the behavior occurs (Bandura, 1989). Similar to the Social Influence Theory of Kelman (Kelman, 1961), the principles of Social Cognitive Theory suggest that individuals are more likely to adopt a modeled behavior if the model is similar to the observer, has admired status, or the behavior has functional value (Bandura, 1977, , 1986).

Literature reviews indicate that researchers have studied the role of self-efficacy in a wide variety of situations (e.g. AbuSabha & Achterberg, 1997; Fein, von Hippel, & Spencer, 1999; Marks & Allegrante, 2005; Marks, Allegrante, & Lorig, 2005). For example, self-efficacy is often incorporated in health-related research because it appears to have a direct effect on many health promoting behaviors, often with diffusion applications (AbuSabha & Achterberg, 1997; Anderson & Baym, 2004; Greenwald & Leavitt, 1984; Holloway & Watson, 2002; Pajares, 2002; Winett, Anderson, Whiteley, Wojcik, Rovniak, Graves, Galper, & Winett, 1999).

Ongoing work in diffusion theory has accelerated due not only to expanded theoretical paradigms and widespread application, but also in response to the changing media landscape. Today, the impact of the ability to communicate rapidly is combined with massive numbers of people in an ever-expanding communication “diffusion network” in ways that are not yet understood or even fully appreciated. This topic will be re-visited in Part 2 of this paper as there are some interesting diffusion applications being examined with new technologies.

Agenda setting

We have discussed a number of theories which developed in response to the direct effects theoretical framework, replaced by a limited effects model that was also questioned as time went by. In a world where media is ever-present, it simply did not seem reasonable to believe that mass media had no effect. Kosiciki connects the simple and intuitively friendly agenda setting hypothesis with a rediscovery of moderate effects, and the move away from persuasion research to an “agenda of issues”(1993, p. 104). Further strengthening this desire for change, he claims, was the fact that after decades of studying persuasive communication, many researchers simply concluded that “Attitudes were not clearly connected to behavior, and media were not clearly and consistently connected to either” (Kosicki, 1993, p. 103). The time was particularly ripe for a paradigm shift, as cognitive psychology was becoming the dominant approach, and researchers questioned if they were looking for effects in all the wrong places: perhaps they should evaluate perceptual changes rather than attitudinal change? The agenda setting hypothesis investigated the development of issue salience, and is characterized by the twin core of media content and audience perception. McCombs and Shaw (1972) are credited with first using the term, but other researchers, are often acknowledged as laying the groundwork (i.e. Lang and Lang or Norton Long). The concept is regularly linked to Cohen’s statement concerning the power of the press: “It may not be successful much of the time in telling people what to think, but it is stunningly successful in telling its readers what to think about,” which Kosicki asserts is a clever tweak indicating the shift away from persuasion to something entirely new and fresh (1993).

McCombs and Shaw (1972) conducted the first systematic study of the agenda setting hypothesis during the 1968 presidential campaign by conducting a content analysis of the mass media serving the “undecided” voters in Chapel Hill, North Carolina, and then interviewing those voters to determine what they perceived to be the major problems in the country at that time. They found an extremely high correlation for both major and minor news items with voter perception of the relative importance of that issue (.967 and .979, respectively.) These results suggested a very strong association between the amount of emphasis placed on a topic by the news media and voter perception of issue salience. Correctly interpreting the correlational findings, McCombs and Shaw (1977) conducted a follow-up study to determine the causal order of agenda setting: did the media influence the public or was the media influenced by the public agenda? Their analysis provided mixed results, demonstrating only a trend towards the media (specifically, newspapers) causally shaping the public’s agenda, but the hypothesis has been more fully supported in subsequent research (e.g. Iyengar & Kinder, 1987; Iyengar, Peter, & Kinder, 1982).

In a citation analysis of agenda setting research, Rogers (1993) suggests a useful taxonomy for organizing the research: media agenda setting, public agenda setting and policy agenda setting. Kosicki (1993) agrees with this tactic and recommends broadening the study of the effect to incorporate all three sub-areas. Media agenda setting literature considers published media content and the antecedents of issue definition, selection and emphasis; media agenda is the focal dependent variable. Public agenda setting deals with linkages between mass media issues and public priority issues, using the relative importance of issues as the main dependent variable. Policy agenda setting comes from a political science perspective, most often evaluating the issue agendas in light of legislative bodies or elected officials and their connections to media. Kosicki suggests referring to agenda setting as a “model of media effects” (1993, p. 102), as that term distills the essence of agenda setting to one of many complex hypotheses of media effects linking media production, content and audience effects together. Severin and Tankard (2001) review a variety of other issues that have been investigated in conjunction with the agenda setting hypothesis, including priming, issue abstractedness, the individual need for orientation, and the role of both the amount and length of exposure. They describe the concept of agenda building as an expanded version of agenda setting, designed to encompass a collective process of influence among the media, government and public. Agenda building assigns roles that involve variables such as the use of symbolic code words and framing (often called the “second level of agenda setting”), although not everyone agrees about the terminology and processes involved in framing, as a recent JOC issue attests (Entman, 2007; Kinder, 2007; Reese, 2007; Weaver, 2007). In related work examining the issue of how the media agenda develops, Funkhouser (1973a) describes a disconnect between media coverage and reality, proposing that mechanisms such as selective and over-reporting, and the creation of pseudo-events, operate in conjunction with the actual flow of events (1973b). Not surprisingly, the actual process of how agenda setting works has not been as clearly defined as what it “looks like.” Further research is needed in this area, but some of the findings reflect multiple dimensions presented earlier, such as the study by Wanta and Miller (as cited in Severin & Tankard, 2001), who found that agenda setting involves information processing, which in turn is affected by selective perception and retention. Kosicki summarizes the upcoming challenges and need for synthesis well:

Knowledge gap hypothesis

Another attempt to understand the role of media in society is proposed as the knowledge gap hypothesis, which asserts that as the amount of information entering a social system increases, those with a higher socioeconomic status selectively receive that information, creating castes of the “information poor” and “information rich.” The operational form of the first knowledge gap hypothesis demonstrated clear differences according to education level, regardless of the amount of publicity generated on a given topic (Tichenor, Donohue, & Olien, 1970). The authors suggested a number of reasons why this gap occurs, such as relevant social contact or selective exposure, acceptance and retention. They later explored ways to reduce or eliminate the gap entirely (Donohue, Tichenor, & Olien, 1975), modifying the hypothesis to account for issue involvement and interest. Rogers (1976) recommended re-framing the situation as a “communication effects gap,” as he pointed out that not all gaps involve information, but may occur because of attitudes and behavior.

Criticism of the knowledge gap hypothesis is that it centers on the traditional transmission paradigm of communication, and some have suggested that the surveys employed don’t always test relevant information, leading to skewed interpretations of the data. A need for better theoretical development is also an issue, and there is a need for distinguishing between macro and micro levels of knowledge acquisition (Pan & McCleod, 1991). Most of the knowledge gap research is conducted at a single time point, and since there is some evidence that knowledge gaps are widening, rather than shrinking, it becomes even more essential to expand understanding of how this phenomena changes over time (Gaziano, 1997).

Cultivation theory

Television is a primary form of communication in society today, and Gerbner and colleagues have attempted to shift away, much as agenda setting did, from the traditional effects model towards a cultivation perspective, viewing the dynamic role of television as a multi-directional process. Gerbner et al. describe television as “a key member of the family, the one who tells most of the stories most of the time” (1980, p. 14), and argue that the end result, particularly for heavy viewers, is a common world view, cultivated through chronic message absorption. The first studies compared light and heavy viewers, and they found that heavy television viewers most often perceived the world as it is portrayed in their living rooms, rather than reflecting a realistic understanding of the world as it is outside their door. Despite the powerful story cultivation theory weaves, there are also strong criticisms, particularly from those who re-analyzed the data while controlling for a number of different variables and found limited effects attributable to watching television (e.g. Hirsch, 1980). In a similar critique, Hughes put it this way:

Gerbner’s group revised their theory in response to criticisms such as these, adding the concepts of mainstreaming and resonance, and no longer claiming uniform, consistent effects for heavy viewers (as cited in Severin & Tankard, 2001). They did maintain, though, that even if short-term effects are minimal, the cumulative effects on those with substantial exposure would not be negligible, and encouraged continued study of accumulating effects. One additional theoretical refinement has been the division of cultivation effects as involving first-order beliefs (facts concerning the real world) and second-order beliefs (extrapolation from these facts to a general belief.)

Ongoing debates surround cultivation theory. Potter (Potter, 1993) and others (as cited in Severin & Tankard, 2001) have proposed an extended cultivation hypothesis, suggesting that some, but not all, forms of programming might function this way. Gerbner and his associates have dismissed this permutation as not addressing the fundamental theoretical assumptions, so consensus remains out of reach (as cited in Severin & Tankard, 2001). Further research in the cultivation area has included consideration of related issues, such as the news refraction hypothesis, which recognizes the heightened influence that local news might have because it hits “close to home,” yet the greatest promise appears to be determining whether the theory extends to other media forms. McLuhan’s media determinism will be an interesting template from which to consider the heightened sensory involvement of interactive mediums, such video games, and we will explore this and the role of cultivation theory in the context of the “new” media in Part 2.

Third-Person Effect

An interesting interpretation of the third-person hypothesis, which states that individuals believe a persuasive media message will affect others more than themselves, has both a perceptual and behavioral component. Paul et al. (2000) contend that the behavioral hypothesis has received only qualified support, yet the perceptual hypothesis appears to be a strong dimension. Despite consistent empirical findings, however, the factors that increase or diminish this perception remain unknown, as does the reason for some individuals are susceptible while others are not. Their meta-analysis of the perceptual component finds that attribution theory and biased optimism are the most promising theoretical frameworks to incorporate the effect (Paul et al., 2000). They conclude that the robust findings and significant effect size support the hypothesis, and also identify moderators that can affect third-person perception, including issue characteristics, sampling and types of respondents. One practical application of this hypothesis is seen in a study investigating female response to programs and commercials of attractive women on perception of ideal body weight. They found that women believed others would be more influenced than themselves, a perception which increased with increasing social distance from the subject portrayed in the visual (David & Johnson, 1998).

Powerful effects model

Concurrent with the return oscillation of the effects pendulum to the more moderate and contingent effects approaches (such as agenda setting), a few studies have demonstrated powerful effects on large numbers of people. Mendelsohn (1973) describes three projects which effectively changed driving behavior and participation in community organizations. He credits the success of these studies to spelling out objectives, pinpointing the target audience, working to overcome audience difference of opinion, and finding relevant themes to stress in the developed messages. Another communication campaign targeted behaviors thought to impact the development of heart disease, and they found that both the mass media and intensive instruction methods effectively changed behavior, but the town receiving only the media messages also showed significant improvement (Maccoby & Farquahr, 1975). The Great American Values Test developed a model for changing values based on consistency theory (Ball-Rokeach, Rokeach, & Grube, 1984). In an experimental design rarely achieved in the real world, they were able to eliminate competing programs and show their program on all networks, comparing the towns that viewed it with one that was blacked out during that time. Achieving an impressive Nielsen rating of 65%, they found that this single half-hour of television was able to change viewer’s attitudes, their rankings of basic values, and demonstrated an increased willingness to engage in political behaviors.

Spiral of Silence

The development of public opinion through a process called the spiral of silence attributes more power to mass media than any of the others powerful effects theories (as cited in Severin & Tankard, 2001). This theory rests on three characteristics of mass communication describing a process that included a widespread buildup of themes, presented in a unified manner: ubiquity, cumulation and consonance, respectively. The “spiral of silence” develops around a structure that has people attempting to conform to public opinion, remaining silent if they are in the minority. The powerful role of mass media is crucial in this process, as they shape impressions about which opinions are dominant, increasing or socially acceptable. This theory has been challenged by others who do not demonstrate the same extent of fear of isolation (e.g. Lasora, 1991; Neuwirth, Frederick, & Mayo, 2007) or ability to estimate public opinion and respond accordingly (e.g. Neuwirth et al., 2007; Rimmer & Howard, 1990).

Noelle-Neumann’s (1983) concern about the media’s power is even more pronounced in later years, as she suggests that the media has been complicit in encouraging the belief that they have weak or limited effect, and actively discredited those who found otherwise. She lists a trail of neglected studies that supported a powerful effects model, and recalls a conversation with Paul Lazarsfeld, intimating his “minimal effects” research reflected his Institute’s need for media support, rather than reflecting a true interpretation of the data. She describes what she perceives to be his selective reporting of The People’s Choice as a “kind of scientific self-censorship” (Noelle-Neumann, 1983, p. 160), and her fervent diatribe continues in a rant against journalists in particular, and social science research in general, which had “liberated the media from responsibility for the evils of society” (Noelle-Neumann, 1983, p. 161). Despite her sarcastic dismissal of the uses and gratifications (U&G) approach as a pablum served up to avoid conflict with journalists, we will close this narrative history of communication research with a look at just that concept. U&G was one of the primary vehicles putting signal transmission theory into reverse, shifting the focus from the purposes of the communicator to the purposes of the receiver, asking “What do people do with the media?”

Uses and gratifications approach

Katz (as cited in Severin & Tankard, 2001) prescribes asking the question in this manner, rather than remaining in the source → receiver mode of the early models, as the antitdote for the body of communication research, perceived by some to be “withering away” (Berelson, 1959). He points out that the only dying going on was in thinking about and studying communication as simply persuasion, citing a number of studies to support his contention, including Berelson’s work that showed different people “missed” the paper for different reasons, depending on how they planned to “use” it. Klapper (1963) enthusiastically welcomes U&G into the functional fold, particularly because the approach avoids the effects dichotomy of seeking a “yes or no,” “limited or powerful” answer. He challenges researchers to apply truly functional methods in conducting their studies, rather than simply generating lists of the uses people make of communication and the derived gratifications. His concern stems from studies such as the first direct measure of the U&G approach, conducted during an election campaign. Blumler and McQuail (as cited in Severin & Tankard, 2001), used qualitative interviews to determine motivation for watching political broadcasts, compiling a list of eight “uses.” Their data indicated that rather than using media to reinforce existing beliefs, most used it in what the authors called “surveillance of the political environment.” A more complete list was later compiled, to include diversion, personal relationships, personal identity/individual psychology and surveillance as descriptive of U&G in that context.

In an excellent review titled Uses and gratifications theory in the 21^st century, Ruggiero (2000) states that the early U&G studies we