The Lost Doctrine: Suggestion Theory in Early Media Effects Research

The “Natural History” of Effects Research

The history of media effects is often described in terms of eras or phases. Denis McQuail (2010) identified four phases in what he called “the natural history of media effects research and theory” (p. 455). The history is sometimes presented as three eras (Bryant & Thompson, 2002), sometimes four (Esser, 2008), five (Baran & Davis, 2006; McQuail, 2010; Robinson, 1996; Vorderer et al., 2020), or even six (Neuman & Guggenheim, 2011). The various schema may diverge on the nature of stages three through six. But all agree on the first two: the direct effects and limited effects eras.

Hypodermic Assumptions

The hypodermic model is typically described as a now-discredited theory of direct media effects that rested on two theoretical pillars: one sociological, the other psychological. The sociological element garners widespread contemporary agreement, with an assumption that a flawed theory of “mass society” dominated early sociological thinking. This theory was characterized by a concern with the transition of collective social life from a rural, agrarian setting of close interpersonal ties to an urban, industrialized “mass” society devoid of communal relationships. Drawing on the work of sociologists such as Emile Durkheim, Herbert Spencer, and Ferdinand Tonnies, some early observers argued that the lack of social anchorage resulting from that transition to modernity left individuals easy victim to powerful, mediated voices. According to the received history of the field, research done starting in the 1940s overturned these assumptions (Baran & Davis, 2006; Bauer & Bauer, 1960; DeFleur & Ball-Rokeach, 1989; Hardt, 2001; McQuail, 2010).

The theory of the mind, or the psychological mechanism assumed to produce influence, has gained less agreement. Some modern authors offer no psychological explanation, holding that pre–World War II scholars simply assumed that opinion could be changed immediately and directly (Straubhaar et al., 2018; Tan, 1981). Richard Petty et al. (2009), for example, said that because researchers of that period assumed the audience was “captive, attentive and gullible,” they believed that mass communicated information “produced direct effects on attitudes and behavior” of “helpless victims” (p. 126). People were, as Werner Severin and James Tankard (1979) explained, “naïve and they believed lies” (p. 125).

Other modern scholars have proffered cognitive explanations. Early behaviorism or stimulus–response (S-R) theory has frequently been identified as the model of influence used by earlier social scientists (DeFleur, 1966, pp. 115–116; DeFleur & Dennis, 1981, p. 502; Miller, 2005, p. 250; Vorderer et al., 2020). Others have offered Freudian interpretations (Barnouw, 1956; DeFleur, 1966), combined behaviorism and Freudianism (Baran & Davis, 2006), or called on 1920s instinct theory as the foundation for a belief in direct effects (DeFleur & Ball-Rokeach, 1989). A few media historians have referenced crowd theory, citing Gustave Le Bon and Gabriel Tarde (Bineham, 1988; Butsch, 2001).

No matter which cognitive process is invoked, the received history holds that the early theorists saw media effects as uniform across individuals (Jowett & O’Donnell, 1999; Lowery & DeFleur, 1983). Sparks (2010) observed “Little consideration was given to the possibility that people might respond differently to the same message,” because “the theoretical framework that allowed for individual differences was not yet in place” (p. 51). Using instinct theory, Melvin DeFleur (1966) explained that at the time, human behavior was seen as governed by non-variable, inherited characteristics. Consequently, media messages, “tapped inner urges, emotions, or other processes over which the individual had little voluntary control. Because of the inherited nature of these mechanisms, each person responded more or less uniformly” (DeFleur, 1966, p. 116). An S-R perspective also has provided a firm foundation for this uniform effects narrative (Bineham, 1988; Vorderer et al., 2020). As with the mass society assumption, this early belief in direct, uniform psychological effects was said to have been overturned in the transition to the better informed and more sophisticated limited effects model.

The Received History—Points of Contention

A growing body of literature critiques, modifies and extends media studies’ institutional memory (Bryant & Thompson, 2002; Delia, 1987; Gitlin, 1978; A. Lang, 2013; K. Lang & Lang, 1993; D. Park & Pooley, 2008; Robinson, 1988; Wartella & Middlestadt, 1991; Wartella & Reeves, 1985). As Annie Lang (2013) observed, “many excellent articles” refute our field’s received history (p. 12). Scholars from social scientific, critical, and cultural traditions have indeed criticized the discipline’s spotty, sometimes misleading, and often over-simplified recollection of early mass communication effects research. Unsurprisingly, many of these revised histories differ not only with the received history, but with each other.

For social scientists, the debate often starts with the issue of what constitutes “scientific research” and when something of that description actually began in media studies. Most of the extant histories state or imply that the hypodermic model was a set of untested assumptions held by both social scientists and laymen. “Such views were not based on scientific investigation, but on observation of the enormous popularity of the press and . . . film and radio” (McQuail, 1983, p. 176). By this view, true scientific investigation of media effects did not begin until the late 1930s or early 1940s, principally with the work of Hovland and Lazarsfeld. Brown (1970) pointed to Hovland when arguing that empirical research in mass communications field began in the 1930s. David Morrison (1978) asserted: “When Lazarsfeld entered the field in 1937 it was almost totally unmapped as an area of enquiry” (p. 347). Richard Perloff (2010) said social scientific research on persuasion began in the 1930s, with Gordon Allport and Hovland. Everett Rogers (1997) also cited Hovland, stating that “persuasion study did not get underway until World War II” (p. 367). A related origin story offers communication’s “founding fathers.” Wilbur Schramm (1963), following Bernard Berelson (1959), marked the true start of communication research with the work of what he termed “the founding fathers”: Lasswell in political science, Lazarsfeld in sociology, Lewin in small group communication, and Hovland in psychology.

Steve Chaffee and John Hochheimer (1985) acknowledged early, isolated studies of media influence, including the Payne Fund studies of children and the movies and Walter Lippmann’s (1922) classic Public Opinion. Generally, however, they held that systematic social science investigation did not begin until the late 1930s. They also suggested that the hypodermic model itself was created in the 1940s to serve as a foil or “straw man” in support and advancement of the emerging limited effects view (pp. 287, 289–290). At least with respect the hypodermic appellation, Chaffee and Hochheimer are well supported by the historical record (Lubken, 2008; Pooley, 2006). Both the hypodermic and magic bullet labels were post–World War II inventions. There is, however, debate with respect to the broader claim that little or no research was conducted on media effects prior to Hovland and Lazarsfeld.

Jennings Bryant and Susan Thompson’s (2002) “revised history” pointedly eschews a chronology that moved “in pendulum swings from ‘all-powerful’ to ‘limited’ to ‘moderate’ to ‘powerful’ effects” (p. 42) and gives greater credence to work done before World War II. Ellen Wartella (1996) called it “historically inaccurate to locate the dominant paradigm for media research with Lazarsfeld; the roots of the paradigm are older and derive from wider influences in U.S. social science” (p. 175). Likewise, Wartella and Reeves (1985) argued that researchers had a more sophisticated understanding of media influence than subsequent writers gave them credit for. Supported by parallel research from Richard Butsch (2001), Wartella showed that social scientists keenly appreciated individual differences in effects in the earliest mass communication studies; early theorists never assumed uniform effects. Jeffrey Bineham (1988), responding directly to Chaffee and Hochheimer, suggested that these varying positions may turn on how media effects research itself is defined, although he ultimately aligned with those who see research, albeit steeped in mass society theory, forming well before Lazarsfeld and company.

Outside the debate within media psychology, scholars have pointed to media and society research beyond the confines of the strictly social-scientific. Rejecting the “myth” of a magic bullet model, K. Lang (1996) noted that a healthy body of nuanced media-related research emanated from Europe, especially from Germany, well before 1940. Cultural scholars have repeatedly argued that the established origin story neglects much important pre–World War II work, especially from the Chicago School (Carey, 1996). Glenn Robinson (1988) has placed the start of media studies in the Chicago School, as has Karin Wahl-Jorgensen (2004). Daniel McDonald (2004) similarly anchored the pioneering phase of media research in the works of Charles Horton Cooley and the early cinema scholar, Hugo Munsterberg. Garth Jowett and Victoria O’Donnell (1999) and J. Michael Sproule (1997) explored the extensive writing on propaganda during and after World War I. Jesse Delia’s (1987) review of work on media between 1900 and 1940 included studies on political communication, business and commercial applications, educational communication, as well as cultural and critical scholarship. Media studies encompasses a host of traditions beyond the purely positivistic and each has its own rich heritage (Vorderer et al., 2020).

With that appreciation in place, I focus on that narrower part of the history that engages the social scientific, specifically the social psychological, tradition and the claims and assumptions that weave through theories of attitude formation and change. The point is that none of these histories, with the possible exception of Butsch, noted below, includes any meaningful treatment of suggestion.

The Missing Piece

If contemporary mass communication literature mentions early suggestion theory at all, the reference is circumscribed, often in conjunction with the work of Le Bon or Tarde. DeFleur (2010), for example, observed that what “Le Bon stressed, and worried about, was that the psychological isolation of the individual in a crowd results in a great increase in suggestibility”; this had “significant implications for understanding the [associated] fear of mass media” (p. 128, italics in original). Delia (1987) observed in passing that Lasswell’s (1927) perspective in Propaganda Technique in the World War was “reminiscent” (p. 26) of Le Bon’s crowd theory and associate processes of imitation and suggestion, but said those processes had no lasting influence in U.S. research.

Bineham (1988) anchored the rise of a hypodermic model in Le Bon’s crowd theory and referenced Trotter’s writing on suggestibility; nonetheless, he saw mass society theory as the foundation for a pre–World War II direct effects belief. Suggestibility as a possible cognitive avenue for influence was left undeveloped. Larson (1992) mentioned Leonard Doob’s use of suggestion in propaganda analysis, but only tangentially (p. 351). Bryant et al. (2013) described Frances Fenton’s suggestion-based studies in 1910 and 1911 of the impact of newspaper coverage on anti-social behavior That reference, however, was nested in a larger description of early mass communication research, with the Fenton research serving only as an illustration of the contemporaneous public and social science concern about powerful media effects (pp. 34–40).

Butsch (2001, 2008), a sociologist and film scholar whose study of the history of audiences emphasized socially marginalized groups (by virtue of class, ethnicity, or gender), was the only contemporary writer I found who described suggestion theory in any detail. Butsch examined suggestion theory as it developed in French crowd theory; he described Le Bon’s concept of emotional contagion and suggestive influence, and demonstrated the application of these ideas by several early film and radio scholars. It was not his intent, however, to situate that review in the context of the evolution of social psychology and media effects theory.

In summary, a robust conversation about the roots of the discipline is on-going. This monograph joins that conversation, arguing that the extant reviews and critiques are useful but incomplete. The published history remains misleading in several important ways. At least with respect to the tradition emanating from social psychology, the critical missing piece in the received history involves suggestion theory. Below, then, I trace the doctrine’s trajectory from its roots in Europe to its application in U.S. social psychology, its role as a theory of media influence, and its later paradigmatic decline and eventual disappearance from the historical record.

Suggestion and Imitation

From the beginning, the terms suggestion, imitation, and contagion were poorly distinguished, even by Le Bon and Tarde (Nye, 1975, p. 71). U.S. scholars frequently used suggestion-imitation as a compound adverb describing a single process; some used the two terms interchangeably (Asch, 1948; Ellwood, 1922, p. 224). George Stratton (1914) treated imitation as “a special form of suggestion” (p. 200). As the field matured, however, the labels came to represent separate mechanisms for the transmission of influence. Imitation was defined as the replication of an observed behavior (Ellwood, 1901). Suggestion involved the influence associated with the communication of a concept or idea (Dunlap, 1925).

Robert Park and Ernest Burgess’s 1921 anthology, Introduction to the Science of Sociology, for example, dealt specifically with suggestion in some chapters, imitation in others. Charles Judd (1921) reviewed just the process of imitation, noting its role in social learning, especially among children. Meanwhile, Bechterew (1921) focused solely on the nature of suggestion, offering one of the more colorful definitions, as a process whereby ideas circumvented our conscious minds, or egos, and entered “[n]ot through the front door, but—so to speak—up the back stairs . . . directly into the inner rooms of our soul” (p. 410). Reviewing the power of “social suggestion” in the context of classic crowd theory, he noted: “The multitude can . . . ever be led according to the content of the ideas suggested to it” (p. 417). Luther Bernard’s Introduction to Social Psychology (1926) similarly distinguished the concepts, providing two chapters on suggestion and four on imitation.

Imitation and suggestion were, in short, conceptually disentangled and set on theoretically separate paths. Imitation increasingly became the study of influence through observed behavior and was intuitively (although not exclusively) focused on children (Faris, 1926). Suggestion, on the other hand, involved a process of influence by general communication and was therefore applied more widely, including in the study of mediated communication.

Perceptual and Attitudinal Suggestion

With its roots in hypnotism, suggestion was of long-standing interest to psychologists seeking to better understand the mind and methods of dealing with mental disorders. The study of hypnotism and “waking suggestion” became part of this broader project. Experimental work even before the turn of the century (Small, 1896) and into the 1930s in Europe and the United States focused on the role of suggestion in changing perceptions about things such as the length of lines, or the temperature or weight of an object (Coffin, 1941). Dozens of such studies were reported in the Psychological Bulletin’s annual review of suggestion research in this period (Scott, 1910–1916). The ability to induce or influence a visual perception or physical sensation by suggestion had clear implications for the influence of attitudes, and although some psychologists looked closely at the manipulation of beliefs and opinions, the emphasis on opinion formation and modification by suggestion came to full blossom in social psychology, and sometimes sociology.

Suggestion, Attitude Change, and Instinct

In his 1916 presidential address to American Psychological Association, John Dewey (1917) observed, “For more than a decade [Tarde’s] work and that of his followers in France and in the United States—among whom we may cite in diverse directions Baldwin and Ross—dominated social psychology and almost sociology” (p. 267). But it was time, he said, for the field to abandon the “wrong track in which the Imitation and Suggestibility schools had set it going” (pp. 267–268). Instead, he pointed to the promise of models founded on the emerging principles of behaviorism and instinct theory. He was especially enthusiastic about the potential of instinct theory, which became a powerful, if short-lived, paradigm in social psychology in the 1910s and 1920s and, despite Dewey’s cautionary note, carved out an important role for the suggestion process.

Building on Darwinian evolution, William McDougall’s (1908) foundational social psychology text proposed a theory of human behavior and group interaction based on evolutionary psychology. All behavior, he said, was a manifestation of basic human instincts, the “innate or inherited tendencies which are the essential springs or motive powers of all thought and action” (McDougall, 1909, p. 19). Each instinct was tied to a particular emotion or set of emotions, such as the instinct to flee and an associated emotion of fear, and the instinct of repulsion and an emotion of disgust.

McDougall (1909) also proposed a second-tier classification of “innate tendencies” or “pseudo-instincts,” among them, suggestion, imitation, and sympathy. Suggestion was based on the “subjection” instinct and associated with an innate deference to authority or prestige (p. 99). His formal definition of suggestion was frequently cited by subsequent scholars of the period, “Suggestion is the process of communication resulting in the acceptance with conviction of the communicated proposition in the absence of logically adequate grounds for its acceptance” (McDougall, 1909, p. 97, italics in original). McDougall’s instinct formulation swept through social science, spawning important works including Edward Thorndike’s The Original Nature of Man (1913), Robert Woodworth’s Dynamic Psychology (1918), and Graham Wallas’s Human Nature and Politics (1908). “The decade following the appearance of McDougall’s Introduction to Social Psychology was almost wholly dominated by instinct theory” (G. Allport, 1968, p. 57).

For a brief period, then, social psychologists considered suggestion-based attitude change through this lens of an innate instinct toward submission. Trotter (1917) focused on “herd instinct” and “herd suggestion.” Drawing from Sidis, he grafted Le Bon’s suggestibility and “group spirit” onto McDougall’s evolutionary instinct theory, and concluded that individuals submit to the will or “the voice of the herd” (p. 114), what he termed “the acme of the power of herd suggestion” (p. 115). Because this tendency is biologically innate, suggestibility was not abnormal, as others proposed, but rather “a normal quality of the mind” (p. 33). “Man is not, therefore, suggestible by fits and starts, not merely in panics and in mobs, under hypnosis and so forth, but always, everywhere, and under any circumstances,” he declared (p. 33). Instinct theory burned brightly but briefly. It was soon displaced by a more lasting paradigm, behaviorism.

Suggestion, Attitude Change, and Behaviorism

Arising in part out of Pavlovian conditioned-response experiments in animals and as a reaction to psychology’s failed introspection model, behaviorism became the dominant paradigm in social psychology. Also labeled S-R theory and learning theory, behaviorism (Watson coined the term in 1913) was analytically deployed across the full range of social behaviors, including questions of media influence. In the earliest days of behaviorism, suggestion was central to ideas about opinion change.

F. Allport was a powerful voice in early behaviorism; Daniel Katz (1979) called him “the father of experimental social psychology” (p. 351). In addition to the 1908 texts from Ross and McDougall, Allport’s, 1924 book, Social Psychology, was foundational in the development of social psychology (Katz, 1979; Kruglanski & Stroebe, 2012). An original voice in bringing an empirical, laboratory-based examination of social behavior into the discipline, Allport began by moving the focus of social psychology analysis away from groups and institutions and toward individuals: the individual was seen as the only real, concrete unit of social analysis. He rejected the concept of a “crowd consciousness” or “crowd mind,” as well as the notion of invariable instincts (F. H. Allport, 1924, pp. 4–9). He did believe in inherent human predispositions, which he termed “prepotent reflexes,” but argued that these were molded, through a conditioning process, by the social environment (pp. 79–82).

While strongly privileging the individual, F. H. Allport (1924) was also concerned with the behavior of individuals in group and crowd settings and the increased suggestibility of individuals in real or imagined aggregations. Allport embraced the process of suggestibility as described by Baldwin, McDougall and others. “When we accept an opinion uncritically, using it as a basis for our belief or action, we may be said to respond to a suggestion” (F. H. Allport, 1924, p. 242). He explained the process, however, in behavioristic terms, defining suggestion as a form of conditioned response:

Perfectly normal individuals . . . show at times an immediate and undeliberated response to commands. These effects are based upon deeply fixed habits of association between words, sounds and the bodily movements which they signify. It is convenient to regard them as sub-cortical or “short-circuited” modes of response. (F. H. Allport, 1924, p. 243)

The process of suggestion, for Allport, worked in conjunction with the formation and activation of attitudes. The development of the concept of the attitude as a measurable mental characteristic (Bogardus, 1925; Likert, 1932; Thurstone, 1929) paralleled the rise of behaviorism in the 1920s. Suggestions could help create an attitude based upon a prepotent tendency or could tap into and release pre-existing attitudes or “motor settings” (F. H. Allport, 1924, pp. 245–248). In all cases, the process involved either the creation of a new conditioned response or the activation of an existing one. Ultimately Allport’s characterization of the suggestion process echoed Sidis’s, 1898 description, wrapped in a S-R cloak. The nature of the process was

. . . that the one who gives the stimulus controls the behavior and the consciousness of the recipient in an immediate manner, relatively uninfluenced by thought, and through the method of building up motor attitudes, releasing them, or augmenting the released response as it is being carried out. (F. H. Allport, 1924, pp. 251–252)

Similarly-inclined social scientists looking at the forces of attitude change took up that behavioristic emphasis on “prepotent” tendencies and behavioral conditioning. Bernard (1926) observed: “Suggestion operates in almost every sphere and aspect of life. It is a short cut method of controlling effectively conditioned behavior” (p. 320). Frederick Lumley (1928) brought F. Allport’s behavioristic explanation of suggestion into his survey text, The Principles of Sociology, observing that “our impressionability is found in our predispositions, native and acquired” (p. 222, italics in the original). William Biddle (1931) outlined various categories of mediated influence including direct suggestion and indirect emotional conditioning which relied on “the drive of certain prepotent emotions” (p. 284). Irving Lorge and Carl Curtiss (1936) looked specifically at the role of positive and negative reinforcement in modifying attitudes, confirming a positive relationship in experiments on attitude change through prestige and suggestion.

Even those who rejected biological prepotent reflexes and adhered to a cleaner version of S-R (Dunlap, 1925; Morgan, 1924) or conditioning-based behaviorism (Krueger & Reckless, 1931) saw a role for suggestion in the formation and modification of attitudes and opinions. Looking back, the suggestion process—in crowd theory, instinct theory, and behaviorism clearly—offered a foundation for what later scholars would describe as a pre–World War II belief in direct effects. This view tells only half the story, however, which is one reason that most contemporary media effects histories fall short.

Demographic and Contextual Conditionality

Following Le Bon, early social scientists usually assumed vulnerable audiences included children and the poorly educated, with differential effects also being misogynistic, racist, or both. Ross cited the increased suggestibility of certain nationalities and ethnicities, including the French, Slavic peoples and Native Americans along, of course, with women (Ross, pp. 13–16; see also, P. C. Young, 1929). J.J.B. Morgan’s (1924) list of influences on suggestibility included both demographic and circumstantial factors. Bernard (1926) likewise stated that especially vulnerable groups included children, the “feeble minded” and “the uneducated and those inexperienced in the problems of life adjustment,” although even men and women of culture were susceptible to “skilled manipulators” (pp. 308–309). Lumley (1928) echoed Bernard:

The young, like the feeble-minded and the lower animals, are highly suggestible in the direction of their relatively few preconditioned responses . . . The uneducated and inexperienced are like children with respect to suggestibility. The same is true of the so-called backward races. (p. 223)

While Knight Dunlap (1925) conceded the significant variability among men and women in their degree of suggestibility, and “many women rival or exceed in independence of thinking the most original and unimitative of men”; nonetheless, overall, women were more suggestible. “Woman tends to perceive, to think and to feel that which is suggested to her,” as “careful experiments have determined” (p. 44). While agreeing that women and non-whites were more vulnerable to suggestion, Robert Gault (1919) maintained that this was due to educational and social disadvantages rather than inherent differences in gender or ethnicity.

Contemporaneous researchers, moreover, moved beyond social categories to allow for variability of suggestive influence across any and all personal traits and conditions. A partial list included congenital disposition, inexperience, emotional state, reverie, fatigue, hysteria, and appreciation of prestige (Conlon, 1929). Thorndike (1921), speaking in broad sociological terms, noted, “Each (person) has an individuality that marks him off from other men” (p. 93). The sentiment was shared by the early suggestion-based writers, including Trotter (1917), Weeks (1923), and Dunlap (1925). Ellwood (1922) observed, “The normal individual in everyday life is . . . more or less suggestible” (p. 229).

As a sensitivity to the mutability of suggestion unfolded, it encompassed not just individual characteristics but situational contexts (Murphy & Newcomb, 1937). E. Prideaux (1919) proposed that “suggestibility is more marked in those who live in the South and warm climates than those who live in the North and colder climates” (p. 231). In the American Journal of Sociology, Bernard (1917) argued that farmers and people living in rural areas were more susceptible to suggestion than city dwellers. Ultimately, McDougall (1909) observed,

The suggestibility of any subject is not of the same degree at all times; it varies not only according to the topic and according to the source from which the proposition is communicated, but also with the condition of the subject’s brain from hour to hour. (pp. 97–98)

Exploring the parameters of suggestibility, psychologists and social psychologists conducted scores of laboratory experiments in the 1910–1940 period. Many of them, especially those involving children, dealt with perceptual suggestion (Aveling & Hargreaves, 1921; Hurlock, 1930; Messerschmidt, 1933; Scott, 1910; Town, 1916). Many others, however, dealt with attitudes and attitude change and tested a range of factors that might increase or decrease vulnerability to suggestion. Unsurprisingly, demographics were often studied. Noting that “various age, economic and social groups exhibit different degrees of suggestibility” (p. 176), Claire Marple (1933) compared high school seniors, college seniors and adults and found younger subjects to be more suggestible. P. C. Young (1929), in an experiment with 323 white and 317 African American children, concluded that white children were less suggestible. He even measured differences between “light skinned” and “dark skinned” African American students. Henry Wegrocki (1934) tested whether suggestibility and intelligence were inversely related. James Bridges (1914) probed the relationship between suggestion and decision-making ability.

Suggestion was examined across different subject areas. An Army program of indoctrination, aimed at the “Americanization” of new recruits, used an inventive technique of suggestion through letter writing (Myers, 1921). The author concluded, “[l]ittle doubt is there that in education, one of the most effectual means for controlling conduct is by suggestion” (p. 31). William Burnham (1912) argued that suggestion was important “in all matters pertaining to the health of children” (p. 228) and advocated pedagogical suggestion to improve hygiene in schools. Researchers were able to manipulate the evaluation of paintings through suggestion (Farnsworth & Beaumont, 1929). Suggestion was found to sway religious beliefs (Burtt & Falkenberg, 1941) and, importantly, political ideology (Arnett et al., 1931; Kulp, 1934).

The nature of the message in its form and substance was found to affect suggestive power. P. C. Young (1931) tested “indirection” in suggestive messages, where the context diverted the subject’s attention. Pitirim Sorokin and John Boldyreff (1932) compared “persuasive” expert suggestions to “dogmatic” false ones; and Coffin (1941) studied the ambiguity of the suggestive content. Ross (1908), Alfred Conlon (1929), and others offered lists of additional mediating factors that included the duration of the suggestion, the volume (loudness) of the suggestion, even the time of day the suggestion was presented (Hollingworth, 1931). Researchers also considered the channels through which suggestion was communicated, comparing face-to-face interaction, speeches to groups, and printed material (Knower, 1935, 1936).

Prestige Suggestion

One of the more heavily studied variables was the perceived authority of the message source, what later came to be known as source credibility but at the time was labeled prestige suggestion. By the turn of the century, social scientists recognized “the immense importance of prestige in suggestibility studies” (Murphy & Newcomb, 1937, p. 174), and directed studies became popular in the 1920s and 1930s. Le Bon (1895/1903) himself remarked, “Whatever has been a ruling power in the world, whether it be ideas or men, has in the main enforced its authority by means of that irresistible force expressed by the word ‘prestige’” (p. 147).

McDougall (1909) described prestige suggestion as a lever in the efficacy of suggestion. Albert Poffenberger (1923) declared the powerful influence of an authoritative source to be a “well known law of suggestion” (p. 7). D.H. Kulp (1934) compared the effect of prestige suggestion across educators, social scientists, and lay citizens (the impact varied by respondent). Lorge and Curtiss (1936) experimented with prestige using behavioristic, reward, and punishment methods. Muzafer Sherif (1935), in a study with implicit early Gestalt undertones, associated prestige suggestion with stereotypes. Wegrocki (1934) examined the impact of prestige suggestion on emotional attitudes, concluding that girls tended to react more strongly than did boys to propaganda material and “there is a tendency for the more intelligent to be less influenced by propaganda than the less intelligent” (p. 394). Bowden et al. (1934) examined the relationships between prestige suggestion, age and gender, as well as emotion; he found younger people to be more suggestible via emotion, and women more suggestible generally than men.

Suggestion and Group Opinion

An individual’s perception of the group’s opinion, or group suggestion, was also seen as an important variable. As noted, Trotter employed instinct theory in considering the role of group opinion, or the “voice of the herd,” in shaping individual views. Looking at group opinion through the emerging lens of behaviorism and attitudes, F. H. Allport (1924) proposed that individuals could develop an “impression of universality,” the perception that a particular opinion was widely held within their reference group (F. H. Allport, 1924, p. 305). Doob (1935) elaborated that once individuals acquire an impression of universality, they tend “to be submissive and, therefore, more suggestible”; he emphasized that “all the empirical evidence gathered from crowds and the experimental data collected in the laboratory or under controlled conditions substantiates this contention” (p. 65). Edward Wheeler and H. Jordan (1929) offered laboratory confirmation that perceived group opinion tended to suppress the expression of individual opinion (although it was not, per se, a suggestion study). Leonard Ferguson (1944) added that susceptibility to group opinion “could be considered as a general personality trait” (p. 243). Herbert Barry (1931) even devised a specific test for susceptibility to majority opinion. (The impression of universality would later be reflected in the popular notion of the bandwagon effect.)

Among the experimenters comparing group suggestion to expert suggestion, H.T. Moore (1921) declared the suggestive influence of group opinion “to be beyond dispute” (p. 16), although he found little overall difference between expert and group opinion. Harold Burtt and Don Falkenberg (1941), in their study of religious attitudes, similarly concluded that both were effective at about the same levels. Dunker (1938) used both group and prestige suggestion to influence children’s food preferences but did not test for differences.

Even the general force of group opinion was conditioned, however (Ferguson, 1944). “That an impression of universality increases suggestibility may not necessary be true in every instance,” cautioned Doob (1935, p. 65):

When the existence of individual differences is acknowledged, as it has to be, then it becomes perfectly obvious that at least in social psychology, the sphere of unpredictability is going to remain a problem . . . No matter what system of motivation the psychologist cares to employ, it will remain true that some people do one thing for one reason and others for another. (p. 129)

In short, while suggestion could be powerful in shaping attitudes, no theory claimed it was omnipotent. There were always contingent conditions. In social practice, especially in the application of the suggestion doctrine to mass media, this conditionality was almost always acknowledged. Clearly then, experiments in attitude change, with an emphasis on the malleability of effects, did not begin with Hovland’s work in World War II, as some versions of the received history have it. Moreover, the laboratory work done in those pre-war decades helped provide the conceptual building blocks for an extended theory of media effects which were subsequently applied in broader, more practical contexts, outside the laboratory.

Suggestion, Media, and Anti-Social Behavior

Since the rise of the commercial press, theorists and public commentators had been worried about media’s deleterious social and cultural effects. Tarde, along with many others, blamed newspapers for fostering immorality, pornography, alcoholism, and juvenile delinquency (Bryant et al., 2013, pp. 25–27; Clark, 2010; DeFleur, 2010, pp. 129–130; Hardt, 2001, p. 157), although much of the critical commentary, even in academic outlets, constituted opinion pieces rather than what would be considered empirical research today (see, for example, Yarros, 1899).

In 1910 and 1911, Fenton (1910) published one of the earliest data-based analyses in the American Journal of Sociology. Her two-part study of the impact of newspaper coverage of crime on anti-social behavior drew heavily on Sidis and Ross in considering both “conscious” and “unconscious” suggestion. She concluded, “On the basis of the psychology of suggestion. . .a direct causal connection may be established between the newspaper and crime and other anti-social activities” (p. 370). This suggestive force was directly analogous, Fenton explained, to that of other media, including literature, moving-picture shows, and theater. Individuals varied in their suggestibility: “whether a person exposed to (external stimuli) gets a suggestion or not depends, in general, on the kind of person he is” (p. 368). Fenton’s general theme, however, was the pernicious effect of newspaper content. She therefore urged newspapers to be more circumspect in how they reported news and encouraged restrictive legislation.

According to Butsch (2008), Le Bon anticipated the suggestive vulnerability of audiences in theatrical performances, but Harvard psychologist Munsterberg (1916) was perhaps the earliest scholar to apply suggestion theory to film viewers; he also used it in other areas of study, including education (Munsterberg, 1910), and economics and advertising (Munsterberg, 1913). His 1916 analysis of film and film-going used suggestion to explain certain psychological processes in the viewing experience. Comparing it roughly to hypnosis, he wrote, “The spellbound audience in a theater or in a picture house is certainly in a state of heightened suggestibility and is ready to receive suggestions” (p. 108).

Joseph Geiger’s, 1923 article on the potentially dangerous influences of popular films employed a combination of suggestion, imitation and instinct theory. “This tendency on the part of the individual to experience the emotional states and instinctive behavior of other individuals who come under his observation, is what is meant by an imitative reaction.” (p. 78). Because he saw adults as largely immune from these influences, Geiger’s primary concern was the special vulnerability of children to suggestion: “The tendency to react to whatever ideas are presented in this uncritical manner is characteristic of the immature mind” (pp. 79–80). R. Armstrong-Jones, (1925) concurred in describing the suggestive impact of motion pictures on youth: “[T]he influence of suggestion extends even to the arts . . . Low-class cinemas play a great part in suggesting assaults, malicious damage and theft to the sensitive minds of young people” (p. 709). Albert Beckham (1933) explained that one cause of juvenile delinquency was the personality trait of “over-suggestibility.”

More prominent research on children and film came from the Payne Fund studies, published in 1933, particularly Herbert Blumer’s work. In his broader sociological writing, Blumer drew on Le Bon in the analysis of crowd behavior, but also described collective action as it was expressed in contemporary mass behavior and in the more deliberative form of publics, following Park. He saw mass behavior as a consequence of modernity wherein “migration, changes in residence, newspapers, motion pictures, the radio, education—all have operated to detach individuals from their customary moorings” (Blumer, 1946, p. 187). In some, although not all, contexts, therefore, people “could be influenced by excited appeals as these appear in the press or over the radio” (p. 188).

Blumer’s (1933) Payne Fund studies used diaries of young film goers to assess the impact of their viewing. There he relied heavily on imitation, reminiscent of Tarde, in explaining the effect of motion pictures on youth. He did not directly invoke suggestion theory, but his analysis did propose a concept that resonated with Le Bon’s emotional contagion in providing a pathway for psychological influence. Blumer (1933) stressed the role of emotional possession in young people’s interaction with film. While in the state of emotional possession,

the observer becomes malleable to the touch of what is shown. Ordinary self-control is lost. Impulses and feelings are aroused, and the individual develops a readiness to certain forms of action which are foreign in some degree to his ordinary conduct. (p. 198)

This state made young audiences more vulnerable to problematic models of behavior and susceptible to the manipulation of their general world view. Blumer wrote that “motion pictures may implant attitudes” (p. 194); “motion pictures organize [the viewer’s] needs and suggest lines of conduct useful in their satisfaction” (p. 195). Still, Blumer (1933) was careful to qualify his findings and noted that the influence of film content varied from child to child based not just on the child’s demographics but on his or her personal experience and life situation:

There is a wide variety in what people may select out of a picture. Its influence, consequently, is dependent not solely on its content but also upon the sensitivity and disposition of the observer . . . The differences are not merely matters of age and sex, but of cultural background and personal character. (pp. 179–180)

This conditionality was an important theme in Wartella and Reeves’ (1985) review of early research on children and media. They highlighted Blumer’s attention to individual differences, albeit without clearly contextualizing it within Blumer’s larger set of strong effects lenses. Butsch (2001), again, additionally argued that early media effects research often focused on the presumed special vulnerability of marginalized or subordinate groups, including the poorly educated, ethnic groups, and children.

Similar findings concerning both the power and the variability of media influence emerged in Blumer’s companion work on juvenile delinquency, which credited imitation and emotional possession as the mainsprings of anti-social influence among juvenile offenders (Blumer & Hauser, 1933). Films were only one source of influence and one that varied across individuals, however; overall, “criminal behavior is much too complex to be explained by any single factor” (Blumer & Hauser, 1933, p. 27). “People come to motion pictures from different social backgrounds and consequently are likely to have different attitudes and interests” which in turn “influence the selection and interpretation of what is seen” (Blumer & Hauser, 1933, p. 149).

The conjoint themes of influence via the suggestion process and variability of susceptibility were extended in both general application and theoretical scope in the era of propaganda research.

Suggestion and Propaganda

Before World War II, the term propaganda commonly encompassed both political and commercial communication. Suggestion served as the cognitive vehicle for both, but the doctrine made some its first practical appearances in commercial propaganda.

Twentieth century research on the psychology of advertising, from its start, was based on suggestion theory. In 1900, in one of the first articles on the subject, Harlowe Gale (1900) concluded, based on his experiments, that the “final buying of the often-impressed [advertised] article is very similar to all forms of suggestion” (p. 69, cited in Eighmey & Sar, 2007). Walter Scott, the preeminent figure in the social and psychology study of advertising in the first decades of the century, offered more pointed and more powerful applications of the doctrine. Scott wrote The Theory of Advertising in 1903 and The Psychology of Advertising in 1908, publishing five more editions of the latter over the next several years. For Scott, who “dominated the field of advertising psychology until 1910-1911” (Kuna, 1976, p. 348), the psychology of suggestion was the primary explanation of and tool for consumer influence. Appeals to reason were useful and had their place, he said, but should not be subordinated to the more effective force of suggestion: “the actual effect of modern advertising is not so much to convince as to suggest” (Scott, 1913, p. 83). Suggestion in advertising was especially useful because of its universal effectiveness; “every normal individual is subject to the influence of suggestion” (p. 81). Scott provided a host of recommendations for ways in which the advertising industry might employ suggestion; he declared a preference for word-of-mouth promotion through friends and companions but also advocated suggestion through direct advertising from manufacturers.

According to David Kuna (1976), “Scott’s theory of advertising, with the law of suggestion as its central tenet was the psychology of advertising during advertising’s formative era” (p. 353, italics in the original). Evidence of that influence is readily available. In 1923, Poffenberger called the “law of suggestion” central to effective advertising (p. 4). Conlon (1929) described suggestion as “the basis of the theory of advertising” (p. 56). Burtt’s Psychology of Advertising (1938) devoted a chapter to suggestion and its application. Even F. H. Allport (1924) noted its commercial application, offering a behavioristic approach: “Advertisers notoriously exploit human drives in building up an attitude to purchase their products. Here also repeated suggestion is used in the attitude-forming process” (p. 245).

Herbert DeBower’s 1917 Advertising Principles was published as one in a series of books on “Modern Business” serving a popular correspondence course established by business faculty at New York University. Advertising, DeBower stated, worked by “securing action through suggestion” (p. 91), including “suggestion by repetition” and “indirect suggestion” (p. 92). “Our most important moves and our most sacred conceptions are reached by means of suggestion” (p. 91).

The suggestion doctrine was also the principal model used in the analysis of political propaganda once concerns about state manipulation of public opinion accelerated after World War I. Sproule (1989, 1997) described a near national obsession with propaganda in the 1920s and 1930s. His insightful history and analysis highlighted the activities of progressive propaganda critics in the inter-war years and of their principal organization, the Institute for Propaganda Analysis (IPA). In his reviews—which spoke favorably of the propaganda scholars of this period—Sproule took issue with modern claims that those early observers had an unsophisticated, overly deterministic view of the power of propaganda. He countered that the writers and social scientists associated, for example, with the IPA were more focused on the practical task of identifying and countering propagandistic forces than with theorizing or uncovering underlying psychological processes (Sproule, 1997, p. 235). They “eschewed a theory of message transmission” (Sproule, 1989, p. 234).

On this point, however, Sproule may have missed the mark, at least in part. While some of the IPA associates may not have bothered with the psychology of influence, others did. Leonard Doob, a member of the IPA Board, and Hadley Cantril, its first president, relied heavily on the suggestion doctrine in their psychological analysis of media influence. Laboratory researchers found they could sway political ideology among students using prestige suggestion (Arnett et al., 1931). Its appearance in inter-war analyses of propaganda, therefore, was no surprise. In 1925, Dunlap described four propaganda techniques, including “logical argument,” appeals to “desire,” “simple suggestion” and “repetition” (pp. 248–255). He said the logical approach was adequate in “a relatively few instances,” but simple suggestion, wherein “the idea is implanted” in the individual, was more useful (p. 252, italics in original).

Some contemporary writers cite Lasswell’s (1927a) Propaganda Technique in the World War as one of the first scholarly statements of a direct effects model and even (controversially) as the origin of the hypodermic label (Lubken, 2008). While Lasswell did not actually use the term hypodermic, he did use suggestion. He defined propaganda as the “direct use of suggestion” (p. 9), adding “propaganda is concerned with the management of opinions and attitudes by the direct manipulation of social suggestion” (p. 9). He identified the “tactical principles” for deploying propagandistic suggestions (p. 208): “Suggestions may be spoken, written, pictorial or musical, and the possible variations in the form of the stimulus-carrier are infinite” (pp. 209–210). At the same time, he explained in a related journal article (Lasswell, 1927b) that he was not using the term suggestion in the typical contemporaneous sense but rather in cultural terms:

Putting the same thing into terms of social suggestion, the problem of the propagandist is to multiply all the suggestions favorable to the attitudes which he wishes to produce and strengthen, and to restrict all suggestions which are unfavorable to them. In this sense of the word, suggestion is not used as it is in individual psychology to mean the acceptance of an idea without reflection; it refers to cultural material with a recognizable meaning. (Lasswell, 1927b, pp. 630–631)

Most other scholars, however, employed the more conventional definition in applying suggestion theory to propaganda. For example, Albert Annis and Norman Meier (1934) tested attitude change through social suggestion in a controlled study of newspaper editorials. They concluded that favorable and unfavorable editorials on a topic unfamiliar to the readers and planted in the newspaper swayed opinion:

The process of suggestion is of an unconscious character, and, in the absence of known causation, the reader attributes the suggested ideas to his own reasoning. Suggestion in the form of selected news and editorial interpretations are accepted uncritically, since the reader is seldom able to verify the accuracy of the new material. (pp. 65–66)

Lumley, a pioneer in radio studies, had a strong interest in the problem of propaganda, addressing it in Means of Social Control (1925), Principles of Sociology (1928), and especially, The Propaganda Menace (1933). Drawing heavily from Ross and Bernard, he analyzed the forms and methods of political and commercial propaganda across all media platforms. He lamented the “ubiquitousness of suggestion,” since “an increasing amount comes from books and magazines and especially the newspapers and the movies. These tell us more and more what to think and what to do and how to feel” (Lumley, 1928, p. 225).

Even more prominent than Lumley was Yale’s Leonard Doob. In his 1935 book, Propaganda: Its Psychology and Technique, Doob combined elements of instinct theory, Gestalt psychology and behaviorism. Propaganda, he declared, “is one of the potent forces acting upon a group of people,” and facilitating propaganda was “the psychological process that controls and changes personalities and therefore simultaneously society itself. This process will be called suggestion” (Doob, 1935, p. 51, italics in original). He organized the psychology of propaganda around the emerging concept of individual attitudes, which he saw as formed on the bases of stereotypes, as articulated by Lippmann (Doob, 1935, p. 36). Suggestion, he explained, constituted the reorganization of attitudes as controlled by the propagandist.

F. H. Allport had proposed that attitudes could be formed and subsequently shaped by the suggestion process. Doob (1935) modified this proposal, arguing that “suggestion results from the manipulation of stimulus-situations in such a way that, through the consequent arousal of pre-existing related attitudes there occurs within the mental field a new integration which would not have occurred under different stimulus-situations” (p. 54, italics in original).

Doob examined all the mediated avenues of suggestion which served the interests of the propagandist: newspapers, radio, the cinema, the stage, books, pamphlets, even art. He sub-categorized suggestions into direct, indirect, and positive and negative forms and illustrated how each worked in practice, declaring that direct suggestion, such as simple commercial advertising, was more common, but indirect advertising could be more useful where the propagandists might wish to cloak their real intentions. He also noted the power of prestige suggestion and, as a sub-category of prestige, the impression of universality (pp. 61–66). He cautioned, again, however, that attitudes were not controlled directly but rather as a result of changes in an individual’s overall “mental field” and these varied across individuals in what he called the “sphere of unpredictability” (pp. 53).

Doob warned that the emerging medium of radio had significant potential to “arouse auxiliary attitudes” and thus serve the interests of the propagandist (pp. 360–361), including advertisers. But, he said, the medium was still young and had only begun to be exploited for such purposes (p. 359). A closer look at this emerging medium, however, was published, also in 1935, by Doob’s Harvard mentor, Gordon Allport, along with co-author, Hadley Cantril.

As radio grew through the 1920s from a hobbyist’s diversion to an integral part of the home environment and then public life, it increasingly became an object of scholarly interest. Much of the initial research involved measuring audience size and characteristics, the nature of the programming, assessing the potential of radio as an educational tool or looking at the preferred characteristics of a radio voice. Its social role, however, was yet to be deeply explored. As Cantril (1934) observed, “In spite of new problems created by radio, this rich field of investigation has been almost entirely neglected by the social scientists” (p. 316). It was a problem he and G. Allport (1935) addressed in The Psychology of Radio, which declared “Radio is an altogether novel medium of communication, preeminent as a means of social control and epochal in its influence upon the mental horizons of man” (p. vii). Their landmark text was the most thorough examination of the use and influence of radio to that date. The authors conducted dozens of laboratory experiments on the persuasive effectiveness of radio compared with face-to-face interactions or reading. The structure and wording of radio scripts, along with the pace of reading, were manipulated. Comprehension and recall of the material were assessed along with the degree to which different types of individuals were suggestible to radio content in various conditions.

Cantril and Allport (1935) borrowed the language of both crowd psychology and behaviorism to describe how radio demagogues worked to create a feeling of mass unity among their audiences. “No crowd can exist, especially no radio crowd, unless the members have a lively ‘impression of universality.’ Each individual must believe that others are thinking as he thinks and are sharing his emotions” (p. 8). By uniting people from across the country, radio fostered democracy and the creation of “common tastes and common attitudes” (p. 20). Nevertheless, recalling Le Bon, they warned that “One of the characteristics of a democracy is the ease with which individuals acquire a ‘crowd mind’” (p. 21). Although radio could not instill the same degree of crowd mentality achievable in a crowd setting because the audience members were not in physical proximity, “still to a degree the fostering of mob spirit must be counted as one of the by-products of radio” (p. 21).

With respect to both political and commercial propaganda, suggestion provided the psychological means of influence. “The mental mechanism upon which the propagandist relies is not reason but suggestion, which brings about the acceptance of a proposition for belief or action without the normal intervention of critical judgment” (Cantril & Allport, 1935, p. 62, italics in the original.) “All propaganda,” they declared, “relies chiefly on the use of suggestion” (p. 242).

Suggestion, following by-then accepted definitions, could be either direct or indirect. Direct suggestion involved connecting a suggested proposition to “some pre-existing attitude, need or symbol rich in meaning” (p. 60). Indirect suggestion involved the process of building up favorable attitudes toward an idea or product and could be employed “when it is essential that the propagandist conceal his purpose to make his propaganda effective” (p. 61). Still, the variability of influence for individual listeners continued to be a prominent theme. “Every listener brings to the radio program his own peculiar attitudes, prejudices, emotions, repressions, moods, and abilities. It is these characteristics of each listener that determines the particular type of psychic experience he seeks from the program” (p. 267).

Suggestion theory remained important for Cantril into the late 1930s; it appeared, albeit less obviously, as the psychological explanation for influence in his famous study of the Orson Welles’ 1938 War of the Worlds radio production (Cantril, 1940/1966). While not strictly a propaganda study, it nonetheless illustrated the then-assumed nature of the effects process. Cantril cited the nature of the program and the tenor of the times as among several important factors causing some listeners to believe the fictional drama about a Martian invasion. But suggestibility was the psychological key: “Those persons who were frightened by the broadcast were, for this occasion at least, highly suggestible. Those who were not frightened and those who believed the broadcast for only a short time were not suggestible” (Cantril, 1940/1966, p. 190). Cantril referred to the radio program itself as both “a suggestion” and “prestige suggestion,” given the credibility of CBS (p. 169).

While popularly portrayed as evidence of the power of radio to shape beliefs and attitudes, the War of the Worlds study is perhaps better known among social scientists for highlighting individual differences in media effects. It has been cited as a pre-cursor to the limited effects paradigm (DeFleur, 2010, p. 140). Indeed, Cantril rejected a universal view of suggestion and saw suggestibility as a variable human trait. It fluctuated largely on the basis of what the researchers called “critical ability,” operationalized as level of education, although other factors played a role, including seven dimensions of personality, such as insecurity, self-confidence, fatalism and religiosity (Cantril, 1940/1966, pp. 131–135). Cantril additionally distinguished “four psychological conditions that created in an individual the particular state of mind we know as suggestibility” (p. 190). These involved the degree to which individuals possessed and used various kinds of pre-existing standards of judgment (p. 191–197).

The reference to pre-existing standards signaled the start of an evolution in Cantril’s thinking about suggestion. His emergent model was articulated in 1941 in The Psychology of Social Movements, which offered an elaborate theoretical framework for analyzing the formation of public opinion, social movements, the process of attitude change, and media influence. Drawing on Gestalt theory, he started with a psychological foundation grounded in an individual’s “standards of judgment” and “frames of reference,” the latter being one’s general world view (Cantril, 1941, p. 20). From these frames of reference arose one’s “attitudes” about specific issues. These attitudes were subject to influence and even manipulation, however, through the process of suggestion. In his separate, detailed discussion of suggestion, he added a new conditional category: beyond the traditional “automatic response” of the suggestion process, he posited that suggestion could occur in periods of personal stress, confusion, or “bewilderment” (p. 64):

A person is susceptible to suggestion when (1) he has no adequate mental context for the interpretation of a given stimulus or event or (2) when his mental context is so rigidly fixed that a stimulus is automatically judged by means of this context and without any examination of the stimulus itself. (p. 64)

He acknowledged that the first condition—vulnerability to suggestion “because they are puzzled”—was under-studied, but he considered it possibly more important than the more traditional context of suggestion. In such conditions, people may seek an appropriate frame of reference; especially when fear or stress characterizes the situation, “they are unusually likely to accept whatever interpretation is offered as long as it seems plausible” (p. 65). The War of The Worlds research illustrated the phenomenon, but Cantril extended the analysis to the nature of 1930s social and political movements, including the rise of the Nazism, and the use of propaganda by those movements. As to the second condition of suggestion, Cantril (1941) said it was a mistake to see the process as a simple behavioristic “conditioned response” (pp. 55–56). Instead, he tied this type of suggestion to the fit between the proffered message and the individual’s frame of reference. The acceptance of the message therefore was automatic only insofar as it comported with a person’s world view and that individual declined to further investigate the substance of the claim.

The book was Cantril’s last major statement on suggestion, and one of the last significant expositions on the doctrine in social psychology. His integration of suggestion into Gestalt theory also signified the growing influence of Gestalt psychology, which critiqued and eventually displaced suggestion theory. The nature and course of that displacement, and the concomitant decline of the suggestion doctrine, are traced below, first looking at the role of emotion in suggestion and then at the political manifestations of a related and critical underlying issue involving the question of cognitive rationality.

Suggestion and Emotion

An enduring sub-theme in much of the above research involved the role of emotions in the suggestion process (Conlon, 1929). Individuals were regarded as more susceptible to suggestion when aroused emotions circumvented rational deliberation, in what one writer termed, the “emotional short circuiting of thought” (English, 1932, p. 402). Emotions were seen as a conduit or pathway for suggestive influence (Hollingworth, 1935).

The study of emotions as powerful drivers of human behavior is ancient. Emotions were important elements in the mid-19th century work of Spencer and Darwin, and later, James (Gendron & Barrett, 2009). For Le Bon and Tarde, as already noted, the role of emotions and emotional contagion in spreading influence through a crowd was important. Blumer, again, saw emotional possession as a vehicle for influence. Angell (1941) described the power of the cinema to lift public morale in World War II by way of its ability “to exploit the principles of crowd psychology” including the contagious transmission of “emotional exaltation” (p. 355).

Propaganda analysis was rife with emotion-based explanations for suggestion. Strong (1922), in his analysis “of propaganda as a psychological problem,” noted “the deliberate process of arousing one’s emotions and desires and then suggesting a line of action by which these desires may be expressed” (p. 242). Scott (1913) saw emotional manipulation as important in advertising suggestion. Edward Bernays’s (1928) classic text Propaganda explained that “the group has mental characteristics distinct from those of the individual, and is motivated by impulses and emotions which cannot be explained on the basis of what we know of individual psychology” (p. 47).

In instinct theory, McDougall’s original articulation of biological tendencies was tied directly to, and even expressed as, emotions, such as fear and disgust. Emotional arousal as a conduit for suggestion was even incorporated into early behaviorism, which advanced a theory of “emotional conditioning” (Dodge, 1920; Doob, 1935, p. 115; Macpherson, 1920). Biddle (1931) proposed that pre-existing, or prepotent, emotional drives were associated with and conditioned by political propagandists and advertisers to establish a favorable relationship between an emotion and a desired attitude or behavior, declaring “the methods of the new propaganda make it become a process of indirect emotional conditioning on large scale” (p. 284).

In 1937, the IPA issued its now-famous list of the seven devices of propaganda, including name calling, glittering generalities, and the band wagon. The IPA stated:

[I]n all of these devices our emotion is the stuff with which propagandists work. Without it they are helpless; they harness it to their purposes, they can make us glow with pride or burn with hatred, they can make us zealots in behalf of the program they espouse. (IPA, 1937, p. 3)

Again, however, as Blumer noted, sometimes people were susceptible to emotional influence and sometimes they weren’t. Contingencies always applied. And while emotional stimulation could in some situations facilitate suggestion, and this was socially problematic, importantly, a much deeper psychological and even philosophical issue was at play. Emotions are the obverse of cool-headed calculation; the suggestion process itself was, by definition, one that circumvented rational thought. The larger question, then, was the extent to which humans were inherently susceptible to such forces at the outset. That is, were humans basically thoughtful, deliberative creatures, or not? Debates about the validity of suggestion theory and the long-term fate of the doctrine eventually revolved around this ancient problem.

Fear and Faith

The question of the extent to which humans are cognitive captives of an intrinsically irrational mind is ancient in philosophy, tracing back to Plato. Enlightenment thinkers offered optimism on this score. Pessimistic German philosophers of the 19th century, along with the French crowd theorists, cast doubts. The survival of suggestion hinged on the eventual outcome of the dispute, and the inherent tension was expressed early the life of the doctrine.

The importation of European crowd theory along with the rise of Freudian psychology threw growing suspicion on the Enlightenment model, leading to what G. Allport (1968) called “the supremacy of irrationalism” (p. 20) in social psychology in the early part of the century. As Scott noted in 1908: “Today we are finding that suggestion is of universal application to all persons, while reason is a process which is exceptional, even among the wisest. We reason rarely, but act under suggestion constantly (pp. 82–83). Suggestion theory, especially coupled with widespread discourse on propaganda, therefore fostered unease in the academy and among the public about society’s ability to resist media manipulation.

Yet, in sharp contrast, Sproule has argued that despite this expanding apprehension about our susceptibility to communicative influence, progressives of the period remained confident in individuals’ ability to surmount the effects of propaganda:

Progressives had faith in the essential cognitive competence of the public, believing that all that was necessary to combat propaganda was to inform the public about how modern institutions diffused their ideologies through news, religion, entertainment, education, and government. (Sproule, 1991, pp. 219–220, as quoted in Jowett & O’Donnell, 1999, p. 106)

In fact, while much of the research on suggestion, especially in its application to propaganda, was implicitly pessimistic with respect to cognitive vulnerability, a thread of optimism ran through most of the scholarship. The potential to reduce or eliminate the deleterious power of suggestion was either explicit or implicit in the earliest writing on the doctrine. This, importantly, was a consequence of the perceived variability of the suggestion process itself. Insofar as those with greater intelligence or sophistication were typically seen as less suggestible, the clear path to countering any cognitively automatic suggestion was to educate and elevate the individual. The irrational or reflexive mind could, with work, be turned rational.

This issue was central to the broader Lippmann-Dewey debate over participatory democracy and was manifest in much of the period’s suggestion research in what was essentially a tension between fear and faith. In 1895, Jeremiah Jenks, writing in the first volume of the American Journal of Sociology, acknowledged that public opinion was composed largely of emotion, sentiment and suggestion. Nonetheless, he urged “thoughtful men” to take the lead in developing means and methods “of training that will lead our people more and more consciously to wish to free themselves from prejudice and to shape their lives in public matters more and more by judgment” (p. 169). Similarly, Ross (1896) looked to strong leadership from social and educational elites:

Scattered through the heedless mass who throw up their caps for bold rascality, and jeer at patient labor for the unmanifest public good, are sagacious men who say “Do we dare encourage this?” “Do we dare discourage that?” By reasoning, by expostulation, by warning of consequences the mass is enlightened, the sober second thought prevails and the public is prevented from quarrelling with its own interests. (p. 765)

Lumley (1925) drew on Dewey to argue that education could be a prophylactic against propaganda (p. 210). Others agreed (Armstrong-Jones, 1925; Dunlap, 1925; Weeks, 1923). While describing, on the one hand, the power of suggestion as “resistless” as “the pressure of the atmosphere” (p. 208), Weeks (1923) nonetheless lobbied against any form of censorship on the grounds that mediated communication provided the public time to think more carefully about suggestive messages. He offered a faith in a pure libertarian model: “The best way to dispose of a perverse suggestion is to oppose it to a better one” (p. 219). Dunlap (1925) concurred:

if we use efficiently against evil propaganda the same weapons which it uses, and which are justified if our repetitions and suggestions and appeals to desires are based on scientific and logical consideration, these latter forces must surely, if slowly, turn the battle in favor of the right. (p. 257)

Even Cantril and Allport (1935) were ultimately hopeful about democratic potential of radio. Having emphasized the conditionality of influence, they noted that people were less vulnerable to suggestion via radio than they were in a crowd setting; radio disengaged an individual from the emotional pull of the mass. As isolated radio audience members, people were “less crowdish,” less emotional, and less susceptible to suggestion (p. 13). So, despite the power of propaganda, they concluded that radio held the potential to break down political and cultural barriers, to foster internationalism and become an “agent of democracy” (p. 22). Cantril (1940/1966) dedicated the closing lines of Invasion from Mars to advocating for stronger efforts at maximizing critical ability: Education was “one of the greatest preventatives of panic behavior” (p. 204).

Dueling Paradigms

This optimism about potential cognitive capacity grew even stronger in the 1930s, carving broad, new channels in social psychology and simultaneously eroding the attraction of suggestion theory. The more optimistic, progressive reading of rationality was expressed in several ways in social science, all with implications for suggestion. Christian Borch (2012) argued that proponents of more rationalist social-psychological models sought to undermine suggestion, in some cases motivated by political ideology:

The reference to suggestion, with its underlying emphasis on features that countered rather directly the ideal of the constituent subject, was hard to digest for liberal sociologists who launched a vendetta against the concept, eventually rendering it sociologically inappropriate. (p. 17)

Borch cited Park in sociology and Mead in social philosophy as examples of an effort to reinvigorate rationalism in the social sciences. Influenced by Le Bon, R. Park (1972) appreciated the problematic suggestive power of crowds and their de-individualizing effects. He saw publics, however, as a different kind of social construction. They were more lasting, hopefully more stable, and capable of enhancing, rather than suppressing, individual thoughtfulness. In contrast to the anarchy of the crowd, the public “is guided by prudence and rational reflection” (R. Park, 1972, p. 80). Importantly, Park saw crowds and publics as steps in the evolutionary development of social structure and hence held positive rather than destructive social implications over the longer term.

Another notable Chicago School figure, George H. Mead, took on suggestion, along with imitation, more directly. Mead’s efforts to fashion a theory of development of the social self in its interaction with others was seen by Ruth Leys (1993) as a pointed counter-theory to Tarde’s imitation process. According to Leys, “we cannot understand Mead’s thought unless we grasp that what is mobilizing and impelling it from beginning to end is the resolve to defeat Tarde’s theory of imitation-suggestion” (quoted in Borch, 2012, p. 158). Both Park and Mead, therefore, nudged suggestion theory toward the margins of social psychology.

The Gestalt Critique

It was Gestalt theory, however, that provided the spearpoint in the rationalist assault on suggestion. Gestalt psychology arose in Europe around the turn of the century and made its way into the United States in the 1930s (Koffka, 1935). It proposed a rational model of attitude change in contrast to irrational cognitive processes—specifically, suggestion. This reorientation of the understanding of persuasive communication was led by psychologists such as Asch and his student, Helen Lewis (1941), who called into question the validity of the “mechanistic hypothesis” of suggestion. Instead of a process of influence characterized by the non-critical acceptance of an idea, they saw a much more considered act of judgment. Opinions changed, they asserted, when a reasonable interpretation of the material changed, through, for example, a change in the authority of the authorship (the prestige of the source). Lewis (1941) explained, “when changes in judgment occur, it is normally because the material to be judged is seen in a new light and has consequently changed its meaning” (p. 230). Such changes “need not turn out to be uncritical or unreasoning” (p. 230).

Gestalt principles quickly gained popularity in the social sciences. By the end of the 1930s, both G. Allport and Cantril were incorporating elements into their work. Paul Cressey (1938) employed Gestalt theory in his analysis of the impact of film on youth. Allport and Faden (1940) used Gestalt concepts and citations in their analysis of the psychology of newspapers. As noted, Cantril’s, 1941 text on the psychology of social movements had a clear Gestalt dimension. Gestalt theory even slipped into media sociology. Robert Merton (1946) considered the highly successful one-day campaign by singer Kate Smith and CBS to sell war bonds as an exercise in mass persuasion. The research was seated principally in the mass society tradition of Durkheim and Tonnies (pp. 142–146) and Merton treated social psychology only in footnote, but there he listed Gestalt psychology, psychoanalysis, field theory, and “the sociology of knowledge” as his theoretical touchstones (footnote 3, p. 116).

With its ascendance, proponents of the Gestalt perspective became more overt in their critique of suggestion. In 1948, Asch penned a sharply worded takedown of the doctrine. He conceded, as previously noted, the success and widespread use of suggestion:

We have today in social psychology a far-reaching and widely adopted theory of the effect of group forces on the formation and change of opinions and attitudes. At its center are the concepts of suggestion, prestige and imitation—terms which in social psychology are virtually interchangeable. (p. 250)

He noted that research on opinion change had gained “added importance in recent times when the production of such effects has been harnessed and institutionalized in the form of mass propaganda and advertising” (p. 250). His intent, however, was to systematically dismantle the doctrine, both theoretically and methodologically. After briefly reviewing the then-extant work on the suggestion process generally and prestige suggestion specifically, Asch (1948) concluded that “the doctrine of suggestion has not been proven” and that “investigations do not support the conclusion that have been drawn from them, or the assumptions upon which they were based” (p. 251).

In 1952, the Society for the Psychological Study of Social Issues released a revised edition of a 1947 textbook designed specifically as a “standard source book for social psychological courses” (p. vii). The earlier edition had offered a section on “suggestion, imitation, and sympathy,” which included Coffin’s (1947) article on the conditions of suggestion, followed by a reprint of Lewis’s (1947/1952), 1941 Gestalt-based critique of suggestion. The 1952 update moved those two pieces to the first section of the book but headed this with a new article by Asch (1952a). There, he declared that despite years of research into attitude change, public opinion formation and propaganda, “[t]oday we do not possess an adequate theory of these psycho-social processes” (p. 2). Approaching these questions through S-R theory produced unsatisfying results, and the inadequacy of the operation of suggestion and prestige were “becoming increasingly obvious” (p. 2). Asch extended his critique in a full chapter of his influential text Social Psychology (1952b).

In 1953, learning theorists Arthur Benton and Albert Bandura refuted the notion that any general trait of suggestibility could be substantiated by the extant literature; consequently “in recent decades the utilization of the concept as an explanatory principle in psychopathology and social psychology has declined markedly” (p. 336). In 1960, Daniel Katz recalled the doctrine of suggestion in early advertising research as an example of irrational psychological models of influence. It was part of his effort to reconcile rational and irrational processes through a functionalist model of attitude change. It was cul-de-sac in the literature, however; for 20 years, no other researchers picked up his proposal (Carpenter et al., 2013). By the end of the 1950s, suggestion theory had become, for most, a distant memory. Even G. Allport (1968) acknowledged its decline: “In recent years, considerable skepticism has been expressed concerning the utility, and even the validity, of the concept of suggestion” (p. 38). Nevertheless, he added, extant questions concerning the nature and definition of rational message interpretation meant “we still need the concept of suggestion” in certain, well-delineated cognitive contexts (p. 39).

It is worth noting that the abandonment of an irrational cognitive model was not just a theoretical affair constrained within the high, cloistered walls of academia. In the socio-political context of the period, it had a significant public policy dimension. The term propaganda became increasing suspect in the late 1930s as the United States began facing the reality of another world war and the government sought to rekindle its own campaigns of public information and motivation (Stole, in press). For more than 30 years, social scientists and public commentators had been warning of the pernicious effects of propaganda. But with war brewing, sentiment changed. As described by Sproule (1997), propaganda critics generally and the IPA specifically were increasingly seen as fostering unpatriotic skepticism toward modern communications apparati, both public and commercial. Pressure mounted for them to shift toward attitudes more supportive of the war effort. Many did. The IPA shuttered its doors and psychologists and social psychologists began lending their skills to the War Department. And within this broader cultural swing, the nomenclature of media influence itself underwent significant change. As Sproule (1997) described:

Beginning in 1941, propaganda, the muckraker’s signifier, underwent a series of metamorphoses whereby the term first was replaced by morale and later by psychological warfare. By 1948, persuasion, communication and information were the favored locutions for what had formerly been called propaganda. (Sproule, p. 217, italics in original)

As described below, a similar fate awaited suggestion.

Hovland’s Model

Hovland and various colleagues conducted what are now landmark studies in mass communications effects during World War II and afterward at Yale University. Hovland, of course, was one of the named “founders of the field” in the Berelson-Schramm origin story; as noted, many contemporary histories trace the start of effects research back to his work, along with that of Lazarsfeld. Hovland’s role in the process of paradigm change was twofold. First, he never adopted suggestion as an explanation for attitude change and, especially given his stature in the field, he thereby contributed to its continuing general erosion. Second, in the form of his writing, specifically his use of references and terminology, he aided in the erasure of suggestion from the field’s history.

Hovland and Janis famously considered a host of factors that impinged on the process of attitude change, including manipulation of source credibility, the organization of arguments, types of media used in the communication process, and demographic and personality traits of audience members, what Lowery and DeFleur (1983) would later call “the magic keys” of media influence (p. 148).

Hovland’s theory of attitude change drew primarily from learning theory, minus suggestion, as developed by his mentor at Yale, Clark Hull. William McGuire (1996) described Hovland as a “convergent stylist,” who used “low-level theorizing eclectically” (p. 52) to consider more direct experimental issues in independent—dependent variable relationships. Hovland’s earliest work, in fact, contained little articulated theorizing. Hovland et al.’s (1949) Experiments on Mass Communication emphasized method and findings. The authors implied a learning model but repeatedly alluded to the applied nature of the studies; discussions of psychological processes were minimal.

Hovland’s theorizing did evolve; he would later propose social judgment theory with Sherif and Hovland (1961). One of the first elaborated discussions of his approach appeared in 1948 (Hovland, 1948), where he offered a reinforcement-based, S-R learning model in which irrational suggestions had no place. If anything, he privileged a Gestalt-inclined rationalism. Communication and Persuasion (Hovland et al., 1953) presented opinion change as a form of incentivized, or rewarded, learning with Hovland and his co-authors viewing the mental operation as largely reasoned: “A major basis for acceptance of a given opinion is provided by arguments or reasons which, according to an individual’s own thinking habits, constitute ‘rational’ or ‘logical’ support for the conclusions” (p. 11).

But while Hovland and his team did not use suggestion theory, they were clearly aware of it. Hull, before he switched to behaviorism, was a leading figure in suggestion and hypnotism research. Hovland and Mandell (1952a) employed suggestion terminology, if not theory, in a 1952 experiment. Hovland was also close colleagues at Yale with Doob, who, as noted, relied on suggestion in his studies of propaganda and, in fact, was gratefully acknowledged for serving “informally as the editor” of Communication and Persuasion (Hovland et al., 1953, p. vii). That landmark text, along with Personality and Persuasibility (Janis et al., 1959), included several citations to work previously done under the rubric of suggestion.

Suggestion as a process, therefore, was absent in Hovland’s work in large part as a reasonable consequence of his choice of theory. More interestingly, and perhaps more powerfully for historical purposes, the associated terms themselves—suggestion and suggestibility—were also largely absent.

Hovland’s Linguistic Pivot

Hovland and his colleagues referenced suggestion research sparingly and sometimes altered the relevant nomenclature when they did. Hovland’s earliest publications did not cite prior suggestion studies or even use the term suggestion. This, again, was arguably due to his use of learning theory but possibly also because of its pre-war association with propaganda. Hovland’s research unit appreciated the general suspicion and distrust reserved for propagandist communications (Sproule, p. 199). They even used skeptical attitude toward propaganda as a variable in their Army studies (Hovland et al., 1949). In this context, it is easy to see how suggestion, along with propaganda—terms associated with purposive cognitive manipulation – might have been avoided during the war years. In fact, most of the language in his initial publications was rhetorically neutral; the subject under study was termed opinion change, message acceptance or communication effectiveness (Hovland, 1948, 1951; Hovland et al., 1949).

The one overt appearance of suggestion, across all of Hovland’s work, came in a co-authored 1952 article on “conclusion drawing” (Hovland & Mandell, 1952a). The question under study was whether it would be more effective for a communicator to provide a conclusion to an audience or provide only information and arguments and have the audience determine the implied conclusion on its own. The test was, as they phrased it, whether “indirect suggestion is more effective than direct” (p. 581). They divided the experimental subjects into those “resistant to suggestion” and those who were “suggestible” (p. 586), although beyond the labels, there was no reference to underlying cognitive processes.

Later that year, however, Hovland’s nomenclature took a significant and lasting turn, and what had been for him attitude change, or briefly suggestion, became persuasion, arguably a term with a more rational valence than suggestion. The reasons for the change are unclear but there are grounds for supposition. Hovland’s adoption of a learning model with a rationalist Gestalt dimension may have had an influence. The term persuasion was known in media studies, although it had been used only sparingly, for example in Merton’s (1946) Mass Persuasion and in a 1949 Journalism Quarterly review of communication research (Schramm, 1949a). Persuasion, however, was closely associated with the study of rhetoric and Hovland’s use of an older rhetoric-based experiment may have been a factor. To study order effects in message transmission, Hovland and Mandell (1952b) had reached back to Frederick Lund’s, 1925 study on the subject. While a psychologist, Lund (1925) nonetheless grounded his analysis in the discipline of rhetoric and speech. His context was “textbooks dealing with composition, oratory, debate and argumentation” with special emphasis on debate and “the art of persuasion” (pp. 183–184). Hovland and Mandell (1952b) replicated Lund’s experiment in both its method and terminology. Their 1952 paper, “Is there a ‘law of primacy in persuasion?’” was the first significant appearance of the term in Hovland’s work. In 1953, with the publication of Communication and Persuasion, it became the de facto label for his research.

The adoption of the term had field-altering import. Persuasion implicitly, if not explicitly, entered the rationalism-irrationalism debate and indicated some degree of thoughtful evaluation (Larson, 1992) or freedom of cognitive choice (Perloff, 2010) against suggestion’s irrationalism. More importantly, perhaps, Hovland began using the term even when recalling previous suggestion studies, sometimes turning what had been research on suggestion into research on persuasion. This, again, was in conjunction with an infrequent use of prior suggestion research generally. The 1952 “conclusion drawing” study with Hovland and Mandell (1952a), for example, had only four citations, three of them to Hovland’s own work, none to suggestion studies despite their use of the term.

A heavily-cited Public Opinion Quarterly article with Walter Weiss (Hovland & Weiss, 1951) on source credibility illustrated both the issues of citation and language. Despite decades of research on prestige suggestion going back to Le Bon, Hovland and Weiss began their influential piece with a historically problematic claim: “An important but little-studied factor in the effectiveness of communication is the attitude of the audience toward the communicator” (p. 635). They referenced prior work, but briefly and tangentially, stating that studies of “prestige” had offered only “indirect data on this problem” (p. 635). They provided only three citations, all of them to Gestalt-based studies. And the text of the article employed the labels “prestige,” “prestige studies” and “prestige technique,” but never “prestige suggestion,” although the latter term was in the title of their citations.

This linguistic distancing from the term suggestion became characteristic of Hovland’s writing, even when previous suggestion studies were cited (Hovland, 1954; Hovland et al., 1953; Janis et al., 1959). For example, in Communication and Persuasion, Hovland et al. (1953) described approvingly and in detail the work of Ferguson (1944) who, they wrote, “used objective indices of susceptibility to persuasion” (Hovland et al., 1953, p. 177). Nowhere in their description of his work does the word suggestion appear. Yet, the title of Ferguson’s article was “An analysis of the generality of suggestibility to group opinion,” and the “indices” noted by Hovland in fact employed “suggestibility scores” and “suggestibility coefficients.” Ferguson never used the term persuasion; Hovland never used suggestion. Another suggestion study, by Wegrocki (1934), underwent a similar linguistic transmutation in the book (Hovland et al., 1953, pp. 181–182).

The change of nomenclature did not go unnoticed. G. Allport’s (1968) history of social psychology observed that “recent years” had seen an increase in studies on “the conditions of suggestibility . . . employing the term persuasibility rather than suggestibility,” specifically citing Hovland (p. 37, italics in original). The importance of Hovland’s choices in theory and more critically in citation and language for the remembered history of the field cannot be overstated. In this way his work severed connections with the suggestion doctrine, as he embarked on a scholarly journey characterized by learning theory and persuasion.

Hovland’s “Persuasible Personality”

At the same time, it is worth noting that while the language of suggestion theory did not appear in Hovland’s principal work, its underlying cognitive functionality—that of bringing about “the acceptance of a proposition for belief or action without the normal intervention of critical judgment”—did. As his model evolved, it began absorbing essential elements of unreflective, automatic attitude change. Hovland’s research program at Yale began with a long list of interacting variables said to condition the persuasive process. Many of them were mechanical or contextual, such as the difference among media channels. Over time, however, the program grew more interested in the psychological dimensions of attitude change (Farr, p. 5). And while Hovland started by citing Gestalt theory, eventually some of his conclusions about the persuasion process, despite the rationalist implications of the term, turned back to explanations of effects based on more irrational cognitive models.

Hovland and Janis came to believe that at least some people were inherently subject to a process of non-rational acceptance of influential communication. Chapter 6 of Communication and Persuasion (Hovland et al., 1953) was devoted to a preliminary investigation of the traits that spoke to an individual’s “general susceptibility to various types of persuasion and social influence” (p. 174). Studies of what the authors described as “topic-free” (p. 175) or “general” (p. 199) persuasibility became a special focus of their interest, especially for Janis. As they explained, “It seems probable that there are certain types of individuals whose personality needs incline them to be highly gullible” (p. 184).

Ideas about which personality factors related to this inherent susceptibility were advanced tentatively in Communication and Persuasion and extended in their book-length treatment of the subject, Personality and Persuasibility in 1959. What they called “persuasibility factors” included personality traits such as self-esteem, hostility, authoritarianism, and personality disorder. Intelligence, age and gender also were considered. But, with the possible exception of self-esteem (which showed some signs of an inverse correlation with persuasibility), most of these root-cause explanations proved unsatisfying. Intelligence and demographics did not show consistent relationships to persuasibility. Ultimately, the researchers concluded that:

personality factors appear to play only a very minor role as determinants of persuasibility. Almost all the correlations obtained so far are relatively low, and even those that are statistically significant can account for only a small portion of the variance. (Janis et al., 1959, p. 248)

They decided that it was “probably premature, however, to draw any definite conclusions,” adding that future testing with improved methods of measurement might yield better insights (p. 248). They proceeded to lay out 10 “primary persuasibility profiles,” describing patterns of traits associated with high to low persuasibility. The two most persuasible patterns were labeled “motivated gullibility” and “unintelligent gullibility” (p 267).

The results of the Yale team’s research on general persuasibility received several pages in The Effects of Mass Communication (Klapper, 1960, pp. 72–76), which observed that “recent research strongly suggests that some persons are in general more persuasible than others” (p. 72). For decades, suggestion theorists had plumbed the nature of a possible trait of suggestibility. But once again underscoring the loss of all memory of that history, Klapper declared, “The fact that some persons are, at least at a given time, consistently more (or less) persuasible than others was apparently first demonstrated by Janis (1954)” (p. 73).

Lazarsfeld and Suggestion

While Hovland explored the psychological conditions of persuasion, Paul Lazarsfeld and his team at Columbia’s Bureau of Social Research (BASR) looked at the social system. Here, suggestion theory trod a different path, albeit with the same destination. Lazarsfeld (1940) acknowledged, and even employed, suggestion theory early on. In 1940, he used it to explain the potential impact of radio and the importance of “the accuracy and impartiality of the flow of information coming over the air” (p. 256). “That this qualification is of extreme social importance is plain when we consider that people on the lower cultural levels are apparently more suggestible than those on the higher cultural levels” (p. 256, italics in original). And, insofar as his research indicated that people with lower levels of education preferred radio over newspapers, and radio listeners were “more suggestible” than newspaper readers, there was reason for concern. “Of all the facts that make radio a powerful social institution, probably the most imposing one is that radio is the preferred medium of the more suggestible man” (Lazarsfeld, 1940, p. 257).

By the time of Lazarsfeld’s survey work on political decision-making, however, the “suggestible man” was gone. The People’s Choice (Berelson et al., 1944) found that propaganda converted only a small minority of subjects; the authors emphasized the stability of political attitudes and the practices people adopted to protect their pre-existing beliefs. While they discussed the mechanics of selective attention in mediating communication effects, (along with the role of personal influence and the two-step flow), their study provided little elaboration of psychological processes. Asch (1952b, p. 531) subsequently cited this absence as a failing of the research. In response, in their next book, Voting, Lazarsfeld and Berelson (Berelson et al., 1954) pointed out the daunting complexities of identifying generalizable “psychological mechanisms”; they described relevant research, broadly, as “studies of suggestibility,” but it was a passing notation (p. 291).

Published in 1955, Katz and Lazarsfeld’s Personal Influence offered “prestige suggestion” as one component in persuasion, acknowledging and footnoting the “very large number of studies” on the subject (p. 24). But the phenomenon was presented as a process just recently discovered by Gestalt psychologists: “The study of how such changes come about . . . was almost completely ignored, and apparently misunderstood, until Gestalt psychologists began to investigate what went on in the subjects’ minds when they responded to prestige suggestion” (p. 69). The topic was dropped at that point—“we shall not deal with that subject here” (p. 69)—and the narrative moved on to a discussion of group influence. As with Hovland, there was little looking back to suggestion.

Jeff Pooley (2006) has argued that Lazarsfeld, and the field of public opinion generally, “was largely cut off from, and uninformed by, its media research predecessors” (p. 147). This, he credits in part to Lazarsfeld’s well-known disinterest in communication per se, which “contributed to his ignorance and lack of curiosity about those currents of media analysis that preceded his own” (p. 147, italics in the original). Delia (1987) also noted this absence of significant citation to pre-World War II writing in the publications of both Hovland and Lazarsfeld: “Hovland et al. (1953) largely ignore previous persuasion research” (p. 63), and Lazarsfeld was “working largely independent of past scholarship” (p. 65). This characterization may be a little unfair to Hovland, who did provide significant citation in his later work. Nonetheless, as a consequence of choices in theory, citation and language, modern scholars who looked back to the Columbia-Yale corpus as the beginning of media research, as many do, found little evidence of a widespread use of suggestion before World War II.

This absence of suggestion was unsurprisingly mirrored in Klapper’s, 1960 review. His index listed multiple references to persuasion but had none for suggestion (although the bibliography listed a few suggestion studies). Likewise, his treatment of selective perception, conceptually central to the limited effects model, was illustrative. Klapper observed that the phenomenon had “long been noted and often detailed” in social psychology and communications research (p. 22). He cited prior studies on the topic, several from Gestalt theorists, one (Bruner & Goodman, 1947) that referenced suggestion theory, albeit briefly. Klapper might, however, have taken note of Coffin’s extended 1941 review of suggestion in which the author reported:

The results of our first experiments are univocal in speaking for the importance of pre-existing attitudes in determining the acceptance or rejection of suggestions . . . It is found that subjects accept, in general, those suggestions which “fit” with their existing attitudes, even when alternative suggestions are available, offered simultaneously and in similar manner. (p. 110)

Across the post–World War II literature, then, references to suggestion diminished almost to the point of vanishing, leaving little evidence of its existence for future scholars. Meanwhile, the very nature of research into opinion formation and change was conceptually reversing. Instead of assumptions about a powerful media, the emerging social science models powerfully mitigated against influence. Sociological structure-functionalism identified factors that intervened between sender and receiver, not the least of which was the two-step flow. There was a growing interest in cognitive consistency (Festinger, 1957) powering the mental selectivity processes that impeded attitude change. Attention turned away from forces that worked toward attitude change and went to forces that repelled it in a limited effects world that made little room for non-critical acceptance of direct suggestions. At the same time, the World War II generation of communications scientists did glance back just long enough to sketch their own compact version of the history of mass communication effects research.

The BASR and the Hypodermic Model

Pooley (2006) has provided a rich, detailed review of the intellectual cauldron out of which the new history arose. He has argued that Lazarsfeld, in search of a coherent narrative for his research findings, discovered Edward Shills’ mass society critique. Mass society theory, as noted, provided a sociological basis for a fear of powerful media. Shills’ counterarguments, based on people’s inherent social networking needs and behaviors, even in the transition to modernity, gave Lazarsfeld a backdrop with which to highlight his own conclusions about limited persuasive effects.

Deborah Lubken (2008) described the steps in rolling out this vision. Lazarsfeld’s team included young scholars such as Katz, Klapper, and Berelson, each of whom contributed to the creation of the new history. According to Pooley (2006), Berelson, who was particularly influenced by Shills, provided perhaps the first glimpse of that history in 1948. Describing then-current work on public opinion formation, Berelson (1948) observed:

To speak roughly, the 1920s propaganda was considered all-powerful—“it got us into the war”—and thus communication was thought to determine public opinion practically by itself. In the 1930s the Roosevelt campaigns “proved” that newspaper had lost its influence and that a “golden voice” on the radio could sway men in almost any direction. (p. 171)

In 1953, Katz, then a graduate student drafting a BASR report, followed Berelson’s lead and gave this history its new label:

Exaggerating only slightly, we might say that the “model” of the campaign-like persuasion process which research had in mind, at least at the outset, resembled nothing so much as it resembled a giant hypodermic needle. Until very recently, it was widely assumed that the media were all-powerful, capable of reaching out, and influencing nearly every eye and ear. (quoted in Lubken, 2008, p. 22)

The report was an early version of the first chapter of Personal Influence (Katz & Lazarsfeld, 1955), published 2 years later. There, Katz described what he saw as the prevailing image of media effects prior to the 1940s. That view, “first of all, was of an atomistic mass of millions of readers, listeners, and movie goers prepared to receive the Message; and second, (one that) pictured every Message as a direct and powerful stimulus to action which would elicit immediate response” (p. 16).

While a passing reference to “hypodermic stimulus” appeared in Voting (Berelson et al., 1954, p. 234), Katz and Lazarsfeld did not use the label “hypodermic” in Personal Influence (1955). The section of that landmark book that included Katz’s general observations on history, however, was described by Pooley (2006) as “15 pages that shook the field” and “had more influence on the field’s historical self-understanding than anything published before or since” (p. 131.) As noted above, Katz and Lazarsfeld were familiar with suggestion theory. Nonetheless, Elihu Katz especially continued to promote a direct effects research tradition shaped by mass society perspectives (Katz, 1960; Katz & Lazarsfeld, 1955, p. 17).

Klapper (1957-1958), also a Lazarsfeld graduate student and a Project Director at BASR, echoed Katz in a 1957-1958 report. More famously, in The Effects of Mass Communication, Klapper (1960) observed, “The new orientation . . . can perhaps be described, in a confessedly, oversimplified way, as a shift away from the concept of ‘hypodermic effect’ toward an approach which might be called ‘situational’ or ‘functional’” (p. 5).

The idea of an early but erroneous model of media influence characterized by direct, uniform effects was soon taken up by other media scholars (Bauer, 1964; Bauer & Bauer, 1960; Davison, 1959). David Berlo’s, 1960 textbook, The Process of Communication, stated, “Much of the early discussion of the effects of the mass media were of the ‘hypodermic-needle’ variety” (p. 27). DeFleur (1966) was a particularly strong legitimizing force, featuring in the first edition of Theories of Mass Communication an extended review of mass society theory as a basis for an early belief in powerful effects. “We may refer to the first mass communication theory as a ‘mechanistic S-R theory,’” he declared. “It has been given other more colorful names such as ‘hypodermic needle theory,’ ‘transmission belt theory,’ etc.” (p. 115). But while DeFleur, through his many texts, played an important and on-going role in popularizing the hypodermic account, Wilbur Schramm was probably the most powerful voice in spreading and legitimizing the BASR’s pocket history.

Wilbur Schramm and the Magic Bullet

Rogers (1997) declared Schramm to be “the founder of the field” (p. 29, italics in the original), and without question Schramm, more than anyone else, shaped media studies in the 1950s and 1960s, in part through his string of anthologies. His first edited review, Communications in Modern Society in 1948, provided the overture when he declared: “Communications research is a development of the last few years. Fifteen years ago, the term would hardly have been heard” (Schramm, 1948, p. 5.) Acknowledging the early efforts of Lasswell, Lippmann and the Payne Fund researchers, Schramm nonetheless indicated that true social science research did not begin until the work of Lazarsfeld at Columbia and the group at Yale. (Later in life, Schramm said Hovland “was assigned to study a field that never before had been systematically examined” (Schramm, 1997, p. 92). The 1948 anthology also included Hovland’s (1948) chapter on “The psychology of the communication process,” and featured Berelson’s original description of 1920s “all-powerful” propaganda.

Schramm’s (1949b) enthusiasm for the Columbia-Yale paradigm was underscored in his extended 1949 reader, Mass Communications, which was dedicated to Lazarsfeld and included chapters from Lazarsfeld, Berelson, and Katz (as well as important pieces by Park, Lippman, and Lasswell). The next anthology, the first edition of The Process and Effects of Mass Communication, appeared in 1954. It contained a reprint of the 1951-1952 Hovland and Weiss study on source credibility, and a reprint of Klapper’s 1949 BASR report that would serve as the nucleus for the 1960 text. The 1948 Berelson piece was still there, as it was in the subsequent and updated 1960 edition of Mass Communication. This comprehensive reader was designed as an introductory overview of the field and was sparse on treatments of psychological theory. Nevertheless, in his forward, Schramm (1960) highlighted the two-step flow findings of Personal Influence that he said helped bring about the “healthful and general realization that the emphasis on ‘mass’ audience (associated with a direct effects model) was never accurate” (p. viii).

Schramm’s 1963 reader, The Science of Human Communication, offered his elaborated story of the “founding fathers”—Hovland, Lazarsfeld, Lasswell and Lewin—along with minimal effects-oriented chapters from the Columbia and Yale groups and a reinforcing piece by Festinger on cognitive dissonance. In 1971, in the revised edition of The Process and Effects of Mass Communication, Schramm (1971) presented his own label for the BASR direct effects history, coining the term “Bullet Theory”:

Communication was seen as a magic bullet that transferred ideas or feelings or knowledge or motivation almost automatically from one mind to another . . . To sum up, then, in the early days of communication study, the audience was considered relatively passive and defenseless, and communication could shoot something into them, just as an electric circuit could deliver electrons to a light bulb. (pp. 8–9)

Sue Curry Jansen (2008) described the 1971 anthology, as “for many years the best-selling and by far the most influential textbook in the field,” adding, Schramm “virtually defined what would become the dominant paradigm in the field for the next three decades” (p. 83). Arguably, while the Bureau may have created the hypodermic model, it was Schramm who made it professional gospel.

Toward the “Natural History”

While the seal was set on the magic bullet, the multi-stage chronology was not yet fully assembled. In 1966, looking to the future, DeFleur (1966) stated, “We know more about mass communication than we did in 1920; we also know more about it than we did in 1940 or even 1960” (p. 99). And, indeed, by the 1970s, a string of new research initiatives in the social sciences and in cultural studies were expanding the view of media’s relationship with society. Maxwell McCombs and Donald Shaw published their agenda setting study in 1972. Gerbner and Gross (1976) were advancing their cultural indicators project and cultivation analysis. Interest was moving beyond the restricted concern with short-term attitude change, and Elisabeth Noelle-Neumann (1973) encouraged a “return to the concept of powerful mass media.” In 1977, McQuail (1977) unveiled his three-stage “natural history” of effects research and what had been Klapper’s articulation of state-of-the-art social science became for McQuail the “no effects” era (p. 73).

Critical and cultural studies also were gaining academic traction and standing (Curran et al., 1977; Schiller, 1973). As effects research moved into its third era, critical and cultural scholars especially began to question the efficacy, if not the validity, of the Columbia-Yale minimal effects model (McQuail, 1969). That same critical eye was not, however, cast upon the magic bullet. Carey (1978) considered the received direct effects history and, while stripping it “down to the point of burlesque,” conceded it was “correct as far as it goes” (p. 117). He added that there were solid political and economic reasons for observers to assume hypodermic-like powerful effects in the 1930s.

In 1978, the critical sociologist Todd Gitlin dissected the limited effects paradigm of Personal Influence with an unsympathetic scalpel, but accepted the hypodermic narrative as described by the BASR at face value. His larger project in the essay was to illuminate the administrative and marketing-oriented ideology of the positivistic “dominant paradigm,” which he saw as a myopic perspective on broader social and cultural media effects. He offered a detailed analysis of Lazarfeld’s background, the Decatur study, and the two-step flow in particular. He did not treat, in the same detail, the BASR’s hypodermic history, describing it only briefly (pp. 209–210), but noted that the empirical findings from Personal Influence were more “consistent with the old ‘hypodermic’ notion” (p. 219) than they were with the two-step flow. In short, despite keen and exacting critiques of the limited effects model, post–World War II scholars continued to assume that pre-War theorizing was characterized by a direct, uniform model of hypodermic effects.

By the 1980s, therefore, a hypodermic history of early media studies was entrenched and a half century of social psychological investigation of communication influence, along with the primary tool of that investigation, suggestion theory, had vanished, replaced by a sanctioned tale of needles and bullets.