Part VI: Opening

Large file: Patience please.

Insights through intrapersonal communication are found in various contexts of communication research. Each of these scholars is focused on something other than intrapersonal communication processing. Yet, each perceives a relationship to intrapersonal communication processing in her or his area of study. As we expand our perspectives, these articles can help provide an opening to other ideas that will help our field advance.

Television's Native Americans: A Search for Identity (Annette M. Taylor)

Intrapersonal Processing in the Congregation: Recommendations for Enhanced Laity Involvement (Deloris McGee Wanguri)

Intentionality Attributions for Naturally-Occurring Nonverbal Behaviors in Intimate Relationships (Valerie Manusov)

Faces and Voices: Comparing the Impact on Person Identification (Patricia Rockwell)

The Discourse Made Less Commonplace: A Mark of Discrimination (Philippe Castel and Marie-Francoise Lacassagne)

Intrapersonal Health Communication (Gary Meyer and Kim Witte)

Willingness to Communicate, the Need for Cognition and Innovativeness: New Zealand Students and Professionals (Craig Johnson, Bruce Dixon, Michael Z. Hackman, Larry Vinson)

Perceived Understanding, Nonverbal Communication, and Relational Satisfaction (Kori L. Egland, Michael A. Stelzner, Peter A. Andersen, and Brian H. Spitzberg)

The Pedagogical Effects of Interactive Video Instruction in Oral Communication (Michael W. Cronin)

Penetrating the Glass Wall: Creating and Retaining Imagined Interaction in Televised Distance Education (Gregory Gutenko)

Televisionís Native Americans: A Search for Identity

Annette M. Taylor

About the Author: Annette M. Taylor is an assistant professor of Journalism at the University of Dayton, Dayton, Ohio 45469-1410. She teaches writing and publication design. Before entering the academy in 1988, the Detroit native worked many years as a daily newspaper reporter and editor in Montana and Alaska. She also worked as a free-lance writer for publications based in the Pacific Northwest. Besides teaching, she is completing a dissertation in American Culture Studies at Bowling Green State University in Ohio.

Abstract: Since colonists first claimed the New Land, Native Americans have been presented in the popular culture as slowly fading into oblivion. Despite their importance as a source of dramatic tension in televisionís early Westerns, Native Americans usually were relegated to the background, portrayed in such generalized and homogenized ways that became transparent and anonymous. Later television programs, even with contemporary settings, pigeonholed Native people to fabricated tribes and cultures and in terms only relating to the land. The Native American characters on Northern Exposure offer an improved conceptualization of Native Americans, but the series still fails to escape televisionís racist past.

Perception: Watching this award-winning and highly popular series, we might consider what television critics have not:

1. Does Northern Exposureís representations express something new and better or merely cloak and redefine racism?

2. Does Northern Exposure, the one show now on television with Native Americans as regular, featured characters, challenge the stereotypes that have persisted in the popular culture for nearly five centuries?

 

Northern Exposure brought more than visions of Alaska to the rest of the United States when it premiered on CBS television in the summer of 1990. The series restored Native Americans characters to the little screen and became the first television program in more than 15 years to feature them in regular, prominent roles.1 Despite this phenomenon, only a handful among the scores of reviewers who critiqued the show during its summer debut and return the following spring even mentioned Native Americans. Millman (1991, April 8) in the San Francisco Examiner and Kitman (1991, April 20) in TV Guide noted only that Native Americans inhabit the showís fictional Alaskan community of Cicely. OíConnor (1990, July 12) in The New York Times specifically identified the character of Ed Chigliak half-way into his review, describing him only as an "American Indian" with an interest in films and music. A writer for Entertainment Weekly (1990, July 6) also noted Ed, calling him a "crazy Indian" who "proves" to be smarter than expected. Kasindorf (1991, May 27) in New York magazine briefly pointed to the character of Marilyn Whirlwind, describing her as "a round, beatific American Indian woman."

Geranios (1991) in the Boston Herald was one of the few commentators to allude to televisionís past treatment of Native Americans, thus intimating that Northern Exposureís characterizations might be different. Well into his review of the show, Geranios quoted Darren E. Burroughs, the actor who plays Ed Chigliak, as saying: "We try to stay away from stereotypical things, like Tonto or something. I like the way it [the character] was written, and it wasnít with a bad accent or stupid or anything like that." In the next paragraph, Geranios quoted actress Elaine Miles as saying that she was relieved that her character, Marilyn Whirlwind, was portrayed as a "normal person."

And that was that.

Maybe the lack of serious attention to the Native American characterizations on Northern Exposure by most of the media during the showís first runs was a positive sign. Maybe it suggested acceptance of the race and cultures that the United States government was still trying to annihilate less than sixty years ago. Maybe it demonstrated successful integration of Native Americans and native culture with the dominate European-American society. If Cicelyís multicultural community merely reflected real life, then critics would have little cause to expatiate upon it in their reviews. More likely, however, television critics did not notice the Native American characters or found them trifling and their cultural presentations nebulous. Critics simply failed to recognize the significance and potentialities of such characters and their placement in the show and on prime-time television.

Why should we be surprised? Ignoring Native Americans and dismissing their rich and distinct cultures have been habits in this country since the colonists first staked their claim to the New Land. At least until the Indian Reorganization Act of 1934, European Americans in government and business championed policies of genocideñmurder, assault, and neglectñto make the natives disappear. Essayists, novelists, journalists, filmmakers, and television producers essentially accomplished the job by presenting Native Americans in fiction and non-fiction as perpetually fading into oblivion (Berkhoffer, 1978; Mander, 1991, November/December). For all the war whoops, tomahawk chops, and assaults on wagon trains in literature, film, and newspaper columns, Native Americans usually have been cast only as anonymous extras or secondary figures primed for defeat. For all the psychological drama in modern-day culture clashes, Native Americans still too often have been portrayed in such generalized and homogenized ways that they became illusive.

To most everyone else, Native Americans have been simply "Indians," the European misnomer for the thousands of cultural groups that once populated the Americas. From that mistaken tag was created a whole new identity enveloped in buckskin and feathers, mysticism and Mother Earth. "Indians" became monosyllabic heathens, noble but doomed leaders, drunken and pathetic fools, or trusty servants. While the popular culture periodically has presented both historical and modern-day Native Americans as complex individuals with distinct heritages, the term, "Indian," in many minds remains inextricable from the fabrications. To many people, "Indians" still are defined by what they are notñwhite European-Americans. As Berkhoffer (1978) pointed out, for "Indians" to be considered like everybody else, they must reject their culture. For "Indians" to be faithful to their heritage and traditions, they cannot, by definition, be part of the larger society dominated by European-Americans. Thus segregated, they face economic disaster and, consequently, illness and death. In any case, "Indians" face but one end: obliteration.

While Northern Exposure has revived Native Americans from virtual extinction on television (Gerson, 1980; Greenberg & Heeter, 1983), the mere presence of such characters does not necessarily alter the subtextual meanings that the media has historically associated with natives and native culture.

 

Popular Culture Images of

Native Americans

For many Americansñat least those of us born and reared east of the Rocky MountainsñNative Americans have existed only on television and film and usually within the confines of the nineteenth century. Education through books, schools, documentaries, and first-hand witnessing may alter the focus but rarely erases the image we were exposed to when young. Such is the power of myth (Barthes, 1972). In the late 1950s and early 1960s, television offered us as many as 28 Westerns per season, including The Lone Ranger, Cheyenne, Wagon Train, and Gunsmoke. Sunday afternoon movies on the little screen replayed such Hollywood classics as They Died with Their Boots On, Stagecoach, and Northwest Passage. At the cinema, we watched The Searchers, Cheyenne Autumn, and How the West Was Won. Memory tells us that all of those black-and-white and Technicolor Westerns detailed the lives of cowboys and Indians, albeit through fictionalized storylines.

In actuality, most Westerns focused on the cavalry, honorable gunslingers, father figures, and rich settlers (Himmelstein, 1984; Parks, 1982). Except for a battle or two with the "savages," the new Americans spent most of their time braving the elements or chasing cattle rustlers and bank robbers. The heroesí favorite steeds often got more screen time than any Native American (Brauer & Brauer, 1975; Tompkins, 1992). Hollywoodís "Indians" loomed over the horizon and provided the dramatic tension but rarely appeared as natural, sharply honed personalities with whom audiences could readily relate, let alone identify. Even as sidekicks, Native Americans were marginalized. The most famous fictional Indian of all, Tonto, always remained to the side of the Lone Ranger and rarely said anything. Daniel Booneís Indian companion, Mingo, was little more than a stooge despite his Oxford education.

Hollywoodís "Indians," when they did appear, generally fell into one of the following roles. They were bad Indians, who scalped, raped, and burned. They were good Indians who adopted white manís ways, served the white man, and helped him to clear the Plains of other Indians. They were noble but doomed Indiansñheroic, virtuous, and living according to their own cultural code of conduct. They had to be killed because there was never enough room for them and the new Americans on the same piece of property. There were saintly Indians who clung to their traditions but were too spiritual, too serene, and usually too old to pose a threat. And besides, they usually were safely confined to the reservation or a designated zone near the fort. There were defeated and pathetic Indians who needed protection. If they were to be saved, the new Americans were the only ones to do it.

When Hollywood tried sympathetic portraits and brought Native American figures to the storyís forefrontñmost notably during the social-consciousness-raising period of the late 1960s and early 1970sñit tended to merely reverse the stereotypes. It made European-Americans the bad guys, turned pro-progress tales into anti-progress tales or "anti-Westerns," and continued to distort historical evidence (Bataille and Silet, 1980a; Tuska, 1985).

Television rarely attempted any image adjustment and only a few times did it pull Native Americans from the subplot. Throughout the history of television, Native Americans have been featured as the lead characters in only about a half-dozen series, four of which confined them to the past (Brooks & Marsh, 1992; McDonald, 1987; Terrace, 1976a and 1976b). The first series, Brave Eagle, was a bold attempt to suggest the historical mistreatment of Native Americans. The underlying message, however, was doom. The program, which ran one season during 1955-56 on CBS, concerned a Cheyenne chief trying to safeguard his homeland from the oncoming settlers. Episodes related hardships of native people as perceived by the showís European-American creators and portrayed through the characters of a young chief and his foster son, Keena. Other characters included a "Sioux maiden turned Cheyenne," called Morning Star, and a "half breed" called Smokey Joe (Terrace, 1976a, p. 116).

Hawkeye and the Last of the Mohicans ran in syndication the following season in 1957-58 and featured characters created by James Fenimore Cooper, not necessarily the most credible source for accurate and complex representations. Set in the 1750s in New York State, the series related the escapades of Nat "Hawkeye" Cutler and his Indian blood brother, Chingachgook. These fur traders and frontier scouts helped the pioneers in settling and helped the army "in its battle against the constant Huron uprisings" (Terrace, 1976a, pp. 342-343). The showís characterizations departed not at all from the stereotypical roles of nineteenth-century literature and Hollywood films.

Cheyenne, an ABC series from 1957-63, featured actor Clint Walker as frontier scout Cheyenne Bodie, a man of Cheyenne descent and trained in the ways of both worldsñNative American and European-American. A drifter, the character might be found one week working as a scout for a wagon train or another week as part of a ranch crew. Like other Native American representations, this one was familiar to majority audiences: Cheyenne served a role, serving the new Americans. Stories had little to do with Native American culture, except for what the scriptwriters apparently believed was an "Indian" trait: Cheyenneís ability to track.

Law of the Plainsman, which ran on NBC from 1959-62, offered the most fantastic premise. A 14-year-old brave in Arizona is about to scalp a wounded army captain, but inexplicably relents and nurses the soldier back to health. The captain takes a liking to the boy and gives him the name, Sam Buckhart. Audiences apparently were supposed to believe that the boy had been nameless. Two years later, the captain dies and leaves his estate to the boy. Sam uses his newfound wealth to go to Harvard, which, amazingly for this time period, accepts him. Sam becomes a U. S. Marshal and moves to New Mexico. The weekly episodes related Samís capture of criminals and his efforts to establish peace between his native people and the new settlers (Terrace, 1976b).

It took Burt Reynolds, himself of Native American descent, to bring a modern-day Native American to the small screen in a lead role. The 1966 ABC series, Hawk, featured a half-Iroquois New York City police lieutenant, who was "a prowler of the night" for the district attorneyís office (Terrace, 1976b, p. 342). The series did not deal with Native American issues or identity unless one believes that Hawkís talent to see in the dark was endemic of native people. The series lasted only three months.

Seven years later, television tried another Native American character in a lead role. Launched during the 1974-75 season on ABC and pulled after only three months, Nakia featured a modern-day Navajo working as a sheriffís deputy in a rural New Mexico county. The basis of the series was that Deputy Nakia Parker sometimes finds his loyalties divided between his beliefs and tribal customs and the modern police practices and the law he was sworn to uphold (Terrace, 1976b). Whatever cultural conflicts he had, they were relegated to subplot. The series was a traditional crime drama with murders, kidnappings, and thefts (TV Guide, 1974, September-November). Only one episode emphasized Nakiaís "Indianness." He used his knowledge of the landñapparently considered by the writers as an Indian traitñto track a kidnapped boy. Another allegedly "Indian" characteristic was Nakiaís preferred mode of transportation: a horse.

These all-too-familiar images can be traced back to the early sixteenth century when Europeans not only began dominating the aboriginal people of the Americas but also information about them that was conveyed to the rest of the world. The Europeans and, later, European-Americans shaped images that embraced a dichotomy between savagism and civilization, which ultimately served to justify dispossession and extermination of Native Americans (Pearce, 1953). Writers, most notably James Fenimore Cooper, capitalized on the tensions, establishing a new American literature that celebrated that moment of confrontation "when social order and anarchy meet" (Berkhoffer, 1978, p. 97). As important as Native Americans were to the story, however, they had to remain shadowy figures. Cawelti (1971) writes: "[I]f the Indian represented a significant way of life rather than a declining savagery, it would be far more difficult to resolve the story with a reaffirmation of the values of modern society" (p. 38). In other words, to understand the natives as distinct human beings would make it harder in real life to kill them and steal their land.

The popular media confined Native Americans to a handful of tribes and cultures and then redrafted them to suit popular conceptions and political agendas. Alfonso Ortiz (1977, April) found that all Native American imagery comes from a mere five sources: the Tupinamba of coastal Brazil, the Algonquins of the United States coast, the Iroquois, the Cherokee and five civilized tribes, and the Sioux, the most dominant and long-lasting image. The core of this image, according to Scholer (1981), is that of befeathered, breach-cloth-clad horseriders on the open plains, living in tepees, hunting buffalo. With Hollywood perpetuating and reinforcing the image, modern-day audiences got the idea that Native Americans belonged only to Plains tribes, spoke the same language, dressed in the same clothes, and practiced the same religion.

Despite the 1960s counterculture interest in Native American spirituality, philosophy, and artistry and the pro-Indian movies in the early 1970s, Native American characters remained tied to the land and bound by images born 500 years ago. They were the children of a benign nature or, most often, beasts of a the dangerous wilderness. The underlying messages, whether the setting was modern-day or turn-of-the-century, tended to remain the same: the noble savage was always doomed and the assimilated Indian always abandoned her people.

According to a survey of minority representation on television by Mauricio Gerson (1980), 23 percent of all Native American characters on television during the 1970s were involved in fatal violent actions. No other ethnic group was shown so frequently involved in violence. Compared to European-American characters, Native Americans were three times more likely to be involved in violence. When looking at TVís violent villains, Gerson found that Native Americans again led all ethnic groups. Looking at victims, Gerson found Native American characters were the least likely to be the innocent party.

Even when they were portrayed as nice, average men and women, Native Americans were nondescript. According to Gersonís survey, only 25 percent of the Native American characters on television in the 1970s had an identifiable occupation, the lowest of all ethnic groups. By comparison, occupations of whites could be identified 82 percent of the time. Of those natives with occupations, less than 3 percent were portrayed as "professional."

Such distinctions would be difficult to notice, however, since so few Native American characters appear on television. When the popularity of Westerns waned in the early 1970s, so did the Native American presence on television. By the early 1970s, they represented a mere 0.3 percent of all TV characters (Gerson, 1980). Their representation since then has hardly been noted.2 For todayís youth, who have all but escaped exposure to the Hollywood parade of Apaches and Comanches, the principle image that remains, however faded, is that of war-painted warriors attacking innocent pilgrims. Native Americans otherwise are virtually invisible.

Popular Culture Images of Alaska Natives

Conspicuous by their near absence on film and television are Alaska Natives. Alaskaís history would seem to have all the elements for a good "Western" because of its indigenous peoples, outside invaders, and the conflict over the rich resourcesñotters, fur seals, whales, fish, gold, timber, and oil. The three major gold rushes to Alaska and the neighboring Canadian Yukon in the late 1800s and early 1900s especially provide all the drama of the Wild West stories set in Dodge City or Tombstone, as Jack Londonís stories demonstrate. But the gold rushes were stories of European-Americans, not Alaska Natives.

The most likely reason that Alaska Natives escaped the Hollywood treatmentñsome would say "mistreatment"ñis that much of the history of white contact lacks combat. Except for a few battles during the Russian occupation days, Alaska Natives never warred with the invaders and never posed a major threat. The Aleuts of the Alaska Peninsula and Aleutian Islands were quickly subdued by the Russians, enslaved, and nearly exterminated. After purchasing Alaska from the Russians in 1867, the United States government continued the abuse of the Aleuts until circumstances during World War II exposed government policy and forced some changes. Yet, in one of the most egregious examples of tyranny, the government kept the Pribilof Aleuts in a state of virtual servitude until 1984.3

The Tlingits in Southeast Alaska won a few battles early on against the Russians, but eventually they were forced to share Alaskaís southern panhandle. The Tlingits nevertheless escaped Russian dominance and maintained their way of life and distinctive culture by keeping the Russians and Americans otherwise occupied: playing them against each other in the trading business. The mainland, occupied by the Inupiaat Eskimos of the north and northwest, the Yupíik Eskimos of the west, and the Athabaskan Indians of the interior and upper Cook Inlet, was simply too cold, too bleak, and too dark to appeal to many non-natives. Those who did decide to immigrate found plenty of room.

Not until 1971, after thwarting construction of the 800-mile-long trans-Alaska pipeline that the federal government and oil industry desperately wanted, did the Alaska Natives sign a treaty and land settlement claim with the United States government. Alaska Natives were awarded 40 million acres, additional title to lands of special religious and cultural significance, and $1 billion. Throughout their history of contact with outsiders, Alaska Natives have managed to retain their traditional ways well into the modern era. Today many continue a subsistence lifestyle while driving three-wheelers and snowmachines past satellite dishes on the frozen tundra.

The image many Americans have of Alaska Natives comes primarily from film documentaries, such as Nanook of the North (1922), Eskimo (1934), The Alaskan Eskimo (1953), and Disney nature films, along with a handful of feature films, such as The White Dawn (1974) and White Fang (1991). Television has had virtually no input in the popular conceptions of Alaska Natives. In the history of television, only four shows, including Northern Exposure, have been situated in Alaska.

The first Alaska-based TV series, The Alaskans, ran on ABC during the 1959-60 season and told the story of Silky Harris, Reno McKee, and saloon-hall gal, Rocky Shaw. The trio goes to Alaska in 1898 in search of a gold mine that a love-struck prospector had left to Rocky. An avalanche buries the gold, and they head to Eagle City to gently fleece others of their gold (Terrace, 1976a). The following year, NBC tried the same theme with Klondike, which related the struggles of Kathy OíHara as she tries to maintain an honest hotel in a lawless town. She gets a bit of help from prospector Mick Holliday, who also serves as the love interest. The show lasted only four months. None of the material written about these two shows in encyclopedias or TV Guide mentions Alaska Natives or native culture. Other than the weather, then, the shows probably could have been situated anywhere in the West in the contiguous states.

ABC tried a modern-day crime drama set in Alaska in 1974 with Clint Walker as Cal "Kodiak" McKay, an Alaska State Police Patrolman. Alaska Natives evidently had a presence in Kodiak in that the lead character got his nickname from them. In addition, Kodiakís assistant was Native with the Native-sounding surname of Imhook. He was Anglicized, however, with the given name of "Abraham Lincoln." According to the synopses of episodes published in TV Guide (1974, September-October), the series focused more on the sceneryñit was filmed in Alaskañthan on Alaska Natives or native issues. The show was canceled after only four episodes.

In Northern Exposure, then, American television offers its first truly successful series to prominently feature Native American characters and, like Hawk and Nakia, to position them within the modern society as normal, functioning human beings. But where Hawk tried to assimilate its Native American into the dominant culture and where Nakia stressed an irreconcilable conflict between European-American and Native-American cultures, Northern Exposure attempts to portray the native culture as complementary to the dominant European-American culture. The Native Americans on Northern Exposure are not freaks, outsiders, second-class citizens, villains, or sidekicks. And most unusual for TV land, the majority culture here is better for the influence of the native culture.

 

Northern Exposure and

The Western Genre

Northern Exposure is based on the premise of a young New York Jewish doctor being forced to live in Alaska to work off his medical school loans, which were financed by the state. Dr. Joel Fleischman had hoped to serve his time in Anchorage, but instead finds himself in Cicely, a village of 839 persons. Weekly episodes relate the experiences of this foreigner and the other, rooted inhabitants of the frozen backcountry. Many of Cicelyís residents are Native American, as suggested by the crowd scenes, but seven of the nine leading characters represent people of Judeo-Christian backgrounds from predominately European-American or European-Canadian communities.

Just like the classic television or film Western, Northern Exposure has civilization meeting primitivism, social structure encountering anarchy, the gentry confronting the bushmen in the frontier. Fleischman, the well-bred and educated urbane citizen of Americaís quintessential settlement, enters the nascent town of Cicely in the middle of a wilderness. If this were 1950s television, a superior Fleischman would force an inferior, indigenous population of Cicely to adapt to his ways or get out of the way. He would clearcut a pathway to progress for the grateful settlers by subjugating the natives. He would be a Marcus Welby in The Big Valley, the ultimate "Frontier Doctor," nursing the spiritually deprived to moral health.4 Fleischman would be the hero.

If Northern Exposure were a creature of early television, the characters of Maurice Minnifield, the townís richest landowner and owner of all communication systems; Holling Vincouer, owner of The Brick restaurant or bar; and Chris "In the Morning" Stevens, KBHR-radio disc jockey and ordained-by-mail minister, would form the necessary male community like the team of men on Wagon Train who always supported wagonmaster Major Seth Adams. Maurice, Holling, and Chris would help Fleischman battle the elements, outlaws, and, of course, the Indians. As the crew on Rawhide looked up to trail boss Gil Favor, they would look up to Fleischman.

In keeping with Brauer and Brauerís (1975) definition of classic Western series, Northern Exposureís girlsñand they would be girls rather than women, no matter what their age or occupationñwould be adoring, submissive, flighty, fidgety, or meddling. The character of bush pilot Maggie OíConnell would be the schoolmarm with just enough intelligence, independence, and sexuality to suggest Gunsmokeís Miss Kitty. Maggieís role would be to woo and win or forever pine for the showís star. The Shelly Tambo character, Hollingís wife, would be the saloon-hall entertainer with the heart of gold and unwavering loyalty to her man. Storekeeper Ruth-Anne would be the meddling female, who always has something to say.

The Alaska Natives on Northern Exposure are easy to categorize according to the James Fenimore Cooper formula. Ed Chigliak, the teenager and film buff, would be the "crazy Indian." We would expect him to be slower and less intelligent than the non-natives, just as the writer for Entertainment Weekly (1990, July 6) seemed to do. Ed would be like Gunsmokeís Festus and would lend comic relief to the busy, burdened hero. Marilyn Whirlwind, Dr. Fleischmanís office manager and receptionist, would be the lovely and loyal squaw, a female Tonto to Fleischmanís Lone Ranger. She would speak wisely, but in clipped phrases. Having adopted the white manís ways, both Ed and Marilyn would be considered "civilized" and, as such, would be tolerated by the majority European-American community. The status of the other Native Americans shopping in the stores and walking along the street would depend on how well they behaved. All those "Indians" in The Brick would have to be closely watched, however, since everyone knows that liquor and Indians are a lethal combination.

The similarities in characterizations between the old Westerns and Northern Exposure are obvious, as the May 18, 1992, episode on Cicelyís founding, "Cicely and Roslyn," revealed. Indeed, some elements of the old Western genre bleed into Northern Exposure. There is a bit of the hero in Fleischman, the boyís club mentality in the other men, the ornamental in the women, the dim-witted in Ed, and the fidelity in Marilyn. But just as the real Alaska is a place where people can and do re-invent themselves, so Northern Exposure is a show that re-invents Western-genre relationships. The most significant alteration concerns the Native American characters.

Marilyn Whirlwind is nobodyís Tonto. She scrubs the examining room, holds her tongue, and maintains tranquillity, but the audience knows she is in charge, with knowledge and power to make things happen. In every conflict of wills or ways between Fleischman and Marilyn, Marilyn wins. Even in the pilot episode, Marilyn clearly comes out ahead. She needs a job, decides she wants to be the doctorís assistant, and, despite Joelís protests, goes to work. She lets Fleischman know that he will be making house calls and gives him maps along with patientsí charts. She lets Fleischman know that he will fly into bush villages to conduct childbirth classes, and he goes. By the 27th episode, in the middle of the 1991-92 season, Fleischman admits that he depends on Marilyn, trusts her judgment, and essentially considers her a partner.

Marilynís influence extends to the other townspeople, as well. When Maurice tries to launch a multimillion-dollar ostrich farm ventureñwith Marilynís ostrichesñand take most of the profit, Marilyn negotiates the better deal. When Holling goes off on a bender, moping about his insignificant life, she is the only one in town who can get him to return home. She goes to Hollingís mountain still, pokes her head inside the cabin, and announces: "Itís time to go home." And Holling does.

Marilyn can be annoyingly stereotypical when she slips into the role of wise, saintly chief, possessing "the serene wisdom of a tribe of elders," as Kloer (1991, August 5, p. D-1) wrote in the Atlanta Constitution. When descendants of Napoleon and an Alaska Native woman show up in Cicely to claim the frozen body of the French leaderís faithful companionña body that townspeople are considering putting on displayñthe television audience knows that Marilyn is behind the repatriation. The point is driven home at the end of the episode when Marilyn stands amongst the trees, watching the short Natives in their black leather jackets paddle their canoe into the mist. It is a scene right out of Cooper. Any tendency towards stereotyping her as Earth Mother is muted, however, by the fact that Marilyn cannot exclusively claim wisdom and righteousness. Store-keeper Ruth-Anne, as well as DJ Chris Stevens, also share such gifts.

Marilynís clipped phrases also are too reminiscent of Tontoís monosyllables to escape the negative comparison. Her limited speech may have more to do with the inexperience of the actress than the intentions of the showís writers. Although the character is a woman of few words, Marilyn is not stoic like Tonto. Marilyn registers emotion, intelligence, and superiority. She was noticeably angry when Holling, during the mayoral election, implied that all native people voted alike. She was noticeably in love with the silent Flying Man, who visited Cicely when the circus bus broke down. She was obviously sad to leave him. She did not want to be "just the Flying Manís girlfriend," traveling on the bus with no life or identity of her own. "I have a career here," she told him. She has been manipulative and calculating and has outwitted both Fleischman and Maurice.

The audience has seen Marilyn as a normal and average woman, who loves her family, eats at a dinner table, goes to work, and tries to be a decent human being. Marilyn adheres to cultural traditions, just as Christians celebrate Christmas and Jews mark Hanukkah. During the annual Raven pageant in a 1991 December episode, Marilyn related the real-life Alaska Native myth about how the Raven brought light to the world. The portrayal displayed all the respect that would appear in a scene of a Christian relating the story of the birth or resurrection of Jesus.

Although he may talk like a space cadet, the character of Ed Chigliak is nobodyís fool. He tends to be a half-beat behind, but that is due more to his age and innocence than his IQ, which is a high 180 (although it seems to have declined by the 1992-93 season). Realizations "hit" Ed like a tap on the shoulder or an unexpected kiss on the cheek. The actor who plays Ed sees the character as someone who lives in the moment, "each day noticing the sun come up with a childlike innocence," as Kloer (1991, August 5, p. D-1) noted. Ed looks at the moon and sees a bunny rabbit. Ed, too, defies the stereotypes, not only because of his intelligence, but because he shares so many traits with Shelly Tambo, the ditsy yet smart blond with the child-like innocence.

Ed, an orphan, has a strong affinity for his Native American culture and philosophy. He pays attention to spirits, holds a deep respect for his elders, and abides by tradition. He also appreciates his American citizenship, takes his voting rights seriously, and studies American history in preparation for his first time to vote. A child of the later twentieth century, he is a media junkie. He knows everything and anything to do with filmmaking, studies Igmar Bergmanís camera angles, and takes advice from Woody Allenís mother. From his small rented room, he corresponds with the countryís leading filmmakers, including Martin Scorsese and Steven Spielberg. His all-consuming passion is the very medium that has so distorted the image of Native Americans. Ed, however, is no relic but is of a new generation with a promising future.

Northern Exposure mixes up the other traditional Western roles, as well. Mauriceñthe Ben Cartwright of the showñassumes the role of buffoon, but one who has a knack for gourmet cooking, a love of show tunes, and a sincere yearning for love. Chris assumes the role of spiritual guide, waxing eloquent Kahlil Gibran, Buddhism, Zen, and all the great philosophers and poets. But this holy man is an Alaskan Adonis, ex-con, and petty thief with an African-American, half-brother, Bernard. Holling, the old adventurer and shootist, holsters his weapons after dreaming that all the animals he had killed were gunning for him. The Euro-American women are the least removed from their Western predecessors in that they maintain a definitional connection to men, but at least they are self-reliant and assertive. Rather than the Native Americans, the "hero" is the antithesis of the classic Westernís true American: Fleischman is Jewish and of Russian descent. He claims to represent progressivism and civilization, but we soon see that his vision is narrow and his behavior rude. Fleischman is the outsider, the "other," the one who must adapt.

The West As A Multicultural Community

The settlers who came to Cicely before Fleischman seeking tolerance and acceptance, learned to practice them, as well. And so must Fleischman. According to the old Western genre, talk of a 250-year-old spirit, a dead man returning to earth as a dog, and a white man becoming a "person of color" through the power of dreams would all be treated as bizarre, cultist "Indian" things to be stamped out like the misrepresented Ghost Dancers were on an old episode of High Chapparel.5 Instead, in Northern Exposure, when Ed takes his ancestral spirit, One Who Waits, to The Brick for a hamburger, the crew at the bar accepts Ed at his word that the spirit is sitting there on a bar stool. Sure it is possible. Why not?

When Maggie proclaims that a visiting malamute is her dead boyfriend Rick returned, Shelly, Holling, and Chris note how much the dog acts like Rick. Everyone is open to the possibilities except Fleischman, who assumes only his beliefs are valid. Like any Western hero, he points out the errors of these ignorant and silly people. In Cicely, however, nobody listens, and Fleischman, with some redeeming value, spies a crack in his limited world view. He asks two local Alaska Native men about reincarnation, interrupting their conversation about Japanese exchange rates, real estate investments, and the stock market. They respond that while reincarnation in various forms is well known, documentation about people coming back as dogs is inconclusive. They find no contradiction in participating in the modern-day society and holding age-old beliefs. Fleischman ponders the possibilities but cannot quite dissolve the notion of a dichotomy between "Indians" and everyone else, between wilderness and civilization, between tradition and science.

While traditional healer Leonard Quinhagak respects modern medical techniques and seeks to learn by observing Fleischman at work, the doctor dismisses Leonardís practices as quackery. In one episode, Fleischman diagnoses Shellyís rash as an allergic reaction but cannot identify the cause. He sends Shelly home with an ointment, which fails to work. Leonard talks to Shelly about her childhood and then suggests that she is shedding her skin to be reborn. Shelly finds the explanation quite reasonable and returns to The Brick to molt. Fleischman is livid about Leonardís interference but eventually is forced to admit his own limitations. If he could not be certain of the accuracy of his diagnoses, then he cannot be certain of the error of Leonardís. The episode ends with new, pink skin for Shelly and a lesson for Fleischman.

Northern Exposure, unlike its Western-genre predecessors, takes Native Americans seriously and views native culture as a viable alternative and beneficial companion to European-based cultures in the United States and Canada. The old Westerns rarely gave such consideration to native culture because, some scholars contend (Bataille & Silet, 1980b; Brauer & Brauer, 1975; Woll & Miller, 1987), audiences would have had to confront the possibility that those assaults on the wagon trains were justifiable responses to the invasion by outsiders. To find Native Americansí way of life plausible and their actions reasonable would have provoked questions challenging the foundation of Western society beliefs about manifest destiny, progress, civilization, and morality. And so the popular culture, spurred by moneyed and political interests, dismissed Native Americans as simply too different from the rest of society to understand.

Such ignorance can foment fear, as Fleischman discovered in one episode. Raised on Hollywoodís images of Native Americans, Fleischman resists a patientís invitation to adopt him into her tribe in gratitude for easing her arthritis. He says he has his own tribe and cannot see any wedding of the two. Then, reminded of stories about torture and mutilation in initiation ceremonies, he becomes terrified. When Marilyn threatens never to speak to him again if he fails to go through the adoption, he relents. The process begins with his "purification," orchestrated by Marilyn, who ultimately teaches Fleischman the value of seeking knowledge. First, Fleischman must give away all his worldly goods. Pained by the realization of his own materialism, Fleischman nevertheless explodes in anger over each disappearing possession. Then he must fast, and he gets hungry and cranky. Then he must search for a vision, but he does not believe in visions and becomes frustrated. All the while, he fears the upcoming initiation ceremony. At each step, Fleischman moans and complains, and Marilyn says nothing. Only when he finally asks when his present trial will end does Marilyn end it and move him to the next step.

Later, at the community center, Fleischmanís grateful patient speaks a few minutes about why she wants the doctor as a grandson. Tribal leaders give Fleischman a plaque and shake his hand. Fleischman stands awkwardly at the podium. His fears had been his own making. He offers his thanks, and then everyone sits to eat a potluck, thanksgiving meal. Fleischman learns, as he does periodically throughout the series, that all peopleñdespite differences in rites, religions, lifestyle, and languageñneed other people.

More than anything else, the series focuses on family, home, and community. Yet, while every non-native character has an identifiable family and cultural and/or religious background, the two native characters have been stripped of their heritage. Although Northern Exposure attempts to celebrate both the similarities and differences of all peoples and cultures, it has simply identified Ed and Marilyn as anonymous "Indians." The absence of clearly defined and specified countervailing imagery has put Ed and Marilynís ancestors in buckskins and feathered war bonnets, which in real life are associated with only a few tribes in the lower 48 states and no Alaska Native group.

Although "Indian" is an accepted anthropological term for one of the three major ethno-linguistic groups of Alaska Nativesñthe others being Aleut and Eskimoñit is not a term generally used by Alaskans. Most Alaskans refer to Indians by their different subgroupsñAthabaskans, Tlingits, Haidas, Tsimshian, and Eyak. While most all Alaska Native groups are represented in Alaskaís urban areas, specific groups tend to dominate populations of small rural villages. Athabaskans historically have lived in the interior of Alaska, and the Tlingits, Haidas, Tsimshians, and Eyak call the archipelago of Southeastern Alaska and the Cordova area home (Lantis, 1970; Laughlin, 1980; Morgan, 1979; Vanstone, 1974). Eskimos have made their homes along the coastal regions of Alaskañthe Inupiat in the north and northwest, the Yupíik in the southwest, and the Koniag along the Alaska Gulf coast. As far back as 12,000 years, the Aleuts have lived on the Alaska Peninsula, the long archipelago of more than 100 islands that stretches westward into the Bering Sea, and the two Pribilof Islands to the north.

Alaska Native groups developed diverse cultures suited to the many environments and available resources in the state. The Athabaskan and Eskimo cultures are rooted in a system that promotes independence. Athabaskans, whose cousins are the Apache and Navaho, and the Eskimos in years past lived in small family groupings and relied on the family head for food and direction. Athabaskans, more so than the Eskimos, were highly mobile, with no permanent villages, because of limited resources and need to search for food. Aleuts and Southeast Alaska Natives, who had the vast resources of the oceans at their front door, were much more sedentary and socially and community conscious. The Tlingits and Haidas, in particular, developed complex social structures and elaborate ceremonies and concentrated on amassing great wealth.

The art tradition of the Aleuts continues in fine grass baskets and ivory carvings. In Eskimo communities, men and women sculpt wood, bone, and ivory and make fine fur garments and baskets. Athabaskans, until recent generations, generally lacked much interest in art because the lifestyle afforded little opportunity. Tlingits, on the other hand, are renown for their art, especially the highly refined styles of form lines of animals representing clan crests or stories. This work can be seen carved on totems, screens, bentwood boxes, and silver, and can show up in non-traditional ways in watercolors and T-shirts. Tlingits also are noted for the Chilkat blankets, which originated with the Tsimshians and are made of mountain goat wool and cedar bark (Campbell, 1993).

Despite these major distinctions among real Alaska Natives, Northern Exposure confines its Native Americans, for the most part, to a generic group of "Indians." Marilyn comes simply from "Marilynís tribe." Until the end of the 1992-1993 season when he was briefly identified as a Tlinglit, Ed came only from "Edís tribe." By contrast, the showís writers from day one have identified Fleischman as Jewish from New York. Maggie, despite her Irish name, is a rich Protestant from Grosse Pointe, Michigan. Shelly is poor Catholic from Saskatoon, Saskatchewan, Canada. Chris of Wheeling, West Virginia, came from a broken home, served time in a state penitentiary, and has adopted an eclectic blend of religious and mythical beliefs. Holling, of royal French ancestry, grew up on a farm in Quebec, and Maurice of Oklahoma is primarily a child of the military, having enlisted at the age of sixteen.

Yet without saying so, the program increasingly has appropriated the Tlingit culture for Cicely. The town features totems, carved ravens, and T-shirts with the black-form lines distinctive of the Southeast Alaska Native culture, even though all locator references since the series premiere have placed the village in the Alaska Interior, home to Athabaskans. To add to the confusion, Marilyn Whirlwind early in the series wore a Cayuse-Nez Perce costume during a powwow attended by six-feet-tall Crowsñnovice actors from Montana. Marilyn then performed a dance endemic to the Washington-Idaho area Native Americans. The implication to the viewing audience was that the dress and dance were Alaskan when, in fact, they were not.

Then, at the end of 1991, Marilyn suddenly and silently adopted the ways of Chilkat Tlingits. During the community Christmas pageant, she wore a Chilkat robe and danced under an archway of classic Tlingit design. She told the Raven myth, a common Alaska Native story of the creation but then missed the Tlingit distinction of a second raven. Marilyn thus has floated in a tribe of "any Indian," which puts the character at risk of slipping into old stereotypes. By the end of the 1993-94 seriesñfive years into the development of the programñMarilynís cultural heritage remained nameless.

Long without any moorings, Ed already has lapsed into the manufactured "Indian"ña sidekick in some episodes and a New Age spiritual wonder in others. More and more often, he has served as Fleischmanís or Mauriceís servant, even though his tendency to wander makes him a bit unreliable and obviously less obedient than Tonto or Mingo. His new-found cultural identity at the end the 1992-93 season held little promise, however, for a new and improved popular conceptualization of Native Americans. Edís heritage came up only in context of its "dying" language. Ed decides to try preserving the language by convincing an elder to dub an old movie in Tlingit. Once again, Hollywood scriptwriters dim the lights on Native Americans.

An "Indian" on television today without a tribe, language, religion, culture, or occupation is hardly indistinguishable from the homogenized Indians of the old TV Westerns. In TV land, Marilyn might just as well be a Cheyenne in Navajo dress practicing the rights of the Hopis. What native group will she adopt next? Almost assuredly, however, the programís writers would never so readily alter Fleischmanís cultural heritage. He would never inexplicably become a Lutheran from Minnesota.

It is possible that Northern Exposureís producers and writers believe they are rescuing the term, "Indian," from its racist trappings. Northern Exposureís "Indians" indeed counter those in the old Westerns. The effect, however, is to subsume Alaska Natives into tribes of the lower 48 states and gloss them with Hollywoodís manufactured stereotypes, which they thus far had managed to escape.

Some critics contend that the media, in general, avoid dealing with specifics about real-life Native Americans and Native American issues because of the national guilt over a past government policy of genocide and the current legacy of social and economic problems on the nationís reservations (Gitlin, 1983). Guilt makes people feel bad and "it is not considered good programming to make your audience feel bad," a network executive reportedly once said (Mander, 1991, November/December). People who feel badly do not feel like spending money. And the purpose of television, after all, is to sell products.

Certainly the popularity of recent made-for-TV movies, such as Son of Morning Star, and Hollywood films, most notably Dances with Wolves and Last of the Mohicans which strove to be authentic at least in terms of language and dress, have opened the door to more realistic portrayals of Native Americans as distinct groups and cultures. These shows found acceptance, however, because they focused on the past and, in some respects, perpetuated the noble savage image, what some critics are now calling "the new Custerism" (Sarf, 1991, November/December; Seals, 1991, May 13). Mander (1991) argues that European-Americans do not know how to relate to Native Americans other than as relics. They do not see "Indians" who cut their hair and live in the city as "real" (Mander, 1991; Teegardin, 1990, September 6). Americans want to continue to think of Native Americans in romantic terms, even when they carry a briefcase (Martz, 1986, December 7).

The image of Indian as relic, even in modern-day society, has become so fixed in the American psyche that even United States Supreme Court justices, with all of their education, cannot erase it from their own minds. Justice Blackmun, with Justices Brennan and Marshall, in a 1989 tax case accused their brethren of betraying a "stereotyped and almost patronizing view of Indians." The case involved land development and the majority of justices decided that only "pristine" land could be "spiritually significant" to a tribe. The dissenting justices argued that Native Americans were being held to a "single, perhaps quaint view" of what it means to be Indian (Brendale v. Confederated Tribes & Bands of Yakima, 1989). Native Americans again are conceptualized only in relation to the land, nature, and images born 500 years ago.

Northern Exposure brings Indians to forefront and yet avoids provoking guilt because it deals with Alaska Natives about whom most Americans know little. Viewers can watch the series without necessarily "feeling bad," as Mander might say. Because the series ignores present-day issues about jobs, environment, land use, education, and subsistence living in Alaska Native villages, viewers can easily escape into a frontier fantasy, albeit one that twists the old formulas. Northern Exposure is sheer whimsy, a trip to a Brigadoon-like land where a bear can turn into a man, Sasquatch cooks French cuisine, and people steal each others dreams. The showís creators have said that the Alaska in Northern Exposure is not a place but a state of mind.

Hollywood film director John Ford may have felt the same about some of his Westerns. He was only telling rousing, good stories, giving the audiences some innocent fun, creating a bit of fantasy. Yet Ford played a major role in shaping the "Indian" image in the American conscious, an image that for most part is distorted, demeaning, violent, and anonymous. In later years, Ford regretted those portraits. He said his movies "killed more Indians than Custer, Beecher, and Chivington put together. . . Letís face it, weíve treated them [Native Americans] badly" (Woll & Miller, 1987, p. 330).

Conclusion

Northern Exposure is not a documentary. The show is not a history lesson about Alaska Natives. Nevertheless, television entertainment programs and other media need to consider minority groups in more realistic, accurate terms. Arguments that televisionís inattention to detail and looseness with facts is endemic to fiction do not hold up. Television would never consider portraying Abraham Lincoln as a midget (Brauer & Brauer, 1975) or former President Ronald Reagan as a Rhodes scholar. Northern Exposure would never have Joel Fleischman praying in a Catholic Church or confuse the movie-making styles of Ed Chigliakís favorite directors. Yet without hesitation, television will distort and distill Native American groups to just such an extent. Dismissing Northern Exposureís treatment of Alaska Native culture as innocent poetic license is exactly what writers, artists, and film and television producers have been doing for 500 years. We know the results. Incorrect, mistaken imagesñespecially when repeatedñbecome naturalized, as African-Americans know so well. They seep into our subconscious, temper our personal attitudes and reactions and influence government policies and Supreme Court decisions.

Northern Exposure tries on the one hand to battle the old myths by bringing Native Americans into the present and proclaiming them, first of all, to still be alive. The show offers a kind and appreciative view of Native American culture that has continued to develop past the year 1900. Rather than melting the native culture into the dominant European-American culture, the show attempts to applaud its distinctions. Rather than stressing those distinctions as points of conflict and alienation, the show attempts to view them as complementary.

With regard to its native characters, the show avoids the villainy, shuns the helpless victim act, and gingerly skirts the morally superior and noble savage portrayals. The characters here are more complex, with good and bad traitsñalthough Ed has committed only one sin. The actors really are Native American (most of them from tribes of the lower 48 states) instead of Euro-Americans in heavy makeup. The Native American characters speak normally and lead average lives. The show is a gigantic step forward in treating a minority group as a part of the world rather than separate from it or in constant conflict with it.

For all the cultural displays and signals of multicultural acceptance, however, Northern Exposure fails to escape its racist antecedents. The Native Americans here are as displaced as the non-native characters who move into Cicely from Canada and the United States. At least the non-natives have previous claims. The Alaska Natives meanwhile are dispossessed of their land, even though in reality they would likely be the largest private landowners in Cicely. Even more damaging, they also are stripped of their cultural identity. While the show intentionally battles fear of the "other," it adds little new, accurate information that would serve to eradicate and replace the old concepts. Until we confront our ignorance about culture diversity, Native Americans will never completely escape the old Custerism or even the new Custerism, which, although kinder and gentler, is no less a stereotype.

Notes

1An examination of encyclopedias of television programming uncovered no series with leading native American characters in more than a decade. A few shows in very recent years, notably Twin Peaks, have had Native American characters, but none was integral to the series or featured consistently week to week. Walker, Texas Ranger, which features a lead character of Native American ancestry, did not premiere until April 1993.

2Seggar, Hafen, & Hannonen-Gladden (1981) studied minoritiesí representation on television between 1971-1980, but did not create distinct listings for Native Americans. They instead grouped them with other minorities, excluding African-Americans.

3For a history of the US government policy toward Pribilof Aleuts, see Jones (1980).

4A series by the name Frontier Doctor ran in syndication in 1958-59.

5In the episode, Ghost Dancers were presented as renegade, murderous Occultists. See Braver & Braver, pp. 177-178.

References

Response from Leonard J. Shedletsky

Taylorís analysis of Northern Exposure, and more generally, mass mediated images of Native Americans, reminds us again that intrapersonal communication is hardly limited to the boundaries of our skin or any particular level of communication--that it operates as we assign meaning. The power of the image both as fashioned by stimuli and the nature of categorization are central here. This reader wonders about stereotypes: given our tendency to form categories, just how do we distinguish between good and bad categories? How do we keep our categories free of prejudice? This chapter would suggest to begin with information rich not poor representations.

Intrapersonal Processing in the Congregation: Recommendations for Enhanced Laity Involvement

Deloris McGee Wanguri

About the Author: Deloris McGee Wanguri received a Ph.D. in organizational communication from the University of Texas at Austin (1984). She is currently an associate professor of Speech Communication, University of Houston-Downtown. Research interests include: confidential information disclosure in organizational settings, organizational socialization, performance appraisals, ethical issues in organizational communication. Her most recent publication is in Communicating in Organizations: A Casebook. An extended, earlier version of this essay was presented at the 43rd annual conference of the International Communication Association, Washington, DC, 1993.

Abstract: In the local congregation, communication occurs on all levels. This article examines the role that appropriate intraorganizational communication can play in facilitating the assimilation process for congregational laity, in deepening their levels of commitment, and in improving their levels of involvement. Specific suggestions include appealing to the motives of the laity, encouraging development of two-way channels for supportive feedback, and encouraging development of and participation in meaningful learning activities inside the congregation.

Perspective:

1. To what extent do certain communication levels dominate interactions within various mosques, synagogues, temples, fellowship halls, and churches in the United States? In spite of doctrinal differences that may exist among these congregations, do their members share similar communication characteristics as they conduct routine business and participate in worship services and activities within the boundaries of the organization?

2. In addition to the three recommendations offered within this article for instilling a deeper sense of commitment and enhancing laity involvement within the local congregation, are there other recommendations you would suggest to a congregational leader? If so, do your recommendations require the use of normative, coercive, or remunerative power?

 

Within the local congregation in the United States, communication occurs on all levels, and it serves numerous functions. Such communication is intrapersonal, interpersonal, small group, public, and mass in form. It serves religious, social, cathartic, persuasive, and administrative functions, to name a few. In this context, the term congregation refers to any assembly of persons meeting for worship and for religious instruction. Like other organizations, local congregations must continually work to strengthen member commitment and to increase member involvement. Without such maintenance, the congregation atrophies and cannot survive. Given the importance of highly committed laity in this setting, this essay examines member commitment and makes recommendations for improved laity involvement.

Organizational commitment, according to Mowday, Steers, & Porter (1979), refers to "the relative strength of an individualís identification with and involvement in a particular organization" (p. 226). It may be reflected attitudinally by: (a) a strong belief in and acceptance of the organizationís goals and values; (b) a willingness to exert considerable effort on behalf of the organization; and (c) a strong desire to maintain membership in the organization. When one is specifically measuring commitment within the local congregation, however, a fourth variable to examine and to consider is a willingness to contribute financially to the program of the congregation (Wanguri, 1990). According to Wuthnow (1990):

Commitment, then, is the process that deepens oneís involvement in an organization.

Most literature that addresses the nonprofit sector assumes that the average "volunteer" is, to some extent, actively involved in the organization. Such an assumption is, however, inappropriate when examining the participation of laity within the local congregation, for there are actually variations of membership within this body. More specifically, an individual can be a part of a local congregation, and hence, a member or "volunteer" within that organization, while being very passive. This may be an individual who simply attends worship services once a week and who participates in no other activities. At the other end of the continuum are those "volunteers" who not only participate in religious services and religious educational experiences several times a week, but who also share in a multitude of other programs that are sponsored by or through the congregation. Such programs are numerous, and they include nonprofit retirement communities, nursing homes, meals-on-wheels programs, pregnancy clinics, foster home programs, shelters for battered women, counseling centers, alcoholism and substance abuse support groups, gender awareness groups, gay fellowships, training programs for the jobless, soup kitchens, and shelters for the homeless.

In view of these variations in the fiber of congregational voluntarism, the strengthening of laity involvement within this entity takes on increased importance. Whether they are teaching, singing, ushering, or greeting, volunteers within the religious community are actively solicited, continually monitored, and generally affirmed within the context of that congregation. Their services are valued by the employed staff, and for this reason, effective communication with these volunteers is essential.

For any member of an organization, regardless of whether that individual is a newcomer or an insider, communicative openness is extremely important during the socialization process. This process is ongoing, beginning in some cases before an individual actually enters or joins the organization and continuing throughout that personís affiliation with the group. To the extent that a person has access to information regarding the institutional "culture"ñits goals, norms, values, policies, practices, procedures, activities, stories, rites of passage, status symbolsñthat individualís assimilation into the mainstream of the organization, is likely to be facilitated. To the extent that a good "fit" exists between a member and an organization, feelings of alienation are likely to be minimized (Hess, 1993). Regardless of the organizational type, these principles are likely to apply.

As one would expect, these premises are particularly true of the local congregation. Although communication does occur at all levels of the congregational hierarchy, the communicative role that the congregational leader can play in facilitating this assimilation process for the laity, in deepening their levels of commitment, and in increasing their levels of involvement, is undeniable. What follows then is a list of recommendations for the leader who wants to utilize her or his communicative skills in order to instill a deeper sense of commitment and to enhance laity involvement through the use of normative power, as opposed to coercive or remunerative power.

Appeal to the Motives of the Laity. Analyze the congregation, and based upon this assessment, develop or reinforce appropriate programs and call for involvement in these programs. For example, for members who are motivated by a desire to interact with the larger community, development of and/or involvement in a voluntary literacy center, might be appropriate. Other motivations include cause-oriented members who are deeply concerned by a particular social issue, self-expressive members who join groups primarily for the sake of enjoyment or personal expression, occupational self-interest members who are motivated by a desire to enhance their careers and by opportunities for personal development, and philanthropic members who concentrate on donating money or raising funds for groups rather than providing direct service (Ilsley, 1990). These motivations can manifest themselves in a number of ways, including participating in voter registration drives, theater groups, and on a more informal level, networking opportunities.

Encourage Development of Two-Way Channels for Supportive Feedback, Involving the Congregational Leader and Members of the Laity. Understandably, there is a tendency to reward largely those members who participate actively in congregational activities, by providing them with positive "strokes" in return for their involvement. Even though this positive feedback is understandable, it frequently ignores those "volunteers" who share in religious activities minimally. Congregational leaders should personally cultivate as well as encourage their constituents to develop consistent networks of communication with all congregational members. These networks may take the form of support groups, for example, which allow members to discuss ideas, opinions, and feelings; to facilitate friendship and emotional support; and to provide input to the larger congregation. Formal and informal channels of communication should acknowledge and reward all members for their participation, regardless of the involvement level. Although the membersí specific responses will be influenced by a number of other variables, such as perceptions of the intentions of the congregational leader and the kinds of alternatives that are available to the member (Shaw, 1981), the leadership should not fail to provide these channels. Recognition of even minimal participation will frequently lead to a heightened sense of belonging, more commitment, and deeper involvement.

Encourage Development of and Participation in Meaningful Learning Activities Inside the Congregation. According to Ilsley (1990), "Programs and organizations that [offer] a wide variety of learning experiences. . .[are] the most likely not only to attract volunteers but to retain their services for long periods of time" (p. 61). Hence, the creation of a learning climate is directly linked to a memberís duration of service. More specifically, the duration of membership in the local congregation, according to organizational literature, is linked to the amount, kind, and quality of learning that members experience within that particular entity. Congregational leaders would do well then, to develop and to encourage involvement in programs that are goal-oriented (providing learning in order to gain a position), activity-oriented (providing learning for the joy of participation), and learning-oriented (providing learning for its own sake) in order to increase involvement. These three purposes, according to Houle (1961), characterize healthy learning climates in organizations.

References

Response from Leonard J. Shedletsky

Professor McGee Wanguri helps to remind us that aspects of intrapersonal communication cut across levels of communication and are just as important to the working of organizations as to individuals. Motivation and learning are two such issues.

Intentionality Attributions for Naturally-Occurring Nonverbal Behaviors in Intimate Relationships

Valerie Manusov

About the Author: Valerie Manusov (Ph.D., University of Southern California, 1989) is an assistant professor at the University of Washingtonís Department of Speech Communication, DL-15, Seattle, WA 98195. Portions of this article were presented to the Speech Communication Association, Chicago, 1992 and to the International Network on Personal Relationships, Milwaukee, 1993.

Abstract: The attributions partners offer for anotherís behaviors are an important part of how people make sense of nonverbal cues. In this study, participants kept logs recording their meaning and their partnersí nonverbal behavior. Each log entry was rated for valence and intent by the participant. Satisfaction was inversely related to the tendency to provide entries for partnersí behaviors that were seen as negative, and it correlated positively with the likelihood of attributing greater intentionality to facial expressions. Space and touch were most likely to be judged as intentional, and eye behavior and facial expressions were attributed as least intentional. Highly positive behaviors were viewed as most intentional, whereas very negative behaviors were seen as least intended.

Perspective:

1. When you notice your partner using certain nonverbal behaviors (e.g., facial expressions, gaze, vocal cues), how do you know what the behaviors mean? Could the meanings be different from what you think they are?

2. What nonverbal behaviors are most likely to catch your attention? Why?

3. People can choose various meanings for othersí nonverbal behaviors. What are some of the things that are likely to affect this choice? What does this say about the nature of meaning?

4. Much of the time, we do not pay conscious attention to othersí nonverbal cues. Under what conditions are people likely to think more overtly about othersí behaviors? Why?

 

One of the primary ways in which members of romantic dyads try to access their partnersí feelings, beliefs, intentions, and motivations is through interpreting the nonverbal indicators used by the other. That is, when couple members see a facial expression, hear a vocal tone, or note the posture of their partner, they may infer that the nonverbal behavior signified a certain mood, feeling, or attitude or that it was sent intentionally to convey a particular message. Although most of this meaning-making occurs quickly and with little consciousness, other times the meaning of nonverbal behaviors is ambiguous and requires more thoughtful interpretation (Sillars & Scott, 1983).

Social behaviors that are not easily understood are subject to attribution-making as people try to assess the cause, and ultimately the meaning, of the observed actions (Heider, 1958). The ambiguity of certain behaviors allows people to decide whether the cause of the action was internal or external to the other, whether it reflected a stable characteristic of the person or the situation, whether it was a specific or global cause, and whether the other acted intentionally (Weiner, 1986). Each of these choices helps determine what meaning can be taken from the otherís behavior because each leads to different interpretations and evaluations of the action.

Although people can provide rational heuristics for determining the cause of a behavior, interpreters usually have perceptual and motivational biases that influence the accuracy of their decoding (Floyd & Markman, 1983; Jacobson & Moore, 1981). One of these, relational satisfaction, has been found to be a strong determinant of the ways people assign causes to their partnersí actions. Although the interplay between satisfaction and behavior has been applied to other cues, little research has investigated the role of satisfaction as a bias in interpreting nonverbal behaviors. Because nonverbal cues provide an important source of meaning within intimate relationships, the present investigation seeks to add to the growing body of information regarding the interpretation of nonverbal behaviors in intimate relationships. In particular, it investigates the ways in which couples attribute intentionality for the everyday occurrence of nonverbal cues.

Nonverbal Communication in Romantic Dyads

According to Noller (1984) three communicative functions that are particularly important in intimate relationships include 1) conveying interpersonal attitudes, 2) expressing emotion, and 3) handling ongoing interaction. These functions are most often associated with nonverbal messages (see Burgoon, 1985; Patterson, 1987), and the nonverbal means of communicating these goals have been investigated in depth (e.g., Bull, 1987; Buller, 1991; Duncan & Fiske, 1977; Ekman, 1982; Mehrabian, 1981). When studied within the context of ongoing relationships, the focus of nonverbal communication studies has been on determining differences in the use and perception of nonverbal cues between satisfied (nondistressed) and dissatisfied (distressed) couples. Overall, the use of nonverbal cues, especially the expression and interpretation of negative affect, has been found to be a strong discriminator between satisfied and dissatisfied couples (Gottman, Markman, & Notarius, 1977; Levenson & Gottman, 1983; Sillars, Pike, Jones, & Redmond, 1983).

Kahn (1970) was one of the first to study the link between nonverbal cues and marital satisfaction. Using the Marital Communication Scale, couples read a passage (e.g., husbands would ask, "Didnít we have chicken a few nights ago?") with various vocal inflections designed to express different intentions (e.g., that the encoder is irritated at having the same meal, that the speaker was curious if his memory was correct, or that the husband was very happy because the chicken dish was his favorite). Kahn found that satisfied couples were more accurate in reading their partnerís intention from the vocal cues than were dissatisfied couples.

The disparity that Kahn found between distressed and nondistressed couples may have involved an inability to decode nonverbal intentions correctly as implied by the study, a limitation in the ability to encode the intentions clearly, or some interaction of the two processes. In more recent extensions of Kahnís work, Noller (1980) and Gottman and Porterfield (1981) found that, not only are both encoding and decoding processes involved in the perceptual differences based on satisfaction level, but gender differences confound the process further, with dissatisfied husbandsí lack of decoding skill accounting for much of the variance in accuracy (for extensive discussions of gender differences see Noller, 1984; 1987).

Other researchers have looked at the occurrence of affect displays in satisfied and dissatisfied couplesí behavior. Although there is some discrepancy in this area (see Barnett & Nietzel, 1979), it appears that negative nonverbal displays are generally more common in the interactions of distressed couples than in nondistressed dyads (Bradbury & Fincham, 1987; Gottman, 1979, 1980), and this mirrors research on other interaction behaviors (Billings, 1979; Jacobson, Waldron, & Moore, 1980; Wills, Weiss, & Patterson, 1974). The conclusion is less clear regarding positive nonverbal affect, although some researchers suggest that satisfied couples may use more positive cues (Gottman, 1980; Margolin & Wampold, 1981) and/or that they are more efficient in sending positive affect through nonverbal means than are dissatisfied couples (Noller, 1987).

Interestingly, however, differences in frequency do not appear to be the most important characteristic of behavior distinguishing couples. Especially for negative cues, the ways that the behaviors are noticed, interpreted, and the impact they have on the partner are more likely to reflect a coupleís satisfaction level than the actual occurrence of the behavior (Gottman et al., 1977). In fact, in their study of a variety of interaction behaviors, Jacobson, Follette, and McDonald (1982) found that only 13 percent of reported events were negative for dissatisfied couples (versus 6 percent for satisfied dyads). However, distressed couples were likely to be underrepresented by 50 percent the number of positive behaviors that actually occurred (Robinson & Price, 1980). This indicates that the proportion of negative behaviors may be even less than couples believe. Yet, when they do occur, dissatisfied couples are more reactive to and will be more likely to reciprocate negative behaviors than will satisfied couples (Jacobson et al., 1980; Schachter & OíLeary, 1985).

What appears to be the greater discriminator between satisfied and dissatisfied couples is not the frequency of behaviors but rather the cognitive processes that result from the behaviors (Epstein, Pretzer, & Fleming, 1987; Jacobson, 1984). Arias and Beach (1987) contended that "an individual does not respond to the environment per se but rather, the individual responds to her or his perceptions and interpretations of the environment" (p. 109). In particular, partnersí sentiment about one another appear to affect cognitive reactions more than will the nature of the behavior (Floyd, 1988).

Cognitions in Romantic Relationships

If an overall schema or sentiment about oneís partner is unfavorable, the impact of an observed behavior is also likely to be negative. Gottman (1979, 1980) helped confirm this by noting that dissatisfied couples rate their partnerís intent much more negatively than the spouses rate their own action. Likewise, a positive sentiment may bias satisfied couplesí evaluations such that people are likely to rate their partnersí behavior as more positive than the partners judged themselves (Arias & OíLeary, 1981 cited in Beach & Arias, 1983). The tendency for the affective nature of the schema partners hold about one another to influence the interpretations they make led Jacobson and Moore (1981) to note that "spouses are not passive and objective observers of their partnersí behavior; rather they are participants with biases, and these biases need to be understood" (p. 276).

Attribution Theories

Although there has been a significant amount of research looking at the use of nonverbal behaviors in intimate dyads, less focus has been offered to understand the ways in which couples interpret non-language cues within the context of their relationship. Other actions that occur within relationships (e.g., conflict, sexual overtures, household tasks) have been investigated from the perspective of attribution theories, however, and consistent results have been found.

Most commonly, there is a tendency for dissatisfied couples to make "distress-maintaining" attributions and for nondistressed partners to offer "relationship-enhancing" interpretations for one anotherís behaviors (Fincham & Bradbury, 1987; Holtzworth-Munroe & Jacobson, 1985). That is, in looking for the causes of their partnersí behaviors, people who are dissatisfied tend to see negative behaviors as caused by internal, stable, and global factors of their partner, whereas positive behaviors are interpreted as having been caused by external, unstable, and specific circumstances; the opposite is more likely to occur with satisfied couples. The conclusion that dissatisfied couples are likely to attribute causes to their partnersí behavior in ways that perpetuate the negative tone of the relationship and that satisfied couples tend to make their attributions in ways that support a more positive climate, has received strong support (Camper, Jacobson, Holtzworth-Munroe, & Schmaling, 1988; Fincham & Beach, 1988; Jacobson, McDonald, Follette, & Berley, 1985).

One investigation applied this thesis to nonverbal cues. Following other researchersí contentions that people make attributions for their own and othersí nonverbal behaviors (McMahan, 1976; Seibold & Spitzberg, 1981; Sillars, 1982), Manusov (1990) had couples interact in a laboratory where one (a confederate) altered her or his behaviors to act particularly positive or negative. Similar to other studies (e.g., Pszyszynski & Greenberg, 1981; Wong & Weiner, 1981), the results of this investigation showed that negative behaviors were more likely to generate attributions, and that, in general, satisfaction level correlated with the types of attributions made for the nonverbal behaviors.

Prior to the empirical test, Manusov (1990) offered a suggestion for why nonverbal behaviors had not been studied previously within an attribution framework. She noted that the tenets of initial attribution theories included the proposition that, in making causal choices, people must believe that another may have had the intent and the ability to enact the behavior being interpreted (e.g., Jones & Davis, 1965; Schneider, Hastorf, & Ellsworth, 1979). If the other could not control the behavior, the cause automatically becomes one of circumstance, and the choice about the source of the behavior is limited to non-social, or at least non-active, causes. Thus, behaviors must be perceived as potentially intentional and controllable to be subject to causal attribution as it is generally conceived.

It may be for this reason that nonverbal behaviors are usually not studied as legitimate sources of attribution-making. Because researchers have often looked at nonverbal behaviors as spontaneous "leaks" of internal states rather than potentially controllable and intentional behaviors, the cues may not have been seen to fit into the attribution paradigm. Manusov and Rodriguez (1989), however, provided support for their argument that people do see nonverbal behaviors as intentional in certain instances, and therefore the cues are subject to inference-making in much the same way as other social behaviors. Further, the traditional attribution paradigm overlooks another source of action that can account for the meaning attached to communicative behaviors: spontaneous reflections of a personís traits or states. As mentioned previously, the choice of attributional "path" determines to some extent the ultimate choice of what a behavior or set of behaviors may mean to an interpreter (Manusov, 1990).

Intentionality and Controllability of Nonverbal Behaviors

A number of researchers have already investigated differences in actual or perceived control and intentionality across different types of nonverbal cues. Friedman, Riggio, and Segall (1980) and Motley and Camden (1988), for example, documented the ability to pose (and therefore control and intend) facial expressions. Murphy-Berman and Berman (1978) looked at proximity as a potentially intended symbol of friendliness. Touch has also been investigated as a controllable and intentional cue with strong social meanings (e.g., Andersen & Sull, 1985; Jones & Yarbrough, 1985). Looking across behaviors, Rosenthal and DePaulo (1979; DePaulo & Rosenthal, 1979) concluded that nonverbal channels can be seen as falling into a hierarchy of controllability, with the face as the most controllable, followed by the body, the voice, and microexpressions. Although this contention was based on encoding ability, people also make judgments of intentionality behind anotherís nonverbal behavior as part of the decoding process (Manusov & Rodriguez, 1989; Strube & Werner, 1982).

Little of this research has focused specifically on intentionality attributions within ongoing relationships. Manusov (1990), however, did find that intentionality correlated negatively with relational satisfaction for those behaviors encoded with negative valence. Many of the attributions that were offered by the participants were to variables related to the experimental setting, however (e.g., the camera, being at the study). The awareness of the laboratory thus affected the outcome of the experiment and implied a threat to the validity of the results, a concern arising in other attribution studies as well. To counter this, Holtzworth-Munroe and Jacobson (1985) suggested looking at spontaneous attributions that are made in ongoing, naturally-occurring interactions. One such method is the use of diaries or logs to capture the thoughts people have about one anotherís behaviors soon after they happen.

The present study is an extension of the investigation into attribution-making for nonverbal behaviors as they occur in everyday interactions, focusing particularly on intentionality assessments, because of their importance for determining the meaning of nonverbal actions. To reiterate, past research suggests that satisfaction level influences the likelihood of noting certain types of behaviors, with dissatisfied couples taking note of more negative behaviors and satisfied dyads more likely to see positively-valenced cues (Bradbury & Fincham, 1987; Gottman, 1979, 1980). Satisfied couples are also more likely to view behaviors neutrally (Manusov, 1990). Thus, the present investigation hypothesizes that within naturally occurring encounters,

Hypothesis 1a: There is a positive correlation between relational satisfaction and the proportion of attributions made for behaviors seen as positive.

Hypothesis 1b: There is a positive correlation between relational satisfaction and the proportion of attributions made for behaviors seen as neutral.

Hypothesis 1c: There is a negative correlation between relational satisfaction and the proportion of attributions made for behaviors seen as negative.

Second, following the general trend in attributions for dissatisfied couples to make "distress-maintaining" causal explanations and for satisfied couples to offer more "relationship-enhancing" attributions (Fincham & Beach, 1988; Holtzworth-Munroe & Jacobson, 1985; Jacobson, McDonald, Follette, & Berley, 1985), and the argument that intentionality attributions are a key dimension of this attribution-making (Manusov & Rodriguez, 1989), it is hypothesized that,

Hypothesis 2a: As relationship satisfaction increases, the tendency to see oneís partnerís positive behaviors as intentional also increases.

Hypothesis 2b: As relationship satisfaction increases, the tendency to see oneís partnerís negative behaviors as intentional decreases.

Finally, it has been suggested that diverse nonverbal cues are viewed with different levels of intent or control (McClintock & Hunt, 1975; Rosenthal & DePaulo, 1979). Most of this research has been based on researchersí views of encoding discrepancies, however. To further an understanding of the perceived differences in the nature of nonverbal messages and how these are tied into the ways in which relational satisfaction affects this perception, three research questions are posed.

Research Question 1: What nonverbal behaviors do couples notice in their everyday interactions?

Research Question 2: Is there a difference in perceived intent for these behaviors?

Research Question 3: To what extent does relational satisfaction predict different perceptions of intentionality for nonverbal cues?

Method

Participants

Participants (N = 46) were recruited from local newspaper advertisements asking for married or living together couples to become involved in a study on nonverbal behavior. Each couple was paid $35 for their three week involvement. The mean age for the participants in the study was 28.09 (with a range from 20-68). Thirty five percent were married. The average length of time since the couple had been together was 79.63 months (approximately 6 1/2 years) with a range from 8 to 429 months (35 1/2 years). One couple was homosexual, and the others were heterosexual. One couple was Indian, one was African-American, one member was Hispanic, one member was Asian, and the remaining participants were White.

Procedure

The couples were asked to meet with the researcher in small groups to explain the process of log-keeping used for the study. After a brief introduction, each member of the couple was placed apart from one another in the room in order to complete a series of scales, including an assessment of relational satisfaction. They were given both written and oral assurance that their answers would be kept confidential and that their partner would not see their answers. Following this, the couples were told that the researcher was interested in finding out what nonverbal behaviors people notice in their daily interactions. They were given examples of "nonverbal behaviors," including facial expressions, eye behaviors, vocalics (including silence), physical appearance and artifacts, body movements, touch, and the use of time to communicate. A range of behaviors was mentioned, because noting only certain cues would likely bias what the participants recorded.

The members of the couples were asked to keep a log of the times they found themselves thinking about why they or their partner may have used a behavior or set of behaviors. They were asked to record the day and time of the occurrence, the behavior(s) they noticed, and what meaning they attributed to the behavior(s). They were also asked to rate the behaviors on two scales. First, they made an assessment from 0 to 100 as to how positive or negative they perceived the behavior(s) to be. A 0 represented a very negative attribution, and a 100 meant very positive. A score of 50 indicated a neutral response. Second, they were asked to indicate the degree to which they thought the behavior(s) was or were enacted intentionally. In this case, 0 meant that they believed the behaviors occurred spontaneously, without any intent. A score of 100 indicated that they saw the behavior(s) as highly intentional. A 50 meant a moderate degree of intentionality.

Each couple continued this diary for two weeks, after having been instructed not to talk with one another about the logs. Following the two weeks, the couples came in one at a time, and demographic information was collected. They were allowed to ask any questions about the study, offer their observations, and were thanked for their participation in the study.

Variables

Relational Satisfaction. Satisfaction in this study was measured by Nortonís (1983) Quality Marriage Index (QMI). The mean satisfaction level for the sample was 118 with a range from 73 to 141. In this study, the QMI had an alpha reliability of .89.

Coding of Behaviors and Attributions. Two undergraduate research assistants were trained to determine what entries were acceptable for the current study. The investigator went through two complete logs with each coder separately, and then they took home copies of two other logs in order to test their inter-coder reliability. The coders had 94 percent agreement on what counted as acceptable entries (see below). One husband did not provide any ratings for his entries, and two husbands and two wives did not provide intentionality scores. The former was left out of all analyses, and the latter were excluded from those analyses measuring intent.

The coders discarded specific entries from each log where the behaviors discussed were not nonverbal cues. These included such things as participating in an event (e.g., a party, sports), verbal behavior, or actions such as making dinner or studying. In other cases, the participants occasionally noted a mood or state ("She copped an attitude" or "He was tired") without specifying the accompanying behaviors. From the remaining entries, the coders counted the total number of entries and the proportion of these that were positive (greater than 50), negative (less than 50), and neutral (50) according to the ratings provided by the participants. These scores were used to test hypothesis 1.

In order to measure hypothesis 2 , two intentionality means were created. One mean, for negative behavior, was created from the intentionality ratings for all of the valence entries that were less than 50. A second mean, for positive cues, was assessed from those scores with valence ratings greater than 50.

To test the research questions, behavioral categories were created that best represented the nonverbal behaviors mentioned by the participants. For each of these categories, the average intentionality score was computed. Because not all respondents mentioned all of the coded behaviors, only those that were reported by most of the participants were used to answer research questions two and three. Because of the nature of the analysis, mean substitutions were used if any of the behaviors were not included in a personís log.

Results

Tests of Relationships

According to Kenny and Kashy (1991), data that involve information from both members of a dyad should first be analyzed for correlations between the scores of the two dyad members. Scores that are significantly related violate the independence assumptions of parametric tests, and the dyad rather than the individual should be used as the basis of analysis. Using two-tailed Pearson correlations, all of the above predictor variables were analyzed. However, none of the variables were significantly correlated across the partners. Thus, individual scores were used in the following analyses.

Tests of Hypotheses

Hypothesis 1a. Hypothesis 1a predicted that as satisfaction increases the proportion of cues judged to be positive also increases. A one-tailed Pearson correlation, testing the relationship between satisfaction scores and the proportion scores for positive cues, was run. It found no support for the hypothesis , r = .05, p > .05.

Hypothesis 1b. Hypothesis 1b predicted that as satisfaction increases the proportion of cues judged to be neutral also increases. A one-tailed Pearson correlation, testing the relationship between satisfaction scores and the proportion scores for neutral cues, was run. It found no support for the hypothesis , r = .16, p > .05.

Hypothesis 1c. Hypothesis 1c predicted that as satisfaction increases the proportion of cues judged to be negative decreases. A Pearson correlation, testing the relationship of satisfaction scores with the proportion scores for negative cues, was run. It offered support for the hypothesis , r = -.42, p < .01.

Hypothesis 2a. Hypothesis 2a stated that as relational satisfaction increases the tendency to view positive behaviors as intentional also increases. A Pearson correlation was run between relational satisfaction and the mean intentionality score for cues seen to be positive. It found no support for the hypothesis , r = -.09, p > .05.

Hypothesis 2b. Hypothesis 2a stated that as relational satisfaction increases the tendency to view negative behaviors as intentional decreases. A Pearson correlation was run between relational satisfaction and the mean intentionality score for cues seen to be negative. Although a trend emerged, the hypothesis was not confirmed, r = -.21, p > .05.

Research Questions

Behaviors. In answer to research question one, the most commonly cited behaviors overall were facial expressions or looks (150), eye behavior (126), kinesics or body movements (117), touch (113), vocalics or paralanguage (88), silence (54), proxemics, or use of space, (37), kisses (33), chronemics, or use of time, (12), and artifacts (11). Most of these entries involved only one nonverbal cue, but composite behaviors were also mentioned. The most common behaviors to be co-listed with other cues were eye behavior (with vocalics, kinesics, and facial expressions), facial expressions (with kisses and kinesics), and touch, or haptics, (with proxemics and vocalics).

Differences In Intentionality Across Behaviors. A within-subjects ANOVA was run looking for differences in the means across proxemics, silence, eye behavior, kinesics, vocalics, facial expressions, and haptics. Most of the participants did not mention chronemics, kissing, or artifacts, so these were not included in the analyses. The ANOVA was significant, F (6, 35) = 9.41, p < .05. Post hoc comparisons showed that all of the means differed significantly from each other. Proxemic behaviors were seen as most intentional (M = 79.93), followed by haptics (M = 77.95), vocalics (M = 73.17), and kinesics (M = 70.22). Facial expressions were given judgments of the least intentionality (M = 56.51) with eye behavior (M = 60.68) and silence (M = 62.54) also receiving lower intentionality scores.

Satisfaction and Intentionality Attributions. Because this last question was exploratory, a stepwise Multiple Regression was run with satisfaction regressed on proxemics, silence, eye behavior, kinesics, vocalics, facial expressions, and haptics. Only one cue entered the equation. It was found that viewing facial expressions as more intentional predicted greater relational satisfaction, F (1, 38) = 4.23, p < .05, b = .20, R2 = .10.

Discussion

This study was designed to look at the ways people think about the meaning and intentionality of their romantic partnersí nonverbal behavior in everyday interactions as well as how they make attributions of intent for specific nonverbal cues. Consistent with past research, the tendency to pick out what are perceived to be negative behaviors was inversely related to relational satisfaction. However, there was no relationship between satisfaction and the proportion of recorded cues that were rated as positive or neutral. As has been found previously, the occurrence of negative behaviors, or in this case those cues seen by the attributor as negative, were found to be the most important discriminator between couples.

Interestingly, negative actions accounted for almost half of all entries across participants, with positive cues making up 33 percent of the entries, and neutral behaviors accounting for 19 percent. Indeed, in the debriefing following the log-keeping, the largely satisfied sample commented consistently on how much more likely they were to pay attention to negative cues. However, the greater proportion of negatively-labeled behaviors noted by partners in dissatisfied dyads points to the argument that they may be overly oriented to the untoward behaviors used by their partners. This may confound an already spiraling behavioral cycle.

Against predictions, however, satisfaction was not significantly related to the attributions of intent made for nonverbal behaviors. In post hoc explorations it was found that, across couples, people were most likely to assume a high degree of intent behind cues seen as very positive (M = 93.08), less with behaviors viewed as somewhat positive (M = 79.70) and somewhat negative (M = 74.27), and they assumed the least intent behind behaviors they interpreted to be very negative (M = 60.66). These numbers indicate that all of the participants in this study were acting in meaning-making that was more "relationship enhancing." However, there was a trend to see behaviors rated as negative as more intentional when satisfaction decreased. Due to the limitations in finding significance in this relatively small sample, this important tendency should not be overlooked.

Because of the general consistency of past research regarding the role of satisfaction, the present lack of a significant relationship between satisfaction and attributions is surprising. The current results may be due in part to the relatively high level of satisfaction in these couples. Most other research on attributions uses "clinic" couples; that is, couples involved in marriage counseling. However, there was adequate range in the QMI scores in this study to have picked up differences that may exist, despite a negative skew.

Another possibility for the inconsistency with past research may be due to methodology. Past research such as Gottmanís (1980) and Manusovís (1990), had participants interpret one anotherís cues in observed settings. There may be something in the "public performance" of a laboratory that highlights the disparity between satisfied and dissatisfied dyads. The present study suggests that private interpretations may be less clear-cut. When looked at across a period of time and a number of different behaviors and messages, the disparity between levels of satisfaction may be less obvious. This suggestion is very important for relational attribution researchers because it implies, as other researchers have suggested, that attribution-making is somewhat unstable and influenced by the situation in which it occurs. The role of an "audience" may prove to be an important contextual feature.

The above findings also suggest that in everyday interactions, couples may use relatively automatic processing of the meaning for nonverbal cues rather than more complex attributional processing. Following a social meaning model (Burgoon, Coker, & Coker, 1986; Burgoon, Manusov, Mineo, & Hale, 1985), couples in this study appeared to have relatively consistent views of the relationship between valence and intent. Similar to Manusov and Rodriguez (1989) who looked at stranger interactions, there was a strong tendency in these data to see positive behaviors as intended and negative behaviors as unintended, and this was exaggerated by the extremity of the valence rankings.

The belief that people may share largely stable social meanings for nonverbal behavior is given further support by the intentionality ratings provided for particular behaviors. Although satisfied couples were more likely to see facial expressions as intentional, there was no relationship between satisfaction and the other nonverbal cues mentioned by the participants. There was, however, agreement about the degree of intentionality behind any given nonverbal cue across the sample. Indeed, proxemics and haptics were seen as most intentional, whereas facial expressions and eye behavior were given the lowest intentionality scores.

The fact that certain types of behavior were seen as more intended than other cues suggests that there may also be something inherent in the style of the behaviors used. Consistent with past research on encoding (e.g., Motley & Camden, 1988), it may be the exaggerated nature of the behavior that cues people to see the behavior as intended. While this assumption differs from some past research (e.g., Manusov, 1991), it is worth further exploration. Unfortunately, the design of this study did not allow the opportunity to see the actual behaviors on which the participantsí judgments were based.

What is also interesting about this result is that it somewhat contradicts research investigating actual encoding potential of nonverbal cues. Although researchers have argued that proxemics and touch behavior are potentially intentional and controllable (e.g., Andersen & Sull, 1985; Jones & Yarbrough, 1985; Murphy-Berman & Berman, 1978), Rosenthal and DePaulo (1979) have made a good argument that facial expressions are the most controllable. Yet in these data, the participants were likely to view their partnersí facial expressions as the least intended of the cues they recorded. Despite the encoding potential of facial expressions, the participants interpreted the face as the most spontaneous of the nonverbal cues. This suggests an important distinction between encoding ability and decoding tendencies.

The use of logs in this study helped to assess some of the thinking that occurs for nonverbal behaviors in peopleís everyday interactions with their intimate partners. It allowed for a variety of behaviors, meanings, and moods to be investigated. In this way, it helped to show that, across all couples, there is a range in the behaviors people use in their everyday meaning-making. Indeed, couples were likely to note facial expressions or looks, silence, kisses, vocal tone, smiles, and eye behavior, and a number also mentioned artifacts, silence, and the use of time.

Log-keeping can be a problematic source of data, however. The participants varied on the time that they gave to the logs, with a range of entries from 2 to 41, and a mean of 16. This may have reflected differences in the actual number of times that the person thought about the meaning behind certain nonverbal behaviors. It may, however, be indicative of how much energy the participants wished to give to the project. Therefore, it is uncertain whether the entries coded here reflect the majority of the thought processes that actually occurred.

Despite these limitations, the present study is useful in its ability to support some of the findings of past research, although it raises issue with the argument that meaning-making is vastly different based on satisfaction level. This assertion suggests that diverse methodologies employed by cognition researchers may tap into different types of sense-making and may help broaden our view of how attributions occur. In addition to the results regarding the proportion of messages viewed as negative, it also highlights the contention that intentionality attributions are an important part of meaning-making for nonverbal cues, even though they did not differentiate significantly between satisfaction levels.

Across all entries, people were more likely to rank the behaviors used as intentionally encoded. This may reflect a tendency to pick up intentional messages more easily; however, it also helps show that people can and do see their own and one anotherís nonverbal behaviors as potentially intended as communicative messages. This contradicts early theorizing on the use of nonverbal cues as a communication source, and helps show more about the complexity of the attribution system.

References

Response from Leonard J. Shedletsky

Dr. Manusovís review of the literature and her study remind us forcefully that external behavior per se is not enough to understand communication. We must look to intrapersonal behavior to understand what goes on between people. Meaning is central to this research. Furthermore, the symbolic context in which meaning-making occurs is a critical component. How can we direct our energies to studying context?

Faces and Voices: Comparing the Impact on Person Identification

Patricia Rockwell

About the Author: Patricia Rockwell is on the faculty of the Department of Communication, University of Southwestern Louisiana, Lafayette, LA 70504. This article was originally presented at the International Communication Association Convention, Washington, DC, 1993.

Abstract: Conflicting evidence exists regarding how accurately people identify individuals from their faces or their voices. Some research suggests that face memory is excellent; however, data from eyewitness testimony studies indicate rather poor face memory. Although much non-memory research has been conducted that compares audio and visual channels, no research could be located that pitted identification from faces against identification from voices. The aim of this article was to determine if there was a difference in the accuracy of person identification by face and person identification by voice. As a corollary, an examination of induction methodñimpression formation versus recallñwas attempted. Results indicated significant main effects for channel and induction method. The superiority of face identification over voice identification was partially confirmed, but the superiority of the impression formation induction over the recall induction was not supported.

Perspective:

1. Why is our memory for visual stimuli so much better than our memory for auditory stimuli?

2. How do most people remember proper names?

 

For most of us, identifying persons means remembering their faces in connection with their names, but there are numerous situations in which persons must be identified by their voices (e.g., on the telephone, in a dark room). Although extensive research has been conducted on both face perception and voice perception, no research could be located that compared the impact of the two types of perception on person identification. This article attempts to rectify this shortcoming by first, examining the general theoretical components of person perception; second, by examining the specific components of face perception and voice perception; third, by comparing face perception and voice perception; and finally, by testing the two types of perception on person identification.

Person Perception Research

A variety of information can be perceived about a personñappearance, behavior, and traits (Fiske & Taylor, 1991). This information can be encoded directly or mediated by the receiver as attributions (Winograd, 1976).

Many researchers posit a general theory of perception which subsumes the processing of all person information. Some researchers have suggested different types of coding (Fiske & Taylor, l991)ñpropositional, a rational processing reserved for certain logical functions; temporal or sequential, reserved for events or behavior that occur over time, such as language; and analog, a type of coding reserved for holistic or gestalt processes such as vision. If these three types of coding represent the full spectrum of possible coding, it might well be suggested that face perception is an analog process because face perception relies on the visual modality. Voice perception, however, may be either an analog process or a temporal or sequential process. Some researchers go even farther, suggesting that encoding and storing of person information is accomplished by special mechanisms reserved especially for this type of encoding (Ellis, 1975). Researchers note that person memory encompasses a variety of memory tasks. Identification requires the individual to label (name) a target face or voice correctly. Recognition requires the individual to correctly compare a certain characteristic (such as a face or voice) with another stimulus characteristic (such as the same or different face or voice). Recall requires the individual to remember certain characteristics of a person such as physical traits or behaviors. In general, researchers have found recognition and recall tasks easier for participants than identification tasks (Cohen, 1989). This finding is comparable to everyday situations where one might recognize personsí faces or voices but not be able to identify their names. Even so, person identification is a common and necessary social task.

Some researchers argue that impression formation involves a deeper mode of processing than simple memory tasks (Craik & Lockhart, 1972). Thus, researchers sometimes include impression formation tasks along with recall tasks in order to determine if accuracy will increase in these situations. Indeed, some researchers have found that recall accuracy improves when the induction itself is presented as an impression formation task rather than a memorization task (Fiske & Taylor, 1991; Stangor & Ruble, 1989).

Person identification tasks are argued to be difficult because personsí proper names are not implicitly relevant to either their face or their voice, have limited use, and are difficult to imagine (Bruce, 1990; Cohen, 1989). Searle (1969) contends that proper names have different properties from person descriptions such as "friendly" or "intelligent." This argument is supported by researchers who have determined that people are less accurate in supplying proper names for persons than they are in supplying names for objects (Cohen & Faulkner, 1986; McKenna & Warrington, 1980; Reason & Lucas, 1984). Even so, participants in experiments can often recall proper names if given additional time to draw on various idiosyncratic strategies to jog their memory (Cohen & Faulkner, 1986).

Perception of Faces

Perceiving visual stimuli is different from perceiving other types of stimuli. As perceivers we can choose to look at or not a particular visual stimuli. Thus, we can be more discriminating in deciding which visual stimuli we want to remember (Ong, 1982). Faces are such visual stimuli.

Some researchers (Ellis, 1975; Richardson, 1980) suggest that face perception involves a specialized type of visual processing unlike that required for visual processing of other objects. Research (Yin, 1970) with brain injured patients who suffer from prosopagnosia, or the inability to recognize familiar faces, leads researchers to conclude that face perception is a specialized neurological skill. Researchers posit that faces are encoded by a scanning process (Clifford & Bull, 1978) in which perceivers encode facial features from the top of the head to the bottom of the chin. Confirmation of this theory has been found by researchers (Baddeley & Woodhead, 1983; Yin, 1969) studying face inversion, in which participantsí recognition of upside-down faces is inferior to recognition of faces in the canonical position. Other researchers (Tversky & Baratz, 1985), however, have found that distinctive features do not promote recall, a finding that refutes the special mechanism argument.

Remembering faces is a task we perform every day usually with relative skill. However, our performance can be affected by circumstance, context, frequency of contact, and number of faces we must remember (Barrack, 1984). In most social situations face memory is excellent; however, in stressful situations such as those encountered by eyewitnesses to crimes or participants in experiments, face memory deteriorates rapidly (Clifford & Bull, 1978). Additionally, accuracy of face memory in experimental situations depends on other factors such as induction method and task difficulty.

In addition to assessing accuracy of face memory, researchers often test perceiversí attributions (beauty, pleasantness) of stimulus faces (Cross, Cross, & Daley, 1971; Ellis, 1975). A consistent finding in this area of research is a curvilinear relationship between identification accuracy and stimulus attributions; that is, faces eliciting extreme attributions are remembered better than faces eliciting neutral attributions.

Laboratory testing of face memory presents some unique problems for researchers. Researchers must decide whether to test identification, recognition, or recall. They must decide whether to use live or simulated stimuli (real people, photographs, or videotape) to achieve optimum ecological validity. They must determine total number of exposures to the stimulus, length of exposure time to the stimulus, and length of time before testing the response stimulus (Barrack, 1984). Finally, they must decide if there will be any interference or distraction during presentation of the stimulus.

Perception of Voices

Vocal characteristics consist of the speakerís pitch, tempo, volume, quality, and articulation, all overlaid onto the verbal content of speech. These characteristics are affected by the speakerís biological, psychological, and social characteristics (Laver, 1968). In addition, vocal characteristics may be categorized as state characteristics (such as increased pitch during arousal), trait characteristics (such as decreased volume in a reticent speaker); or group characteristics (such as disintegrated vocal quality in the elderly) (Brown, 1987).

The processing of voices is different from the processing of faces. First, a perceiver has no decision as to whether to hear a voice as opposed to whether to look at a face, because sound is pervasive. The ubiquitous nature of sound may make us more attuned to vocal phenomena or, possibly like much background noise, it may make us less (Ong, 1982).

Whereas the face is considered part of a personís appearance, the voice may be considered as either part of oneís appearance or part of oneís behavior. Vocal information is transmitted sequentially, suggesting temporal coding as with behavior; however, individual sounds may be analyzed in a static manner and individual voices may be perceived holistically, suggesting analog coding.

Testing of voice memory shares problems with testing of face memory. As with face memory, voice memory researchers must decide whether to use live or simulated stimuli such as an audio tape (Brown, 1987). They must decide the number of speakers to present. They must decide the amount of exposure time each stimuli is presented as well as the amount of response time allowed participants in answering (Clifford & Bull, 1978; McGehee, 1937). They must determine how homogeneous the pool of stimulus individuals will be, as greater homogeneity has been shown to reduce accuracy of recall (Williams, 1964).

Some problems, however, are unique to voice memory research. For instance, researchers must determine whether the stimulus material will be read or extemporized by the stimulus individual. They must select the presentation mode which may be different from that used in face memory research, such as playing taped stimulus material backwards (Bricker & Pruzansky, 1966; Clarke, Becker, & Nixon, 1966; Williams, 1964) or content-filtering stimulus material. These attempts to remove semantic content (Brown, 1987) must be balanced against the need for authenticity in the stimulus material.

What are people able to remember about a voice? Studies have indicated that experimental participants can determine speaker characteristics such as age, gender, even height and weight (Lass & Davis, 1976) from the voice. These characteristics can be objectively determined by acoustic analysis or by trained listeners. Researchers find that attributions based on vocal cues are generally consistent across listeners (Apple, Streeter, & Kraus, 1979; Aronovitch, 1976; Brown, Strong, & Rencher, 1973; Ray, 1986). Thus, it is possible that listeners utilize vocal cues to attribute characteristics to speakers.

Comparison of the Impact of Face and Voice Perception on Personal Identification

Despite many studies of face perception and voice perception, no study comparing the effects of the two types of perception on person identification could be located. One study that did come close to fulfilling the requirements was done by Bartholomeus (1973) in which children were asked to identify their classmates and teachers by pointing to appropriate photographs while audio tapes of target voices were played. In this study, however, participants were already familiar with the target stimuli. Thus, this situation is not completely comparable to the more naturalistic situation in which participants are exposed briefly to a face or voice and must later recall the accompanying name. Even so, Bartholomeus concluded that participants found person identification from voice more difficult than person identification from face.

In studies outside of the memory field, researchers have experimentally compared processing of faces with processing of voices. Some found greater effectiveness for faces (Dittmann, Parloff, & Boomer, 1965; Gates, 1927; Zaidel & Mehrabian, 1969), some found less effectiveness for faces (Ekman, Friesen, OíSullivan, & Scherer, 1980), and some found results inconclusive (Kraus, Apple, Morency, Wenzel, & Winton, 1981; Mehrabian & Ferris, 1967). These studies concern decoding of face and voice stimuli rather than identification of specific persons by their faces and voices, but they do point to concerns that have relevance for person identification research.

Thus, in order to determine the relative accuracy of identification by face and by voice, and to investigate related methodological concerns, the following research questions are posed:

Research Question 1: Will perceivers identify stimulus individuals more accurately by face, voice, or a combination of face and voice?

Research Question 2: Will perceivers identify stimulus individuals more accurately with an impression formation induction or a recall induction?

Research Question 3: Will extremity of attribution correlate with increased identification accuracy?

Research Question 4: Will perceivers be more confident of their identifications in the face condition, the voice condition, or the combined face and voice condition?

Research Question 5: Will perceivers find identification more difficult in the face condition, the voice condition, or the combined face and voice condition?

Research Question 6: Will perceivers be more confident of their identifications in the recall induction or the impression formation induction?

Research Question 7: Will perceivers find identification more difficult in the recall induction or the impression formation induction?

Method

Individual color photographs of ten target individuals (5 males, 5 females) were enlarged and attached to poster boards. The first and last names of each individual were written in large block letters on the bottom fronts of each poster board. An additional set of these poster board photographs were created without labels to be used in the test portion of the experiment.

An audio recording was made of the same ten individuals. On this recording, each target individual recited the Pledge of Allegiance (selected because it was known by all individuals and eliminated the need to have them ëread" the test stimulus). Two experimental versions of the master tape were then created. One served as the stimulus tape and presented the target individuals in random order. On this tape, each individual said her or his own name and then recited the Pledge. Following each individualís recitation there was a three second pause. The second version of the master tape was used as the test version. On this version of the tape, target individualsí performances were arranged in an order different from that on the stimulus tape and an announcer said each individualís number which corresponded to the number given to them in the photograph test presentation.

During the test presentation of both photographs and audio tape, the researcher controlled the interval between presentation of each stimulus item in order to allow adequate time for participants to answer the test questions. This interval was, of necessity, longer than the three seconds allowed between items in the presentation of the original stimulus material. However, intervals between items were the same for photographs and audio tapes.

Participants (n=61) were drawn from undergraduate communication classes at a large southwestern university. All were given extra credit for their participation.

Procedure

Participants were tested in six groups of approximately 10 people each. In all groups, participants were first presented with the stimulus photographs and/or audio tape. Half of the participants were given an impression formation induction ("Try to form an impression of each person as you look at her or his picture") and half were given a recall induction ("Try to remember each person").

Two of the six groups were presented photographs only (face condition), two were presented audio tape only (voice condition), and two were presented photographs with audio tape (combined face and voice condition). In the face condition, the researcher showed each photograph for ten seconds (an amount of time determined comparable to the average amount of time it took a target individual to recite the Pledge of Allegiance), waited three seconds, then showed the next photograph. In the voice condition, the researcher played the first tape. In the combined face and voice condition, the researcher played the first tape while displaying the matching photographs.

Following the initial presentation, a test questionnaire was distributed and the second phase began. As the researcher presented the second set of stimulus materials, participants completed the questionnaires by writing each target personís name and rating each target person on eleven attributes. Attributes were presented as pairs of adjectives which participants rated from "1" at the low end of the continuum to "7" at the high end of the continuum. Descriptive pairs included "sad/happy," "unintelligent/intelligent," "submissive/dominant," "unfriendly/friendly," "dishonest/honest," "slender/heavy," "unattractive/attractive," "short/tall," "young/old," "unpleasant voice/pleasant voice," and "uneducated/educated."

For this second phase, in the face condition the researcher showed the second set of photographs. In the voice condition, the researcher played the second tape. In the combined face and voice condition, the researcher played the second tape and showed the accompanying photographs. Interval time between presentation of each stimulus item in the second phase was increased in all conditions to allow participants time to complete the questionnaire.

Finally, a post test was presented to determine participantsí reactions to the experiment and their assessment of their own performance on the test. Participants were asked to estimate the number of target individuals they believed they correctly identified providing a confidence measure. They were also asked to estimate both the difficulty of identifying target persons as well as the difficulty of determining the characteristics of target individuals. These estimations were determined on a Likert-type scale with "1" representing "not at all difficult" and "7" representing "very difficult."

Results

Data were analyzed with a 3 (channelñface, voice, combined face and voice) X methodñimpression formation, recall) ANOVA. The dependent measures were accuracy score (number of names identified correctly for each participant, with a full score given for identifying either the first or last name), confidence score (number of names participants believed they identified accurately), and difficulty score (participantsí ratings of difficulty of recall and attribution).

Results indicated no significant interaction between channel or induction method on any of the dependent measures. Thus, interpretation of the main effects can be considered.

The main effect of channel on accuracy of identification was found significant, F(2, 55)=3.61, p<.05, with means for these three conditions indicating that participants in the face condition were more accurate (4.48 names identified out of ten) than those in the combined condition (3.40 names identified out of ten), or the voice condition (2.85 names identified out of ten). Even so, one-way contrasts comparing the three groups failed to indicate any significant differences, thus providing only partial support for RQl.

Regarding induction method, a strong main effect was found, F(1, 55)=75.92, p<.01, with means indicating that the recall induction was superior (6.03 names identified out of ten) to the impression formation induction (1.22 names identified out of ten) in triggering recall. This finding contradicts most previous research, and suggests, in answer to RQ2, that straightforward recall is more effective than impression formation, at least in person identification.

An examination of the correlation matrix revealed no correlations between recall and the various attributions, which provides a negative answer to RQ3 which queried whether extremity of attribution would correlate with recall accuracy. That is, target individuals who received extremely high or extremely low attribution ratings were not remembered better by participants than were participants who received more moderate ratings.

Regarding RQ4 and RQ5, main effects for participant confidence were not found for channel, but were found for induction method, F(2, 55)=52.66, p<.01. Means for this variable on a scale of "1" for no confidence in identification to "7" for high confidence in identification were as follows: M of recall =5.3, and M of impression formation =l.l.

Regarding RQ6 and RQ7 no main effect for channel was found, but a strong main effect was found for induction method on difficulty of identification, F(1, 55)=33.91, p<01. Means for this variable measured on a scale of "1" for no difficulty in identification to "7" for high difficulty in identification were as follows: M of recall =4.5 and M of impression formation =6.61.

Discussion

This study attempted to determine the component parts of person identification. That is, how do perceivers use faces and voices to identify people? Previous research suggests that faces are remembered better than voices and this study partially supports those findings. As a matter of fact, in this experiment the face condition was even more effective than the combined face and voice condition in producing correct identification. It is possible that this occurred because the voices acted as a distraction to participants in the combined condition who focused on the visual stimulusñboth the photographs and the written questionnaire.

A variation implemented in this study that deviates from other person identification studies was the measurement of person recall. Whereas many previous studies have considered only recognition of faces and voices previously seen and heard, this study considered the more difficult but highly more realistic task of person identification. It is possible that some of the more unusual results can be attributed to this particular implementation.

Many laboratory studies have determined that participants remember target persons better if they form impressions of them rather than specifically try to recall them. This study found the opposite. Participants who were specifically asked to remember target persons performed significantly better in the identification tasks than those who formed impressions. As recall of proper names is a common social requirement, participants in the recall condition may have consciously attempted to remember the target individualsí names in anticipation of this expectation. This study did not find any correlation between extremity of attribution and increased identification accuracy. Possibly, the extremity of attribution may function to improve recall only during impression formation and not during directed recall. As those in the impression formation condition were less accurate in identification tasks than those in the directed recall condition, it is not particularly unusual not to find a correlation between accuracy and attribution extremity.

Participants in this study found the recall method was less difficult and made them more confident of their answers than the impression formation method. Possibly, participants find impression formation techniques appropriate for creating a general image of a person such as required in casual social situations, but in more important situations where the motivation is greater such as being introduced to a business contact or meeting a potential mate, the recall method may prove more effective. In these more important instances, we may consciously encode arbitrary and specific information such as a proper name if we believe it will be significant in later interactions.

Although great pains were taken in this experiment to maintain ecological validity, certain problems in this area can be recognized in retrospect. Participants were exposed to target individualsí names longer in the video condition than in the audio condition and were able to view each name as long as the researcher displayed the photograph. In the audio condition participants were exposed to target individualsí names only as long as it took the individuals to say their names. Despite this inconsistency, it may be argued that this presentation is more ecologically valid. For example, when seeing an individualís name written (e.g., on a job resume) one may be exposed to the name for a longer period of time than when one hears an individualís name spoken (e.g., when introduced to someone). However, other factors beside exposure time may come into play. The name may be difficult to pronounce or have an unusual spelling, in which case the audio presentation may be more helpful. On the other hand, the speaker may mispronounce or poorly articulate the name making the video presentation more effective.

The components of person memory have not been totally uncovered. There is much that goes into our ability to remember individuals in social contexts. Connecting names to faces and voices is one important part.

References

Response from Leonard J. Shedletsky

Anchoring the subtle nuances of memory in the real world experience of remembering a personís name is a wonderful teaching device, in addition to a valid practical and theoretical problem. My students typically get excited by the topic of recalling peopleís names. They want to know why it is so hard to recall the name of people theyíve just met. One might have thought that information, in addition to the name, was the distraction: face, clothing, expression, plus thoughts of self-consciousness and so on. Now, following Rockwellís research, one must wonder if focusing on the face isnít the way to go to increase name identification.

 

The Discourse Made Less Commonplace: A Mark of Discrimination

Philippe Castel and Marie-Françoise Lacassagne

About the Authors: Philippe Castel and Marie-Françoise Lacassagne, are on the faculty of G. R. P. (Groupe de Recherche sur la Parole) - Université Paris VIII. Marie-Françoise Lacassagne is at the Université Louis Pasteur, 12 rue Goethe, 67000 Strasbourg, France. Philippe Castel is at the Université of Bourgogne, 36 rue Chabot Charny, 2100 Dijon, France.

Abstract: This research intends to validate the role of the striking feature as a factor of discrimination. The presence of a non racial, but unavoidable feature (tattoo), produces a misappropriation of the same kind as that resulting from the confrontation with a target belonging to an obviously different ethnic group. The results show that the subjects questioned by an interviewer with a pattern on her or his cheek less easily mobilize the arguments generally more available than those addressing a "normal" interlocutor and that they appreciate the theme of the interview less. Note that in this study, the visually identifiable ancestry is discussed using the term "black" to refer to persons who appeared to have predominantly African ancestry and "white" to refer to persons who appeared to have predominantly French ancestry.

 

The general scope of the research covering this study postulates that discriminating behaviors result from the loss of the most available means that a subject undergoes when confronted with another subject bearing some striking feature. More precisely, the odd (abnormal) characteristic shown by the stranger becomes more important than the reason for the existence of the situation, and this results in a de-activation of the replies usually available.

A survey carried out in 1991 (Castel & Lacassagne, 1993) showed that in very standardized situations, the presence of a striking feature entailed a de-routinization of the sourceís discourse together with the discredit of the target.

In the experiment, 60 subjects were asked to write a formal letter (rejecting a candidate) to a person (a musician) who was introduced (by means of a photograph attached with the dossier) as being black, in a striking condition, or white, in a non-striking condition. The results first showed that when the addressee was black, there was a loss of routines both with regard to the long phases that compose a formal letter (e.g., presentation, introduction) and to the elements of the content mentioned (the arguments put forward), or finally to the respect of the closing ceremonial (closing formula). In addition, this de-routinization was selective since the omissions mainly concerned the elements which should have dealt tactfully with the susceptibility of the target. The proposed interpretation was that the obnubilation before a striking feature would lead to the misappropriation of the usual cognitive implements, the resulting "feeling of discomfort" tending to be cast off (exteriorization) on the bearer of the abnormal element. More precisely, the existence of a striking feature in the target would, because of its unavoidability, cause the source to have a feeling of uneasiness. Secondarily, the source would rationalize this condition in assigning negative parameters to the other: in other words, a discriminating process would set in, causing the passage from simply noticing a striking feature to attributing a stigma.

In this first experiment, characterized by the formalism of the discourse support, the misappropriation was therefore translated by a loss of the routines. The question one might ask oneself in the perspective of a generalization of this effect, is that concerning the marks of a misappropriation of this kind when the scope of the expression is informal.

If the "routine - non-routine" dimension characterizes the discourses having a reference model, the "commonplace-original" dimension seems to enable the formally free statements to be put in order. Thus, in certain situations, it would seem that the points raised are largely foreseeable. Certain themes, in particular those dealing with problems of society, refer to largely dominating representations to which correspond, on language level, lexical repertories. Within this scope, commonplaces are particularly pregnant. Platitudes indeed seem to form an informal equivalent to routines. Consequently the general hypothesis being tested is that a striking feature results in a decrease of the platitude of the discourse. 

Table of Results

STRIKING FEATURE (N = 16) ANOVA

discursive parametersnon-striking featurestriking featureF p

number of trivial substantives32.12514.50022.92.001

number of original substantives17.7510.503.52 n.s.

frequency of trivial substantives0.083770.058838.35.02

frequency of original substantives0.043780.42420.02 n.s.

triviality coefficient237.5097.0014.07.01

number of words391.500258.3755.35.05

frequency of references to oneself0.022540.0526013.54.01

Method

In the experiment submitted here, the aim was, on one hand, to refine the support to the striking feature (independent variable) and on the other, to widen the notion of de-routinization (dependent variable).

Independent Variable

It appears that the belonging to an ethnic group, dealt with in the previous experiment, cannot, due to dominating social representations, be exempt from valence. Thus, the noted marks of discrimination could be due to negative evocations, a priori associated with people of African decent and not result from the postulated stigmatization process. The improvement here consists in releasing the independent variable support from any aspect not strictly related to the striking feature

In this particular case, the striking feature was a clock painted on the cheek of the operator (playing the role of the target).

Dependent Variable

The notion of commonplace refers to the frequency with which themes appear. On one hand, however, the frequency is not established a priori (there are no tables of frequencies) and on the other, the operation of delimiting the theme is necessarily rather subjective. Consequently, in order to avoid these two drawbacks, the assessment of the platitude adopted here is based on the objective frequency of the terms used in the corpus studied. In order to identify the commonplaces, we considered all the substantives of the corpus and retained those which were used by at least three people. There is no ambiguity with this method and the results have not been submitted to an inter-judge validity test.

The mark of platitude of a subject will therefore correspond with the number of ordinary substantives she or he will use (including repetition).

Protocol

During 1992, some French women students in Psychology at the University of Burgundy, under the pretext of an investigation on mixed marriages, questioned a series of people chosen at random who were at a railway station waiting for a train. Their clothing and their behavior during the interview were as neutral as possible except for the striking feature (a clock painted on the cheek) which was present on one person out of two: thus, 16 non-directive interviews were carried out, 8 per experimental condition.

Operational Hypothesis

The average platitude of the discourse was weaker when the subjects spoke to a marked operator than when they spoke to a non-marked operator.

Results

The results were treated by variance analysis (ANOVA) when the distributions allowed it and by chi-square in the other cases.

Validation of The Hypothesis

A first measurement of the platitude can be made in counting the number of commonplaces. The number of trivial words indeed seems to differentiate the experimental conditions. The discourses in striking condition include half as many common substantives as the others and the result is significant (see table: number of trivial substantives).

However, the distribution of original substantives (i.e., substantives that only appear in one discourse) lead to certain questions. Here too, they are fewer in number in striking condition than those in non-striking condition. This result might lead one to think that the effect validated on the platitude would only result from the uniform reduction of productivity. This is not so, however, because the decrease of original substantives is not significant (see table: number of original substantives).

Moreover, if the untreated observations are brought down to the number of words in a discourse, the effect on the originality disappears (3%) while the effect on the platitudes is maintained (30%) and remains significant (see table: frequency of original substantives and frequency of the trivial substantives).

The difference of effects observed between originalities and platitudes leads to the elaboration of a sole coefficient taking into account the precise position of each substantive on an "original-trivial" scale. In concrete terms, we allotted a mark to each word corresponding to the number of persons (other than the locator in question) who used it. Thus, the term of "religion," mentioned by seven subjects will be given six points each time it is used, whereas the term "Catholic," used by a single person, will not be given any points at all. The triviality coefficient of each subject will therefore be composed of the sum of the marks of the formulated substantives.

In actual fact, the striking feature led to an important (60%) and significant decrease of this coefficient (see table: triviality coefficient).

These results therefore tended to confirm the hypothesis. The striking feature did indeed appear to make the discourse unique or less commonplace.

Other Results

The Reduction of Words

The first result to be observed, apart from those expected from the hypothesis, is a direct consequence of this: it is the productivity measured by the number of words. Indeed, if the subjects use fewer platitudes (without this reduction being compensated by an increase in the number of originalities) their discourse will necessarily be shorter.

Indeed, the discourses in striking condition include a third less words than those in non-striking condition and this difference is significant (see table: number of words).

The fact that the effect is less important (34% compared with 55%) and less significant (.05 compared with .001) supports the interpretation according to which the reduction of productivity is a consequence of the reduction of the platitudes. The alternative interpretation according to which the striking feature would produce a state of alienation marked by a reduction of generalized productivity (including platitudes) would imply different results: the effect on the productivity would be the most important, the effects on the platitudes and originalities would be less important .

Selective De-Trivialization

To the contrary of the experiment on de-routinization, where the subject had to assess the bearer of the striking feature, the subject of this experiment assessed an object ("mixed marriage") other than the bearer of the stimulus (the interviewer). In this situation, the negative valence resulting from the misappropriation caused by the striking feature should be applied to the theme of the discourse because it cannot be exteriorized on the cause itself.

In the same way as the results had showed a selective de-routinization consisting in "forgetting" the sequences favorable to the target, one might expect the underscored de-trivialization to also be selective in scotomizing the arguments favorable to mixed marriages.

In concrete terms, we have classified the arguments into three main categories :

If the distribution of negative and neutral arguments does not show any significant difference, it is not the same for the positive arguments: only one subject used positive arguments in striking condition whereas six of them did in non-striking condition: this difference is significant, chi-square (2, N = 16) = 15.28, p < .01.

Note, the only positive argument to have appeared in striking condition was a possible interest for an abstraction ("if the couple lives in Europe, this may be an advantage for the European culture"), whereas all the benefits mentioned in the non striking condition were beneficial to the persons concerned or their children.

The de-trivialization already underlined does not uniformly concern the various commonplaces linked with the theme. The subject, disorientated by the striking feature, does not use platitudes such as "itís an excuse for visiting other countries" or "you know more languages," whereas one does not forget "the condition of the woman" or "religious integrism."

This result undoubtedly corresponds with the expression of the negative mood of the subject confronted with the striking element represented by the picture of a clock. This interpretation is based on the observation according to which the operators registered 11 refusals to reply when they were marked with the clock, compared with only 5 when they were not.

Personalization

A final result enabled us to put forward an explanation on a psychological level of the effects previously observed, whether it was detrivialization or selectivity. It is the frequency of the references the speaker makes to oneself (number of "je," "me," "ma," "mon," "mes" ["I," "me," "my"]).

This use of the first person of the singular is more than twice as frequent in striking condition than in non-striking condition and this effect is significant (see table: frequency of references to oneself).

Everything happens as though, the subject, when confronted with the mark, was brought to differentiate oneself by speaking in oneís own name.

The distribution of the subjects having used the hyperbolic form "I, myself" reinforces this interpretation. Indeed, all the subjects in the striking condition use this form whereas only two of them use it in non-striking condition, although the discourses were much longer in the latter condition. This effect is significant, chi-square (2, N = 16) = 9,6, p < .01.

To sum up, fewer subjects in striking condition replied to the request for an interview and when they accepted, they spent less time answering questions; the arguments they "forgot" were the most positive; finally their discourse was hyper-personalized.

Discussion

In a situation of classical investigation, the interviewed subject behaves as a social agent adapting oneself to what one thinks is expected. Thus, in pre-election public opinion polls, the under estimation of votes in favor of parties on the edge of the political checker board (Communist Party) would result from the inference of a "taboo concerning the vote considered as extremist" (Matalon, 1992). While adapting, the subject mobilizes the routines required by the situation and uses commonplaces associated with the theme. The presence of a striking feature in the interviewer prevents this from happening. The subject, shocked by this unusual effect, can no longer behave as a simple agent. Faced with this image, which is not typical of a usual interviewerís appearance, one looses her or his social role. For the person, the picture of a clock represents an element that characterizes this other person and which encourages her or him to reply on the same level. So as to distinguish oneself, the person personalizes her or his discourse more, both in the content (reduction of the number of platitudes) and in the form (increase of the number of "I"). Nevertheless, the inadequacy between the inter-agent communication agreement (Ghiglione, 1985) (characterized by the superficiality and conventionality of the statements) and the need to differentiate, characteristic of the relationship between people, makes the subject uneasy, preventing her or him from using original or positive arguments. In other words, the interview situation calls the subject as representing any kind of category (e.g., average Frenchperson, engineer) whereas the striking feature mobilizes what she or he has as idiosyncratic. This apparent contradiction between stimulus and environment appears to be responsible for discursive perturbations.

Conclusion

The study of racist behavior seems to be progressing. The initial mechanism postulated was that the striking feature born by the target (ethnic feature) caused the loss of the most available answers (misappropriation). The first study had validated this mechanism on strongly ritualized behaviors (deroutinization) and brought another mechanism to the fore (i.e., that the consequence of the resulting state of misappropriation was an under-rating of the target). This study transposes the misappropriation mechanism to non pre-formed behaviors (detrivialization), reproduces the phenomenon of discredit and seems to bring out a differentiation between the source and the target (categorization).

With each of these mechanisms resulting from the same stimulus, it would seem possible to integrate them into a sole process that takes place in several stages. The first stage would be the differentiation directly produced by the unavoidability of the striking feature of the target. The second stage would be the misappropriation caused by the non availability of the usual implements enabling to manage the situation. Finally, the third stage generated by the state of alienation in which the subject finds herself or himself would be the discredit of the other. This process could therefore account for the self-favoritism characteristic of social categorization (Tajfel, 1978).

References

Response from Leonard J. Shedletsky

Castel and Lacassagne paint a picture of observable discourse as reflecting an internal framework, including categorization. This research suggests that our interpersonal behavior is highly sensitive to the intrapersonal domain. Some would refer to the process under study here as involving stereotypes. A question, outside of this study, that ought to be discussed is this: Is stereotyping necessarily a bad thing or even avoidable?

 

Intrapersonal Health Communication

Gary Meyer and Kim Witte

About the Authors: Gary Meyer is a doctoral candidate in the Department of Communication at Michigan State University. His research interests focus on health and environmental communication. He has engaged in research examining the effectiveness of community-based HIV/AID prevention campaigns and the diffusion of environmental technologies. Kim Witte (Ph.D., University of California, Irvine) is an assistant professor at Michigan State University. Her research examines the role of fear in health communication and has appeared in Communication Monographs, Communication Yearbook, Social Science and Medicine, International Quarterly of Community Health Education, International and Intercultural Communication Annual, and elsewhere.

Abstract: This article is designed to bridge the gap between intrapersonal communication and health communication. First, the authors review how individuals come to perceive risks, and how risk perceptions influence health-related decision-making. Then, the authors examine three important health decision models used in health promotion and disease prevention campaigns. Finally, they identify the commonalties between the risk communication literature and the three health decision models and integrate them into one overall intrapersonal health communication model.

Perspective:

1. What intrapersonal decision-making processes might a person go through making health-related decisions?

2. Based on the intrapersonal health communication model (Figure 4), how might a person convince her or his best friend to begin an exercise routine? Would the tactics differ if the person were trying to convince a best friend to quit smoking? Why or why not?

 

Intrapersonal communication research abounds in health communication, although it is not labeled as such. The vast amount of work on individualsí decision-making processes regarding health and health-related behaviors addresses issues of intrapersonal communication. For example, many health communication researchers ask questions like, how do people persuade themselves to adopt a healthy behavior? How do people estimate health-related risks? What kind of self-talk does an individual go through before deciding to use a condom, wear a bicycle helmet, or start an exercise program? All of these decisional processes occur within the individualsí mind, and as such may be termed intrapersonal communication.

The purpose of this chapter is to bridge the gap between intrapersonal communication and health communication, for they often study the same processes. First we will review how individuals perceive risks, and how risk perceptions influence health-related decision-making. Then, we will examine three key health decision models used in health promotion and disease prevention campaigns. Finally, we will identify the commonalties between the risk communication literature and the three health decision models and integrate them into one overall intrapersonal health communication model. This chapter should be of use to both theoreticians who seek to understand the intrapersonal communication processes occurring as individuals wrestle with decisions to choose or not choose health-related activities, as well as practitioners who need guidance about how to motivate individuals to promote health and prevent disease.

The Role of Risk in Health-Related Decision-Making

This story helps illustrate the importance risk plays in the individual decisions each one of us makes about health and health-related behaviors. The precise nature of risk, though, and its specific relationship to health-related decision-making has often been ignored or misunderstood. Here we discuss several components of risk that often are not well differentiated. We suggest that quantitative

Figure 1. Individual-Level Characterization Of Risk Perception

risk assessments along with qualitative or outrage factors, and the way in which each is communicated are all important determinants leading to individual risk perceptions which are extremely important in intrapersonal health communication processes.

Hazard versus Risk

The difference between a hazard and a risk is often unknown or confused. A hazard is an act or phenomenon posing potential harm to some person or thing (National Research Council, 1989), and the magnitude of a hazard may be described as the amount of harm that might result. Phenomena such as earthquakes, tornadoes, and floods are all easily recognizable natural hazards. Air and water pollution, nuclear weapons, and automobiles are all hazards developed by people. Each hazard, natural or manmade, has the potential to cause a certain level of magnitude of harm. The magnitude of an earthquake, for example, may be estimated in terms of potential disruption of normal activity, infrastructure damage, or deaths.

Risk of a hazard differs from magnitude of a hazard in that risk refers to the probability that the potential harm or undesirable consequence will be realized (National Research Council, 1989). Most would agree that the risk or probability of damage associated with an earthquake is different at the epicenter than it is 100 miles away regardless of its magnitude. Being hit by lightening is a very serious hazard with a high magnitude of harm, for example, but the probability of it happening to an individual is fairly small. As such, while the hazard remains the same, the probability of harm and individual-level risk varies.

Risk Assessment

The first important component shaping risk perceptions and intrapersonal risk decision-making is risk assessment. Risk assessment is the estimation of adverse effects stemming from exposure to hazards (National Research Council, 1989). In other words, it is simply how likely a given hazard is to effect an individual. Risk assessment determinations are typically conducted by governmental or industrial scientists and are intended to inform the public about risks they face. Quantitative analytical techniques are used to assess risk and the end result is usually thought to be a factual determination. Risk assessments are generally characterized by a numerical representation often expressed as annual fatalities (e.g., "The risk of dying from the bite of a venomous creature in the United Kingdom is 1 in 5 million") (National Research Council, 1989). Stated as such, the probability appears definite, but because of incomplete knowledge and various assumptions and estimations made during calculation, each of which is a source of uncertainty, the accuracy ascribed to the estimate is limited. Risk assessments can be made more accurate by aligning assumptions with the most up-to-date technical literature (National Research Council, 1983). Although scientifically and technically based, it should be noted that within any risk assessment certain value judgments are made which influence calculations about ultimate risk. These value judgments represent intrapersonal health communication processes by risk assessors.

Risk Communication

The way risk is communicated to individuals also plays a part in shaping individual risk perception. Since becoming a popular area of research, risk communication has been defined in numerous ways and has come to mean different things to different people (Plough & Krimsky, 1987). Risk communication has typically been defined in terms of a one-way flow of information, with a message traveling from scientific experts to the lay public. Here, governmental experts would tell the lay public what risks they face as well as how to perceive and manage them. Intrapersonal decision-making was, for all practical purposes, eliminated. A relatively new model has emerged which defines the process of risk communication as "an interactive process of exchange of information and opinion among individuals, groups, and institutions" (National Research Council, 1989, p. 21). This new conceptualization (1) serves to differentiate risk messages from risk communication, the former which are a necessary subset of the latter and (2) brings risk perception and decision-making back to the individual level.

The process of risk communication historically has been frustrating for all participants involved. Scientists lament that the lay public does not base risk management decisions on facts (technical risk assessments) communicated to them. Individuals, for example, may refuse to eat certain foods because of chemical ingredients but continue to smoke two packs of cigarettes a day. At the same time, the lay public suggests that experts communicate information that (1) is jargonistic, technical, and uninterpretable and (2) has little to do with individual-level decision-making. Most people, for example, have difficulty comprehending the significance of 2.1 parts per billion (ppb) of trichloroethylene in their water. Such characterizations are complicated further when compared to other risks, which is a common practice (Covello, 1991). What does it mean, for example, to compare 2.1 ppb trichloroethylene with driving 50,000 miles per year in an automobile, eating red meat daily, or living in Los Angeles? As a result, individuals often leave the risk communication process more confused than ever, wondering exactly how the risk applies to them, how much risk they face, and what they should do about it.

Risk Perception

Risk perceptions, defined as those judgments an individual makes about the relative danger associated with a particular hazard, play a defining role in intrapersonal decision-making . Difficulties between experts and the lay public stem in large part from the way in

which different people perceive risks (Slovic, 1987). Studies show that while individuals are capable of understanding quantitatively-based risk assessments (i.e., 1 in 2 million will die from being struck by lightening) and do consider them in part, (Chess, Hance, & Sandman, 1988), individual judgments of risk perception are primarily based on qualitative dimensions of hazard (National Research Council, 1989). One study conducted by the Environmental Protection Agency illustrated, for example, that the publicís ranking of health and environmental risks did not correlate well with the EPAís ranking of the same risks (Environmental Protection Agency, 1987). Chess, Hance, and Sandman (1988) refer to these non-technical or qualitative factors as outrage factors because they create outrage in individuals. Depending on the outrage factors present, certain types of risk are much more readily accepted by the public than others. Risks that are voluntary, fair, natural, and familiar, for example, are more acceptable than those which are imposed, unfair, artificial, and exotic. The public will accept risks 1000 times as great from voluntary activities such as skiing than from involuntary hazards such as food preservatives (Slovic, 1987).

The fact that most people base their perceptions of risk primarily on outrage factors is complicated by the fact that one of the primary sources of information about risks is the mass media (Fischhoff, 1985). This is problematic because the mass media rely on images to convey a story and tend to focus on "dread" risks which rank high on outrage and lead people to perceive greater risk (Wilkins and Patterson, 1987). Issues that receive a great deal of media coverage tend to be rated by the public as very risky whereas those risks that do not receive much media coverage tend to be seen as less risky. Individuals, therefore, might have a low perception of risk about cigarette smoking (risk is old, familiar, controllable, and doesnít get much coverage in the mass media), but a high perception of the risk about AIDS (risk is relatively new, exotic, unfamiliar, and gets a good deal of coverage within the mass media).

Risk Management

Individual-level risk perception plays a major role in how risks are ultimately managed. Risk management may be defined as the process of evaluating alternative risk control actions, selecting among them (including doing nothing), and implementing the decision (National Research Council, 1989). Broadly, risk management involves coping with a risk, either actual or objective, that impacts one or more persons or groups, sometimes differentially. Risk management decisions are made at the individual level, within organizations, and even by the federal government. Individuals make daily risk management decisions about what types of food to eat, exercise to engage in, and means of transportation to use. Organizations make decisions about the risk level of certain products or degree of risk employees are subjected to during production, and governmental units make risk management decisions, for example, by setting standards on various types and levels of pollution.

Figure 1 summarizes the intrapersonal decision-making processes about risk-related issues as presented here. In Figure 1, each of the three clusters represents the unique combination of factors leading to an individualís perception of any given risk. These factors include the quantitative assessment of a hazard, the outrage or qualitative dimension of the hazard, and the means by which each is communicated to an individual. As a group, the clusters represent the perceptions of different risks which an individual compares against the others when making a risk management decision. Importantly, and especially relevant to intrapersonal communication research, risks will be characterized differently both within and across individuals. The relative emphasis an individual places on a quantitative or qualitative aspect and the way in which the individual heard about a risk will likely differ for any risk. Quantitative risk assessments may be especially important in some cases, the outrage dimension of the hazard may be important in other cases, and the means by which either is communicated may play a greater or lesser role depending on the risk. Across individuals, hazards will be even more distinct given the likelihood that hazards were communicated to them in different ways through different sources.

Intrapersonal Decision-Making

To illustrate the intrapersonal decision-making processes about risk-related issues presented here, consider the hypothetical situation where an individual has recently been diagnosed with a rare disease and must decide what to do next. His options, along with doing nothing, are as follows:

Option 1 . The first option is to have an operation which has been used successfully over the past 15 years and has a high rate of success (90% chance of survival) and an estimated average future life span of seven years.

Option 2 . The second option is to have an operation which has also been used successfully for several years and has roughly the same estimated future life span. This procedure, however, on average, requires a hospital stay and full recovery period which is half as long as that in Option 1.

Option 3 . The third option is to have an operation which has rarely been performed. This procedure was recently developed with the latest medical technology. There is not enough information currently available to reliably estimate the chances of surviving the operation, but best estimates are around 75%. Estimated future life span is anywhere from 10 to 20 years.

These options might be considered by this particular individual in the following way. The quantitative characterization of the hazards associated with Options 1 and 2 seem similar and slightly less than Option 3. He has read several articles about each in highly regarded medical journals and feels confident about the risk assessments characterized in each. As such, the way the hazards were communicated is important only to the extent that all are believable, but less important insofar as each of the assessments was communicated in a similar fashion.

His outrage associated with the third option is the highest of the three because this is a new and exotic risk which uses technologies unfamiliar to him. His outrage level associated with the first option is small, and somewhat greater for the second option. He has recently seen a television news show which graphically illustrated potential complications with Option 2. It seems that while the recovery period is shorter, it is often extremely painful. Aside from what he saw on television, the only other information came from a co-worker who mentioned that her distant cousin had the second procedure without too much pain. The way in which the information about the second option was communicated is important because the explicit pictures of people in great pain increased his level of outrage greatly. The interpersonal communication from his co-worker might have little effect, because he did not know the cousin personally.

Based on the information provided, we might expect our patient to select Option 1, and if intrapersonal decision-making about health and health-related behaviors was made based solely on those factors presented here, he likely would. But the risk management process, consists of other factors as well, many of which are incorporated into the health models reviewed later in chapter. This particular risk management decision, for example, may also include a financial cost or benefit analysis (e.g., "Is one procedure much more costly than another?" "Will insurance cover each of these options equally?"), consideration of the relative availability of services (e.g., "Is one procedure more readily available than another?" "Is the new procedure (Option 3) available locally?"), and psychological considerations (e.g., "Can I endure extreme pain associated with the second option?" "How long can my family cope without me while I am recovering?"). If Option 1 was the least costly but unavailable for a full year, this individual may select one of the other options, perhaps Option 2. On the other hand, if his insurance company will pay for hospitalization only up to a certain time period (which can be virtually guaranteed by Option 2, but not Option 1) and he did not have sufficient funds to make up any shortfall, then we might expect him to also select the second option rather than the first.

Risk Perception and Models of Health Behavior

Understanding the way in which both quantitative and qualitative factors are communicated to form the basis for individual risk perception is especially important given the major role risk perception plays in health decision models. The important factors in intrapersonal risk perception processes, as described above, include (a) the quantitative assessment of a hazard, (b) the outrage or qualitative dimension, and (c) the means by which each is communicated to an individual. In the following health decision models, risk (probability of occurrence) is often referred to as perceived susceptibility and the magnitude of risk is viewed as severity of the threat or hazard.

Health Decision Models

Three major theoretical approaches dominate much of the intrapersonal health communication research conducted today. The first is the Health Belief Model, a model generated in the 1950s to explain why people do or do not adopt preventive health behaviors. The second is Social Learning Theory, a model that focuses on peopleís perceived ability to perform health promotion and disease prevention activities. And, the third is the Extended Parallel Process Model, which is not so much a single model as it is one that integrates many previous theoretical perspectives in order to explain the intrapersonal processes individuals go through when faced with a health threat. Following is a description of each model.

The Health Belief Model

The Health Belief Model (HBM) is probably the premier health decision model taught to students of public health, and thousands have

Figure 2. Basic Elements of the Health Belief Model

used it to guide the development of health campaigns and interventions. It also was among the first health decision models developed (by researchers in the Yale Communication Research program back in the 1950s). It has had an enormous influence on health communication research.

According to the Health Belief Model (HBM) (Figure 2), preventive health behaviors (e.g., getting immunizations, eating healthy, wearing seatbelts) are influenced by five factors:(a) perceived barriers to performing the recommended response (e.g., beliefs about psychological, physical, or financial "costs"); (b) perceived benefits of performing the recommended response (e.g., beliefs about rewards such as risk reduction, injury prevention, anxiety reduction); (c) perceived susceptibility (e.g., subjective probability of experiencing a hazard); (d) perceived severity (e.g., beliefs about the seriousness of the hazard); and (e) cues to action (e.g., advertisements).

The "combined levels of susceptibility and severity" are said to provide "the energy or force to act and the perception of benefits (less barriers)" are said to provide "a preferred path of action" (Rosenstock, 1974, p. 332). Cues to action stimulate decision-making processes leading to self-protective behaviors. Cues to action can occur internally, as when a persistent cough motivates one to quit smoking, or externally, as when mass media messages highlight the risks of smoking.

According to the HBM, cues such as interpersonal interactions or mass media campaigns must first increase perceptions of susceptibility and severity. For example, when a commercial warns teens that they might get HIV if they have sex, or when a physician warns a patient that she or he has high blood pressure and is at-risk for a stroke, then perceived susceptibility and severity are heightened. When perceived susceptibility and severity are high, then people become motivated to adopt healthy practices. At the same time, people must believe there are lots of benefits (i.e., reduce risk of disease contraction, feel better) and few barriers (i.e., little cost, easy to do) to adopting the recommended response before they will do what an advertisement or physician advocates. For example, the HBM would propose that if individuals feel vulnerable to HIV-contraction (the virus that causes AIDS) and believe AIDS to be a serious or severe disease, and they believe there are benefits to using condoms (i.e., reduce their risk of contraction) and there are few barriers to using condoms (i.e., easy to get, easy to use), then they would use condoms to prevent HIV-contraction.

Therefore, intrapersonal communication processes are key in the HBM. Cues (i.e., mass media campaigns, physician warnings) are the outside influence that motivate action, but they do not directly elicit self-protective behaviors. Rather, cues start and may influence the self-talk where one evaluates whether she or he is at-risk for a given hazard or disease (i.e., susceptibility), whether the given hazard or disease is serious (severity), whether there are significant benefits to adopting the recommended response, and whether the barriers are minimal enough to where the individual is able to adopt the recommended response.

The HBM has received much support since its inception and is the most frequently used model in health education or intervention studies (Janz & Becker, 1984; Leventhal, Zimmerman, & Gutmann, 1984). The HBM offers much heuristic value in conceptualizing how intrapersonal communication processes work in response to cues.

Social Cognitive Theory

Albert Banduraís Social Cognitive Theory (sometimes called Social Learning Theory) has been used in a wide variety of campaigns. It was the theory used in the Stanford 5-Cities project to prevent heart disease and more recently has been used in several AIDS-prevention projects. The focal point of the theory is on perceived self-efficacy. Self-efficacy is defined as "peopleís beliefs that they can exert control over their motivation and behavior and over their social environment" (Bandura, 1977, p. 128). In other words, perceived self-efficacy is what you believe about your capability to perform a certain action (your perceived self-effectiveness). Bandura (1977) views self-efficacy as the driving force of human behavior. "Efficacy expectations are a major determinant of peopleís choice of activities, how much effort they will expend, and of how long they will sustain effort in dealing with stressful situations" (Bandura, 1977, p. 194).

One other important construct in Banduraís theory is outcome expectations. Outcome expectations refer to an individualís belief that a certain behavior will lead to a certain outcome. For example, "I believe that if I get a polio vaccination, I wonít get polio" is an outcome expectation. Itís what you think will happen if you take a certain action. Outcome expectations are different from efficacy expectations in that the latter is a personís belief on whether she or he is able to "successfully execute the behavior required to produce the outcomes" (Bandura, 1977, p. 193). For example, even if outcome expectations are high (e.g., "If I get a polio shot I wonít get polio"), efficacy expectations may be low (e.g., "but Iím not capable of getting a polio shot because needles scare me"). In short, according to social cognitive theory, a person can believe that certain actions lead to a particular outcome (outcome expectations), but she or he may doubt her or his ability to perform the action (efficacy expectations). Thus, it is the self-efficacy perception that causes behavior.

Bandura (1977) states that an individualís self-efficacy perceptions are developed from four sources of information: Performance accomplishments, physiological states, verbal persuasion, and vicarious experience. Participant modeling or performance desensitization are examples of performance accomplishments. Suggestion, exhortation, and self-instruction are what constitute verbal persuasion. Vicarious experience stems from live or symbolic modeling. Finally, physiological or emotional arousal comes from relaxation, biofeedback, and symbolic desensitization or exposure.

According to Bandura (1977), only when efficacy expectations are high will people perform certain behaviors. Efficacy expectations can vary on dimensions of magnitude (level of difficulty of task; people may have different efficacy expectations for simple tasks than for difficult tasks), generality (specific to general), and strength (weak to strong) (Bandura, 1977). Overall, perceived efficacy is one of the most important intrapersonal level variables in health communication research.

 

 

 

        External

        Stimuli



Message Processing Outcomes

(1st & 2nd Appraisals)



Process

 

 

 

 

 

Danger

Control

Process

 

 

 

 

Fear

Control

Process

Figure 3. Extended Parallel Process Model

The Extended Parallel Process Model

The EPPM is a "fear appeal" theory. Fear appeals are persuasive messages that scare an audience into adopting the recommended response. The first part of a fear appeal focuses on the health threat by emphasizing the severity of the threat (i.e., its magnitude of harm) and the probability that the threat will occur (i.e., the audienceís likelihood of experiencing that threat). Fear is aroused when a threat is perceived as relevant and significant. The second section of a fear appeal focuses on the efficacy (e.g., effectiveness) of the recommended response (a) in averting or minimizing the threat (response efficacy) and (b) it attempts to increase perceived self-efficacy by increasing oneís perceived ability to perform the recommended response by outlining specific and easy steps to avert the threat (self-efficacy). The fear appeal may be thought of as the "cue to action" in the health belief model sense.

The research into fear appeals has shown them to be potent persuasive devices, but only in certain conditions. The intrapersonal communication processes occurring in response to fear appeals have been studied for over 40 years. The Extended Parallel Process Model (EPPM) identifies two types of intrapersonal communication processes and specifies what kind of self-talk occurs in response to health threats or scary messages.

The Extended Parallel Process Model (EPPM) is based on Leventhalís danger control or fear control framework and is an expansion of previous fear appeal theoretical approaches (Janis, 1967; Leventhal, 1970; Rogers, 1975, 1983). Readers will note similarities between the Health Belief Model and fear appeal theories . Fear appeal models can be thought of as experimental variants or explanatory versions of the HBM. Clearly, fear appeal research and the EPPM grew out of the HBM. In fact, most of the early fear appeal researchers came out of Yale (as did the HBM).

According to the EPPM (see Figure 3), the evaluation of a health threat initiates two self-talk processes, which result in either danger control (i.e., cognitive processes) or fear control processes (i.e., emotional processes). First, persons appraise the threat of the hazard by determining whether they think the threat is serious (e.g., "skin cancer can be deadly") and whether they think they are susceptible to the threat (e.g., "Iím at-risk for getting skin cancer). The greater the threat perceived, the more motivated individuals are to begin the second appraisal, which is an evaluation of the efficacy of the recommended response. When people think about the recommended response, they evaluate its level of response efficacy (e.g., "Does sunscreen prevent skin cancer?") and their level of self-efficacy (e.g., "Am I capable of using sunscreen?"). When the threat is regarded as trivial or irrelevant (perceived as low), there is no motivation to consider the threat further; the efficacy of the recommended response is evaluated superficiallyñif it is evaluated at allñand no response is made to the health threat. If people do not feel at-risk or do not feel the threat to be significant, they simply will ignore it.

Danger control processes will dominate and people will act in ways that prevent the threat as long as perceptions of efficacy are greater than perceptions of threat (e.g., "I know that skin cancer is a terrible threat that Iím at-risk for, but if I have my physician screen me annually, Iíll be able to prevent my dying from it"). In other words, when danger control processes dominate, individuals are motivated to control the danger by thinking of ways to protect themselves. Danger control processes are primarily cognitive processes where individuals (a) realize they are at risk for a severe danger (high threat) and become motivated to protect themselves, (b) believe they can effectively deter the threat (high efficacy), and (c) deliberately and cognitively confront the danger (e.g., "Iím going to call my physician right now and request a skin cancer screening"). The cognitions or thoughts occurring in the danger control processes elicit protection motivation, which stimulates adaptive actions such as attitude, intention, or behavior changes that control the danger.

At some critical point, however, when persons realize that they cannot prevent a serious threat from occurring, either because they believe the response to be ineffective or because they believe they are incapable of performing the recommended response (e.g., "skin cancer is a terrible disease that Iíll probably get and I donít think thereís anything I can do to keep from dying from it"), fear control processes will begin to dominate over danger control processes. Fear control processes are primarily emotional processes where people respond to and cope with their fear, not to the danger. Defensive motivation is elicited by heightened fear arousal, which occurs when perceived threat is high and perceived efficacy is low, and produces responses that control oneís fear such as defensive avoidance or reactance. Studies have shown that fear appeals with high levels of threat (e.g., "skin cancer is a severe disease that you are susceptible to given your history of sun exposure") and low levels of efficacy (e.g., "sunscreen used now canít undo past skin damage due to overexposure from the sun, therefore you cannot totally prevent skin cancer") result in message rejection, and occasionally in boomerang effects (people do the opposite of what is advocated).

In short, according to the EPPM when people are faced with a health threat, they talk themselves into controlling the danger (i.e., the actual health threat) or controlling their fear about the danger. They weigh their risk of actually experiencing the health threat (i.e., getting cancer) against actions they can take that would minimize or avert the health threat (i.e., does sunscreen really work? can I really use it every time Iím in the sun?). The intrapersonal communication processes in the EPPM occur at both the cognitive (i.e., thinking) and emotional (i.e., feeling) levels. Emotion is an important variable in intrapersonal health communication research that deserves more attention.

Integrating Intrapersonal Health Communication Perspectives

As one evaluates the risk and health decision model literature, several common themes emerge. In this section we identify those common themes and integrate the research into one overall intrapersonal health communication framework. What do each of the theoretical perspectives outlined have in common?

First, all of the perspectives focus on cues to action. The risk communication literature outlines cues (risk messages) emanating from a variety of sources (e.g., doctors, industrialists, government officials) often through the mass media. The Health Belief Model outlines external (messages) and internal (self-talk, physiological signals) cues. Social Cognitive Theory addresses ways to increase self-efficacy (e.g., vicarious learning, persuasive messages, physiological changes). The fear appeal is the cue in the Extended Parallel Process Model. Thus, it appears that most if not all intrapersonal health communication will occur in response to either an external or internal cue.

Second, most of the perspectives emphasize a health threat. Risk communication research has typically focused on environmental and technological threats to physical health, while the health decision models focus more on physical illness or disease. Theoretically, the health decision models combined suggest that individuals must feel susceptible to a severe threat if they are to adopt healthy behaviors. Without a perception of threat, people are not motivated to protect themselves against disease. The risk perception literature outlines the intrapersonal communication processes leading to perception of susceptibility and severity.

Third, two of the health decision models discuss efficacy issues regarding the recommended response. (The risk communication literature focuses more on the health threat and how people respond to that threat, but it does not address perceived efficacy issues). There appear to be two distinct types of efficacy addressed in these models. In Social Cognitive Theory (SCT), self-efficacy and outcome expectations are highlighted. In the Extended Parallel Process Model (EPPM), self-efficacy and response efficacy are addressed. It appears that self-efficacy in both SCT and the EPPM are identical. Thus, self-efficacy is a key concept that must be included in any intrapersonal health communication model.

One other type of efficacy seems to be identified by each health decision model. Namely, SCTís outcome expectations and the EPPMís response efficacy concepts all appear to be similar. All refer to oneís beliefs about what will happen if the recommended response is performed. More specifically, outcome expectations relate to perceptions about whether a certain behavior will lead to a certain outcome, while response efficacy relates to perceptions about whether the recommended response is effective in averting a threat. Combined, this variable might be called recommended response efficacy.

Fourth, barriers appears to be an important concept deserving more attention, although the HBM is the only one to explicitly name the construct. Barriers are not addressed in SCT except perhaps in a passing manner (i.e., there may be certain barriers that prevent the performance of a behavior). In the EPPM, self-efficacy is broadly defined to include the concepts of barriers. Specifically, anything that influences oneís perceived ability to perform a given actionñeither positively or negativelyñis seen as influencing self-efficacy. However, the constructs seem to be conceptually distinct and address different issues in that self-efficacy refers to oneís perceived ability to do something, while barriers are those variables that prevent one from engaging in an actionñregardless of whether you think you can do it or not. Thus, self-efficacy and barriers should each be addressed separately in an intrapersonal health communication model.

Overall, an intrapersonal health communication model must include a threat component, two efficacy components (recommended response efficacy and self-efficacy), and a barrier component. Now that the important intrapersonal health communication processes have been identified, we need to explain how they work together. The model presented here is theoretical only in that it has yet to be tested. However, it offers a valuable integration of dominant health communication perspectives. Figure 4 shows the intrapersonal health communication model based on the previous review of the literature.

Cues about health activate the intrapersonal communication processes. As suggested by the Health Belief Model, cues can be external (e.g., advertisements, PSAs, clinic visits) or internal (e.g., feeling sick, feeling upset). In short, when people are presented with cues that alert them to a health threat, they converse with themselves about their susceptibility to and the significance of the health threat, about potential responses that could avert the health threat, whether they are capable of doing the recommended response, and about barriers to performing those recommended actions.

Thus, cues must first increase perceptions of threat about a health hazard before people will be motivated to act. That is, people must feel at-risk (i.e., susceptible) for a severe health threat before they will think about efficacy and barriers. If individuals feel susceptible to a significant threat then they will evaluate self-efficacy, recommended response efficacy, and barriers to recommended responses. If perceived self-efficacy and recommended response efficacy are at high levels (i.e., people feel capable and able to perform an effective recommended response to avert the health threat), and perceived barriers to performing a recommended action are seen as low, then individuals will act in a healthy manner and engage in disease prevention and health promotion activities. It is important to note that the self-talk that occurs regarding threat, efficacy, and barriers, may not be of a complex cognitive type. Rather, recent research suggests that many intrapersonal

An Intrapersonal Health Communication Model

 

Cues

External

(mass media)

Internal

(anxiety)

Intrapersonal Communication Processes

PerceivedPerceivedPerceivedPerceived

ThreatSelf-EfficacyRecommendedBarriers

Efficacy

        Behaviors

        Health Promotive

        Behaviors

            (exercise)

        Disease Preventive Behaviors

(vaccinations)

Figure 4. Intrapersonal Health Communication Model

 

health communication processes occur "mindlessly" or without complex cognitive thought (Witte, 1994). The exact manner in which intrapersonal communication processes operate

when faced with health threats needs much more research before firm conclusions can be drawn.

One important issue about the intrapersonal health communication model must be addressed. It is absolutely critical that if a health threat is emphasized, then so must self-efficacy and recommended response efficacy be emphasized, and barriers must be minimized. (See Witte, 1992a & b, for a complete exposition on this topic.) Increased perceptions of threat motivate action, but to get adaptive action that promotes health and prevents disease, individuals must have high levels of self-efficacy and recommended response efficacy, as well as perceive minimal barriers to the recommended action. If perceived threat is high but the barriers to preventing the health threat appear insurmountable, then individuals will deny the health threat and perhaps even put themselves at greater health risks. Thus, it is crucial to have high levels of perceived self-efficacy and recommended response efficacy, and low levels of perceived barriers, to accompany perceptions of threat regarding a health issue.

In sum, the review of risk literature and the dominant health communication models suggests that cues affect intrapersonal communication processes (i.e., evaluation of threat, self-efficacy, recommended response efficacy, and barriers). If intrapersonal communication processes result in high levels of perceived threat, self-efficacy, and recommended response efficacy, and low levels of perceived barriers, then individuals will engage in health promotion and disease prevention activities. Great care must be taken to avoid unanticipated harmful outcomes by making sure perceived efficacy is high enough, and perceived barriers are low enough, to balance perceived threat.

Intrapersonal communication is an important and exciting area within health communication and the communication field as a whole. In this chapter we took an in-depth look at factors that influence individualsí perception of risk and reviewed three health decision models. Continued research in the area of intrapersonal health communication will not only serve to advance the communication discipline in important ways, but will also be of tremendous pragmatic import to health practitioners in their on-going efforts to prevent disease and promote health.

References

Response from Leonard J. Shedletsky

Witte and Meyer make the case for merging theory and practice. Moreover, this article represents a grand example of just how attention to mediating variables promises to increase our understanding of communication. Stimuli become messages when they are assigned meaning by people. Having said that, it is likely to be tricky to determine just how health communication can take individual differences into account.

 

Willingness To Communicate, The Need for Cognition and Innovativeness: New Zealand Students and Professionals

Craig Johnson, Bruce Dixon, Michael Z. Hackman, and Larry Vinson

About the Authors: Craig Johnson is on the faculty of George Fox College. Bruce Dixon is on the faculty of the University of Waikato, New Zealand. Michael Z. Hackman is on the faculty of the University of Colorado-Colorado Springs. Larry Vinson is on the faculty of McNeese State University. An earlier version of this article was presented at the 1993 Speech Communication Association convention, Miami, Florida.

Abstract: This study examined the relationship between innovativeness, willingness to communicate, and need for cognition among New Zealand student and professional groups. Positive correlations were found for all groups between all three traits. Lawyersí innovativeness, WTC and NFC scores were significantly higher than those of students and accountants. The discussion centers on the communication implications of these findings.

Perspective:

1. Are some groups (e.g., students, professors, construction workers, lawyers) more willing to talk, more willing to think, and more open to change than other groups? Why or why not?

2. What is the possible relationship between an individualís attitude toward communication and thinking and her or his willingness to try out new ideas?

3. What can managers do to encourage employees to become more open to change?

 

Individual attitudes towards change play a significant role in either the acceptance or rejection of innovations in the organizational setting. Members predisposed to resist change can slow or halt the spread of new ideas while innovators adopt and promote new products and practices. Managers and other organizational change agents have a vested interest in hiring and promoting innovative people. Isolating traits associated with innovativeness makes it easier to identify such individuals. The purpose of this study was to determine if the willingness to communicate and the need for cognition are related to openness to change. Possible correlations between these traits were tested and contrasts were drawn between student and professional groups.

Innovativeness, Willingness to Communicate and the

Need for Cognition

Goldsmith (1986) calls innovativeness an "elusive construct" because it has been defined and operationalized in many different ways. Some researchers treat innovativeness as the motivation to produce new ideas (Jackson, 1976), while others consider it to be a form of risk taking (Cancion, 1967) or a problem solving style (Kirton, 1976). For those interested in identifying innovative individuals in the organizational setting, perhaps the most useful definition is provided by Hurt, Joseph and Cook (1977). They define innovativeness as the willingness to change which contributes either to the adoption or rejection of new ideas.

A personís willingness to talk with others may be related to her or his openness to new things. Innovating is very much a communication activity. To learn of new ideas, innovators must be in contact with others outside their immediate circle of associates. Diffusing innovations throughout organizations requires coalition building and other collaborative efforts. Idea sponsors must share information, trade resources, form task forces, lead meetings and engage in other communication behaviors in order to convince others to "buy in" to the innovation (Kanter, 1988). Those with a positive orientation towards interaction presumably should be more willing to engage in innovative activity.

An individualís underlying, personality-based, trait-like predisposition toward communication has been labeled as "Willingness to Communicate" (WTC) (McCroskey & Richmond, 1987, 1991). This orientation is consistent across varying communication contexts and receivers. Those who are willing to communicate speak more often and for longer periods of time. This increased verbal activity generates a variety of positive outcomes. High WTC sources are more likely to take leadership positions and are perceived as more attractive, credible and sociable. In organizational settings such individuals are more likely to be hired, retained and promoted (Richmond & Roach, 1992). One of the sources of low WTC is fear or apprehension about communication (McCroskey & Richmond, 1991). In a study of the relationship between innovativeness and communication apprehension, Witteman (1976) found that innovativeness scores increased as communication apprehension decreased. His data suggest that high WTC scores (which reflect low apprehension) should be positively linked to innovativeness.

The enjoyment of cognitive tasks may also be related to innovativeness. Need for Cognition (NFC) refers to the degree to which individuals either seek out or avoid cognitive effort (Cacioppo & Petty, 1982; Cacioppo, Petty, & Kao, 1984). Those who enjoy thinking are motivated to elaborate on (pay attention to, think closely about) messages. Those with a low need for cognition tend to avoid demanding cognitive tasks. Innovative behavior appears to require more cognitive effort than noninnovativeness. Innovators must attend to new ideas, think through their implications, evaluate their value and determine how to make modifications in the existing organizational system. Thus it seems likely that innovativeness will increase with the need for cognition.

Sanders, Gass, Wiseman and Bruschke (1992) found a positive correlation between the need for cognition and argumentativenessñ"the tendency to recognize controversial issues in communication situations, to present ideas and defend positions on the issues, and to attack the positions that other people take" (Infante & Rancer, 1982, p. 164). Simply put, those who liked to think were also more likely to argue. The link between the need for cognition and one form of verbal activity suggests that there may be a relationship between cognitive elaboration and the general tendency to interact across all situations. If this is the case, then WTC and NFC scores should demonstrate a positive correlation.

Innovativeness Traits and Professional and Cultural Differences

In addition to identifying innovative individuals, organizational leaders may want to determine if particular constituencies are more or less resistant to change. Professional workers, for example, play a major role in most organizations. To determine if groups of professionals differ in their predisposition to innovate, communicate and think critically, accountants and lawyers were sampled along with students.

Scores on instruments measuring personality-linked communication variables often vary significantly between cultures. Investigators report that the willingness to communicate scores of USA subjects are higher than those of Swedes, Australians and Micronesians (McCroskey & Richmond, 1990). Asian Americans demonstrate a lower need for cognition than Euro-American or Hispanic Americans (Sanders et al., 1992). Residents of New Zealand were tested in this project. In an earlier comparison of US and New Zealand university students, Hackman and Barthel-Hackman (1993) found that New Zealanders were significantly less willing to communicate than US residents and experienced greater communication apprehension.

Method

Subjects

Participants in the present study were from three categories: university students, accountants, and lawyers. The student sample consisted of 179 undergraduate students enrolled in introductory communication and management courses at a New Zealand university. Students completed questionnaires during regular class time. All students were given the option not to participate in the study. Five hundred accountants and 500 lawyers, chosen at random from the directories of New Zealand professional associations, were mailed copies of the questionnaire. Usable questionnaires were returned by 218 accountants (a response rate of 44%) and 190 lawyers (a response rate of 38%). Questionnaires were completed by all participants without personal identification to insure anonymity and to increase the likelihood of honest responses.

Measures

The following self-report measures were used in this study:

The Innovativeness Scale (IS). The IS (Hurt et al., 1977) assesses the extent to which individuals perceive themselves as open to change. The IS consists of 12 positively (i.e., "I find it stimulating to be original in my thinking and behavior") and 8 negatively worded items ("I often find myself skeptical of new ideas"). A 7-point Likert-type response format is used. Hurt et al. (1977) reported split-half internal reliability for the IS at .94. Goldsmith (1986) reported a reliability coefficient of .89. In the present study, the internal reliability (Cronbach Alpha) for the IS was .90.

Table 1

Correlations Among Measures for Students, Accountants, and Lawyers

WTCISNFC

WTC

Studentsñ-.33*.20*

Accountantsñ-.25*.19*

Lawyersñ-.30*.21*

All groupsñ-.30*.21*

IS

Studentsñ-.46*

Accountantsñ-.64*

Lawyersñ-.48*

All groupsñ-.54*

NFC

Studentsñ-

Accountantsñ-

Lawyersñ-

All groupsñ-

*p <.01.

Willingness to Communicate (WTC). The WTC scale (McCroskey & Richmond, 1987) measures the predisposition toward approaching or avoiding the initiation of communication. The WTC is a 20-item probability estimate scale made up of 12 items, which comprise the measure, and eight items which are fillers. The 12 items on the scale assess willingness to communicate in four contexts (public speaking, meeting, group, and dyad) and with three types of receivers (stranger, acquaintance, and friend). McCroskey (1992) reports various studies using the WTC that have found estimates of internal reliability ranging from .86 to .95 with a modal estimate of .92. In the present study, the internal reliability (Cronbach Alpha) for the total scale was .87.

Need for Cognition (NFC). The NFC scale (Cacioppo & Petty, 1982) assesses the extent to which individuals like to engage in complex thought. A shortened 18-item version of the NFC was used (Cacioppo et al., 1984). Participants indicated their agreement (strongly agree = 5, strongly disagree = 1) with each statement (i.e., "I find satisfaction in deliberating hard and for long hours;" "the notion of thinking abstractly is appealing to me"). In the present study, the internal reliability (Cronbach Alpha) for the NFC was .89.

Data Analyses

Pearson correlations were computed to estimate the relationships among the scales. Means scores were computed for each of the scales and subscales administered to the student, accountant, and lawyer sample groups. Differences between the means were tested for significance with a series of one-way ANOVAS.

Results

Innovativeness was positively correlated with both willingness to communicate and need for cognition across all sample groups. WTC and NFC scores were also found to correlate positively across sample groups (see Table 1).

Differences between sample groups were revealed by the analyses of variance. As reflected in Table 2, lawyersí perceptions of their innovativeness were significantly greater than students or accountants. In addition, lawyers perceived themselves as more willing to communicate than accountants or students. A significantly greater inclination to communicate on

Table 2

Summary of Analyses of Variance

Measure

A. Student

Mean

B. Accountant

Mean

C. Lawyer

Mean

F

Sig.

IS

 

WTC

Public

Meeting

Group

Dyad

Stranger

Acquaintance

Friend

 

NFC

94.2ac*

 

60.8ac*

48.6ac*

57.4ac*

68.5ac*

68.8ab*

34.5ac*

66.9ac*

81.0

 

61.9ac*

94.8bc*

 

59.6bc*

50.1bc*

56.2bc*

68.1bc*

64.4bc*

36.6bc*

65.5bc*

76.9bc*

 

63.1

98.5

 

67.8

60.5

65.6

75.8

69.4

46.9

73.7

83.1

 

65.1

4.27

 

12.13

10.03

10.00

10.13

5.38

16.92

8.41

6.78

 

4.61

0.15

 

.001

.001

.001

.001

.005

.001

.001

.002

 

.011

*Means with matching subscripts are significantly different from one another: p. <.05.

the part of lawyers was identified in all contexts and with all types of receivers when comparing lawyers and accountants. When compared to students, lawyers viewed themselves as more willing to communicate in each context other than the dyadic and with all types of receivers other than friends. The only significant difference between students and accountants was in the dyadic context. In dyads, students rated themselves as significantly more likely to communicate than accountants. The highest need for cognition was reported by lawyers. Lawyerís NFC ratings were significantly higher than those of the students and accountants.

Discussion

All three of the traits tested in this studyñinnovativeness, willingness to communicate and the need for cognitionñwere positively related. Those who are more willing to engage in innovative behavior report that they are also more willing to communicate with others and to seek out challenging cognitive tasks. In addition, those who enjoy communicating are also more likely to enjoy thinking critically.

The positive link between IS and WTC scores suggests that we can gain important insights into a personís openness to change by examining her or his openness to communication. If attitudes towards change and communication are connected, then developing communication skills may be one way to encourage innovativeness. People who believe that they possess inadequate communication skills are less willing to interact with others (Richmond & Roach, 1992). Helping these individuals overcome their deficiencies makes it more likely that they will engage in innovative activity which requires interaction.

The strongest correlation was between innovativeness and the need for cognition. Innovative activity apparently attracts those who want to engage in extended thought. In fact, willingness to change and the inclination to think critically may be overlapping constructs. Some questions in the IS scale ("I am challenged by ambiguities and unsolved problems;" "I am challenged by unanswered questions") are similar to items in the NFC instrument. However, the magnitude of the correlation between IS and NFC scores was not great enough to indicate that the two scales measure the same dimension.

The need for cognition is also related to willingness to communicate. Sanders et al. (1992) reported that those with a high need for cognition were more likely to engage in argumentative behavior. The results of this study suggest that NFC is tied to other forms of interaction as well. The WTC instrument measures the inclination to communicate in public speaking, meeting, group and dyadic settings. Participants declared that they were more likely to communicate in all of these settings and with various types of receivers when they also had a strong need to think through issues.

Of the three groups sampled, lawyers reported that they were most innovative, communicative and eager to think. There are two possible explanations for this finding. One is that lawyers may not be any more innovative, talkative or thoughtful than other people but just perceive that they are. This explanation reinforces the popular stereotype that characterizes lawyers as individuals with inflated egos. The second and more likely explanation has to do with the connection between IS, WTC and NFC scores. Intuitively it would appear that the legal profession attracts people high in willingness to communicate. Lawyers spend much of their time talking to clients, interviewing witnesses, delivering arguments in court and so on. Those low in WTC would probably avoid this occupation. If lawyers are attracted to the profession because of its communication demands and if WTC is positively related to innovativeness and the need for cognition, then it follows that lawyers would score high in all three areas.

The significant findings reported here open up promising avenues for future research. First, other communication related traits such as communicator style and argumentativeness could be examined to determine their relationship to innovativeness. Second, other professional groups should be compared to determine if they differ in their self-perceptions of their openness to change. Data could then be collected to determine if those individuals and groups with the highest test scores actually engage in the most innovative activity. Third, additional tests of the relationship between innovativeness, willingness to communicate and the need for cognition ought to be conducted with other cultural groups. Earlier we noted that test scores on personality-linked communication variables differ significantly between cultural groups. The relationship between these traits also varies between cultures. McCroskey and Richmond (1990) report, for example, that there is a much stronger correlation between perceptions of communication competence and willingness to communicate among Micronesians than among residents of other countries. Hackman and Barthel-Hackman (1993) found negative correlations between communication apprehension and humor variables (liking of humor, sensitivity to humor, and coping humor) for US students and positive correlations among these same variables for NZ respondents. Unlike the US students, the NZ students with the most developed senses of humor were less likely to initiate interaction. Before we can say with certainty that innovativeness, WTC and NFC are related, the results of this study must be replicated in other cultures.

References

Response from Leonard J. Shedletsky

Again, we are witness to an instance where looking inside the black box (the mind) helps us to understand observable behavior. Along with Johnson, Dixon, Hackman, and Vinson, this reader is inclined to believe that such variables as innovativeness, WTC and NFC are associated with significant life choices, such as occupation. Evidence from learning styles research suggests the same conclusion. If this is true, it begins to build the case for intrapersonal communication study as simultaneously an applied and theoretical discipline.

 

Perceived Understanding, Nonverbal Communication, and Relational Satisfaction

Kori L. Egland, Michael A. Stelzner, Peter A. Andersen, and Brian H. Spitzberg

About the Authors: Kori L. Egland, Michael A. Stelzner, Peter A. Andersen, and Brian Spitzberg are at the School of Communication at San Diego State University.

Abstract: The purpose of this study was to examine the relationship between perceived understanding, nonverbal messages, and relational satisfaction in heterosexual relationships. Findings indicated that perceived understanding was positively related to increases in relational satisfaction. It was further found that both back-channel communication and immediacy contributed to positive perceptions of understanding, as well as higher levels of relational satisfaction. In addition, specific nonverbal behaviors were identified that lead to increases in perceived understanding and increases in relational satisfaction.

Perspective:

1. In the Western tradition, we have generally believed verbal communication to be the source of "understanding" among people. What role does nonverbal communication play in everyday understanding?

2. Under what conditions is understanding another person likely to be unsatisfying? Why do we generally assume that understanding is a satisfying experience?

3. Does mutual understanding become more or less important to a relationship as the relationship becomes more intimate over time? Why?

 

A middle aged couple sat comfortably in their living room, each reading a lengthy novel. With no words, the woman turned to her husband. Mr. Smith, casting a glance towards his wife, paused for a moment. Nodding his head in agreement, yet saying nothing, he rose from his chair and grasped his jacket. Clutching Mrs. Smithís hand, they departed for a stroll in the park.

Although the husband and wife are fictional characters, they represent a fairly common experience of many satisfied couples. Specifically, understanding can be communicated with little verbal communication. How does perceived understanding contribute to a coupleís success and satisfaction? How is this understanding perceived? Is it, perhaps, the nonverbal messages we receive that confirm this perception of understanding? Hence, the purpose of this article is to examine the relationship between perceived understanding, nonverbal messages, and relational satisfaction in heterosexual relationships. This study will briefly examine the history of understanding research and its relationship to relational satisfaction. The paper will further examine nonverbal communication, including back-channeling and immediacy behaviors and their relationship to perceived understanding.

Generally speaking, research on understanding evolved from the investigation of research on listening. In an attempt to develop a more standardized definition of listening, Glenn (1989) examined more than 50 definitions dating as far back as the mid-1920ís. Glenn concluded that understanding (sometimes also labeled perception) was the most widely accepted descriptive component of listening. Understanding can be approached from a number of perspectives including, but not limited to, simple auditory comprehension, meaning analysis, intelligence analysis, or compassion measures. The aim of earlier research seemed to focus on basic comprehension. Research conducted prior to 1960 was primarily concerned with studying the basic listening and hearing skills that impacted learning and comprehension. Interestingly, interpersonal variables were virtually excluded in listening research until the late 1960ís.

More recently, research has connected the constructs of listening and relational satisfaction. Several earlier studies concluded that listening was related in some way to relational satisfaction (Fretz, 1966; Ivey, Normington, Miller, Morrill, & Hasse, 1968; Royce & Weiss, 1975). Later, Garland (1981) also found that attending behaviors, a component of active listening, increased marital satisfaction. It may be that understanding, a major component of listening, is a more specific variable leading to relational satisfaction.

Research on Understanding

Past research on understanding is fairly limited. However, in one study, Laing, Phillipson, and Lee (1966) examined understanding through the concept of metaperspectives, meaning oneís view of how others view you. They describe that:

Laing and colleagues (1966) further explained that a positive metaperspective conveys a sense of togetherness. A few years later, Luft (1969) explained that "perhaps it is true that unless (one) derives from human interaction the special quality of feeling understood, he will suffer the despair of meaninglessness no matter what he achieves and how well he understands everything else" (p. 145). Our perception of being listened to and our perception of being understood is crucial to relational satisfaction.

It was not until the 1980ís that studies began to explore the concept of perceived understanding, defined as an individualís assessment of her or his success when attempting to communicate with another (Cahn, 1990; Cahn & Shulman, 1984). Cahn and Frey (1990) explained that if "one feels understood, one tends to undertake or continue sending messages based on the belief that she or he has tapped into the appropriate set of rules governing human interaction for the situation [or that person]" (p. 2). In other words, if we are in a communicative interaction and we believe the other person is understanding, receiving, or perceiving, judgments of the relationship, the interaction tends to be positive and communication proceeds.

The phenomenon of perceived understanding has gained respectability since several measures have been recently developed and empirical findings accumulated. Specifically, perceived understanding has been measured and investigated in teacher-student relationships (Cahn, 1984), superior-subordinate relationships (Cahn, 1986), and romantic relationships (Cahn, 1987, 1989, 1990). Cahn and Frey (1989), contended that "perceived understanding is an important variable that affects a variety of perceptual processes and interpersonal relationships" (p. 1299). When a listener is perceived to be responding with understanding, she or he is viewed as cooperative, patient, and modest (Cahn & Frey, 1989). On the other hand, when a listener is not perceived as understanding, she or he is viewed as uncooperative, impatient, and condescending (Cahn & Frey, 1989).

In developing relationships, perceived understanding is positively associated with trust and attraction (Cahn & Frey, 1989). As a relationship matures, perceived understanding becomes most important relative to other variables (Cahn, 1987, 1989; Cahn & Frey, 1990). Cahn (1990) illustrates that increases in relational understanding are associated with:

Thus, measuring the level of relational development is a crucial element when testing for perceived understanding. In essence, perceived understanding becomes so important that it can perpetuate growth or deterioration of a relationship (Cahn 1987, 1990). Cahn (1990) found that partners who feel more understood are significantly happier in their relationships than those who feel less understood. Earlier research supports this conclusion. For instance, Allen and Thompson (1984) found perceived understanding to be related to marital satisfaction. Their study tested predictors of communicative satisfaction in marital relationships by examining individual and dyadic understanding, perceived feelings, and the realization of understanding. Findings indicated that agreement and perceived understanding were indicators of marital satisfaction.

Understanding of oneís partner is imperative for a healthy relationship. Without perceived understanding, a relationship is hard to maintain. The more perceived understanding that partners experience, the greater the probability that the relationship will develop into a stable, satisfactory relationship.

The next important question is what are the cues, if any, that signal understanding? Cahn and Frey (1990) suggest that future research should attempt to separate verbal and nonverbal behavioral cues. A number of researchers have suggested that nonverbal communication is critical in the exchange of meaning (See Burgoon, Buller, and Woodall, 1989 for a summary). Because nonverbal communication seemingly plays such an important role in understanding, it is important to review nonverbal research associated with understanding and satisfaction.

 

Nonverbal Research

To interpret a speakerís internal state, we must evaluate the nonverbal as well as verbal messages. Thus, communicators are specifically interested in the perception that the listener is understanding through information provided by the nonverbal channel.

Cahn and Frey (1992) examined perceived verbal and nonverbal behaviors that were associated with perceived understanding and misunderstanding. Findings indicated that feelings of being understood were regulated by the perceived nonverbal behaviors of smiling, direct eye contact, and the appearance of being relaxed. Feelings of misunderstanding consisted of five perceived verbal behaviors (erred in rephrasing, summarized incorrectly, failed to answer questions correctly, expressed rejection of person, expressed criticism, and verbally expressed disinterest) and eight perceived nonverbal behaviors (shuffled feet, tapped feet, turned away, appeared tense/rigid, frowned, blank stare, stammered/hesitated, and raised voice/yelled/screamed). Cahn and Frey (1992) explain that "people probably pay more attention to their conversational partnersí perceived verbal and nonverbal behaviors that they associate with the negative feelings of being misunderstood than those behaviors that they associate with the positive feeling of being understood" (p. 1063). To summarize, listeners tend to judge perceived understanding from nonverbal more than from verbal behavior, and perceived misunderstanding causes a relational partner to be more sensitive to their partnerís behaviors.

Hecht and Marston (1985) conducted a study on conversational time and satisfaction in long-term relationships. They contended that one criterion of satisfaction in long-term dyadic conversations was based on signs of understanding. In other words, satisfaction was positively related to perceived cues of understanding in more developed relationships. Hecht and Marston (1985) found that these signs are displayed mainly through nonverbal channels.

Given that the speaker judges perceived understanding more from nonverbal behavior than from verbal behavior, one can logically conclude that a listenerís nonverbal behavior is salient to a speaker. Essentially, a listenerís nonverbal behavior (i.e., head-nodís) can signal to the speaker whether or not understanding is being conveyed. This response by the listener occurs simultaneously, for the most part, and in rhythm with the speakerís dialogue. This entire process of conveying a response to the speaker is known as back-channel communication. Previous research referred to similar phenomena as conventional signals (Fries, 1952) and accompaniment signals (Kendon, 1967). Yngve (1970) was the first scholar to use the term back-channel. He defined back-channel as the channel "over which the person who has the turn receives short messages such as ëyesí and ëuh-huhí without relinquishing the turn" (p. 568). Later, Dittmann (1972) used the term listener responses in the same manner that Yngve used the back-channel. Dittmann described listener responses "as specific signals that the listener is paying attention to the speaker, is keeping up with him, or that he has understood what was just said" (1972, p. 405).

Duncan and Fiske (1977) categorized back-channels into five types: m-hm, sentence completion, requests for clarification, brief restatement, and head nods and shakes. It is important to note however, that Duncan and Fiske (1977) show back-channel communication consisting of both nonverbal and verbal responses. Moreover, Krauss, Garlock, Bricker, and McMahon (1977) found the nonverbal back-channel performed an equally important function as the verbal back-channel. To better illustrate the use of nonverbal back-channel communication, recall the opening example of Mr. and Mrs. Smith. Mr. Smith utilized a head nod back-channel to signal his understanding of his wifeís request. Through little or no verbal communication, understanding was successfully conveyed. Nonverbal behavior has an important role in back-channel communication. For example, statements such as "yes-yes," or "uh huh" are often accompanied or replaced by repeated head nods (Yngve, 1970). In addition to head nods, back-channels can be comprised of gestures and facial expressions (Krauss, Garlock, Bricker, & McMahon, 1977). Smiles are a common source of back-channel communication. It has been found that, smiling often perpetuates feelings or thoughts of being understood (Brunner, 1979). In an earlier study, Dittmann and Llewellyn (1968) also found a brief smile when used with a head nod tended to signal listening.

Many situations occur in relationships that obstruct the back-channel. For example, Yngve (1970) illustrates:

In this situation, direct eye contact, a few head nods, and possibly a smile or two would have been a more productive means of conveying understanding. Therefore, it is plausible that a lack of back-channel communication, in either substance or quality, can reduce relational satisfaction.

There appears to be several purposes or goals of the back-channel. Back-channels serve as markers or units in social interaction and they fill a need for the listener to respond without interrupting (Dittmann & Llewellyn, 1968). Another goal, according to Yngve (1970), is to monitor the quality of the communication. The back-channel also tends to signal successful communication from the speaker to the listener (Dittmann, 1972), communicate attention and interest (Dittmann, 1972; Dittmann & Llewellyn, 1968; Yngve, 1970), and signal the listenerís success at keeping up with the conversation (Dittmann, 1972).

Brunner (1979) expressed three additional ideas about what back-channels convey. First, the back-channel signals involvement and participation, and it recognizes that a conversation is occurring in the dyad. Second, the back-channel identifies the level of understanding by the listener, which in turn allows the speaker to alter the communication to effectively convey ideas. Finally, the back-channel identifies the personal response of the listener, which could range anywhere from agreement to scorn, or to a number of alternative reactions.

Shrout and Fiske (1981) found the back-channel to account for a relationship between evaluation and nonverbal behavior. They explained that we evaluate an interaction through the nonverbal cues received from the back-channel. Thus, perceived understanding, an evaluative process, also utilizes the nonverbal back-channel to confirm understanding.

In addition to back-channeling, another nonverbal construct, immediacy, may have a powerful impact on perceived understanding and relational satisfaction. Immediacy is the degree of liking and approach between communicators (Mehrabian, 1971). It has both a verbal and a nonverbal component. An example of verbal immediacy includes the use of phrases such as "we." Nonverbal immediacy is often communicated through behaviors such as touch, gaze, and positive facial expressions.

In an analysis of nonverbal immediacy behaviors, Andersen (1985) proposed four definitional components of nonverbal immediacy. First, immediacy behaviors are approach behaviors. Second, they signal the availability for communication. Third, immediacy increases sensory stimulation, either psychologically or physiologically. Finally, immediacy behaviors communicate warmth and closeness. Furthermore, Andersen (1985) identified six main components of nonverbal immediacy: proxemics, haptics, kinesics, oculesics, vocalics, and chronemics.

First, proxemics, or the use of interpersonal space, include closer physical distance, face-to-face body orientation, interactions at the same physical plane, and forward leans. Second, haptics, also known as tactile communication, involves high degrees of physical contact and touch. Third, kinesics, or communication through body movement, include smiles, head nods, gestures, body relaxation, and open body positions. Fourth, oculesics, which is also known as messages sent by the eyes, include eye contact, gaze, and pupil dilation. Fifth, vocalics, or nonverbal messages in a person's voice, include mm-hmmís, pleasantness of voice, vocal clarity, and fluency. Finally, chronemics, or meanings of time in interpersonal relationships, include spending time with someone. If immediacy communicates warmth, closeness, and availability for communication, it should help communicate understanding and increase relational satisfaction for couples.

To summarize, we are postulating that the perception of being understood is an important antecedent of relational satisfaction in interpersonal dyads, and it is through the nonverbal back-channel and immediacy cues that the strongest perceptions are received. A major goal of this research is to examine nonverbal back-channel and immediacy behaviors, and their relationship with perceived understanding and relational satisfaction.

The following hypotheses represent an attempt to relate perceived understanding, relational satisfaction, immediacy, and back-channel communication. The first two hypotheses replicate previous research whereas the next three hypotheses extend our knowledge of the manner in which perceived understanding is communicated and relational satisfaction is attained.

Hypothesis 1: Perceived understanding is positively related to relational satisfaction.

Hypothesis 2: Perceptions of nonverbal immediacy behaviors are positively related to relational satisfaction.

 

Table 1

Correlation of Nonverbal Behaviors to Perceived Understanding and Relational Satisfaction

Perceived Relational

Nonverbal BehaviorUnderstanding Satisfaction

Eye Contact (+) .4276** .4962**

Body Relaxation (-) .1056 .1609*

Gestures (+) .0967 .2489**

Touch (-) .1773* .3201**

Body Relaxation (+) .1479 .1534

Head Nods (+)-.1208-.0886

Body Direction (+) .4583** .4029**

Smiles (+) .3649** .3566**

Eye Contact (-) .4775** .4273**

Spending Time (+) .4127** .4267**

Touch (+) .2742** .4501**

Distance (-) .3676** .4969**

Body Direction (-) .3427** .2398**

Gestures (-) .1289 .1466

Smiles (-) .3890** .2846**

Distance (+) .1250 .0875

Head Nods (-) .0749 .0443

Spending Time (-) .4987** .5196**

Perceived Understanding .5412**

Relational Satisfaction .5412**

Immediacy .5521** .6009**

Back-channel Comm. .4436** .4211**

* Significant LE .05 ** Significant LE .01 (2-tailed)

(+) positive test, (-) negative test recoded to positive

Hypothesis 3: Perceptions of nonverbal immediacy behaviors are positively related to perceived understanding.

Hypothesis 4: Perceptions of appropriate back-channel communication are positively related to increased relational satisfaction.

Hypothesis 5: Perceptions of appropriate back-channel communication are positively related to perceived understanding.

Finally, we were interested in assessing which nonverbal behaviors had the most significant relationship with understanding and satisfaction.

Research Question 1: Which nonverbal behaviors best predict perceived understanding?

Research Question 2: Which nonverbal behaviors best predict relational satisfaction?

Methodology

Subjects

Subjects consisted of 153 students of diverse majors at a large southwestern university. Of the subjects, 39 % were male and 61 % female. The ages of subjects ranged from 24 years of age or younger (72 %), 25 to 35 (16 %), 36 to 46 (seven %), 47 to 57 (two %), and 58 or older (three %). All subjects were involved in one of four possible heterosexual relationship types: casual dating (26 %), serious dating (34 %), marriage bound (18 %) or married (22 %). The length of the subject's relationships ranged from days, but less than one week (one %), weeks, but less than one month (eight %), months, but less than one year (28 %), years, but less than one decade (48 %) and a decade or longer (14 %).

Procedures

A 45-question survey was distributed to 153 undergraduate students, representing a wide range of majors, enrolled in university required speech communication courses. Questionnaires were filled out in the classroom. Subjects were informed that they needed to be involved in a heterosexual relationship to participate. Subjects were instructed further that there was no obligation to complete the survey and all surveys would be handled anonymously. Questionnaires were completed in an average of 10 minutes.

The questionnaire began with two demographic questions identifying participants age and gender (see Appendix A). The next two questions inquired about the relationship level (i.e., casual dating, serious dating, marriage bound, or married) and relationship length (i.e., days, weeks, months, years, a decade or more), which were attained from Andersen and Guerreroís (1990) relational intimacy measure.

The first series of questions consisted of Cahn and Shulmanís (1984) perceived understanding instrument (see Appendix B). In this section, subjects were asked to indicate the extent to which a number of terms described how they generally felt when and immediately after trying to make themselves understood to their partner. This instrument used eight questions to assess perceived understanding and eight questions to assess perceived misunderstanding. Answers for perceived misunderstanding were recoded to be equivalent with perceived understanding. In the present study, Cahn and Shulmanís (1984) measure produced a coefficient alpha of .89 for perceived understanding, .85 for the perceived misunderstanding, and .91 for combined scales. Given this level of internal consistency the items were combined for all analyses.

The second section of the questionnaire measured relational satisfaction using a slightly modified version of Hendrickís (1988) Relationship Assessment Scale (RAS) (see Appendix C). Questions were modified so that a response in agreement would yield a rating positively related to relational satisfaction. The questions were simple five-point Likert scale questions dealing with love, satisfaction, and conflict. The RAS is a highly reliable, yet simple to administer test, because it consists of only seven questions. In addition, the scale does not limit itself to married couples and thus serves a good measurement for a student subject population consisting mostly of unmarried couples. In the current study, Hendrickís (1988) relational satisfaction scale produced a coefficient alpha of .87.

The final sections of the questionnaire pertained to nonverbal behavior. Questions pertaining to immediacy and back-channel communication were interspersed. Perceptions of nonverbal immediacy were measured with Andersen, Andersen, and Jensenís (1979) behavior indicants of immediacy scale (see Appendix D). The original scale consisted of 28 Likert items referring to multiple nonverbal behaviors. We removed questions that were not related to this study, resulting in a 16-item immediacy scale that achieved a coefficient alpha of .83 in the current study.

There were six items about back-channel communication. Four of the questions (two relating to eye contact and two relating to smiles) were encompassed in both the immediacy and back-channel scales. Two questions about head-nods were unique to the back-channel scale. The six-item back-channel measurement achieved a marginal coefficient alpha of .69.

Results

All hypotheses (p<.05)were confirmed. Hypothesis 1, that perceived understanding is positively related to increased relational satisfaction, was confirmed (r=.54, r2=.29). Hypothesis 2, that perceptions of nonverbal immediacy behaviors are positively related to relational satisfaction, was confirmed (r=.60, r2=.36). Hypothesis 3, that perceptions of nonverbal immediacy behaviors are positively related to perceived understanding, was confirmed (r=.55, r2=.30). Hypothesis 4, that perceptions of back-channel communication are positively related to increased relational satisfaction, was confirmed (r=.42, r2=.18). Finally, hypothesis 5, that perceptions of back-channel communication are positively related to perceived understanding, was confirmed (r=.44, r2=.19, see Table 1). T-tests revealed that there were no gender differences for perceived understanding, relational satisfaction, immediacy, or back-channel communication.

The research questions were tested by correlating the dependent variables of perceived understanding and relational satisfaction with each independent nonverbal behavior (see Table 1). Nonverbal behaviors that represented the negative or absence of the behavior were recoded to represent the positive or presence of the behavior, and recoded variables are signified in the table with a (-) sign.

Several nonverbal behaviors were significantly related to perceived understanding, as posed in research question 1. The four most important nonverbal characteristics that led to perceived understanding were: spending time talking with one's partner (.50), engaging in eye contact (.48), directing body position towards partner (.46), and touching (.45). These variables had a moderately large effect size, each explaining 20 to 25 % of the variance.

The three most important nonverbal characteristics that predicted relational satisfaction were: spending time talking with a person's partner (r=.52), physical proximity or distance (r=.50), and engaging in eye contact (r=.50). These variables also had a moderately large effect size, each explaining 25 to 27 % of the variance.

Due to the large effect sizes of both perceived understanding and relational satisfaction in relation to nonverbal behaviors, a multiple regression was performed. It was found that for perceived understanding, spending time talking with one's partners, engaging in eye contact, and directing body position towards oneís partner, combined explain 36 % of the variance (a strong effect). For relational satisfaction, spending time talking with one's partner, physical proximity or distance, engaging in eye contact, and the use of gestures, combined to explain 43 % of the variance, also a strong effect.

Discussion

The purpose of this study was to examine the relationship between perceived understanding, nonverbal messages, and relational satisfaction in heterosexual relationships. Furthermore, our goal was to address research questions regarding the role of nonverbal communication in perceived understanding and relational satisfaction. All five hypotheses were confirmed and both research questions produced significant results.

Perceived understanding again was found to have a positive relationship to increased relational satisfaction. It has been further found that both immediacy behaviors and back-channel communication contributed to higher levels of perceived understanding. It followed that immediacy behaviors and back-channel communication also contributed to greater levels of relational satisfaction. Similar to Hecht and Marstonís (1985) study, satisfaction was positively related to cues of understanding in more developed relationships. We also found the same result for relationships regardless of length.

Two research questions were proposed: (a) which nonverbal behaviors increase perceived understanding, and (b) which nonverbal behaviors increase relational satisfaction? Findings addressed both questions adequately.

Numerous nonverbal behaviors were found to have a direct relationship to increases in perceived understanding. Specifically, it was found that spending time talking with one's partner, engaging in eye contact, and directing body position towards oneís partner, combined, had strong effects on perceived understanding. Spending time with oneís partner had the largest influence on perceived understanding (r2=.25). Andersen (1985) explains that spending time with someone communicates immediacy and indicates interest. In todayís society, it is not uncommon for couples to find it more difficult to spend time together. It may be that simply taking time out of oneís busy day positively affects perceptions of understanding. Like Cahn and Frey (1992), the present study found eye contact was STET (fairly strongly) related to perceived understanding (r2=.23). Andersen (1985) explained that eye contact is "an invitation to communicate" (p. 8), so these results seem to confirm the theoretical effects of eye contact. Direct body position also was found to be an important nonverbal behavior in fostering perceived understanding (r2=.21). Direct body position may contribute to perceived understanding because it signifies that oneís attention is focused primarily on the speaker, rather than some other person or activity. Therefore, if one spends time with a partner, looks the person in eye when being spoken to, and directs the body toward the partner, more positive perceptions of understanding should result. These three variables, spending time with oneís partner, engaging in eye contact, and direct body positions each independently accounted for 21 to 25 percent of the variance while all immediacy behaviors accounted for only 30 percent of the variance in understanding. It appears that these three variables may be particularly important components of immediacy in generating perceptions of perceived understanding.

Relational satisfaction was found to be associated with similar nonverbal behaviors. It was found that for relational satisfaction, spending time talking with one's partner (r2=.27), close physical proximity or distance from oneís partner (r2=.25), engaging in eye contact (r2=.23), and touch (r2=.20) showed fairly strong associations with relational satisfaction. Given these strong associations, these four immediacy behaviors are evidently critical components of immediacy in producing relational satisfaction.

Two nonverbal behaviors were highly associated with both perceived understanding and relational satisfaction (spending time and eye contact). This suggests a relationship between these nonverbal cues of perceived understanding and relational satisfaction. Physical proximity has been shown to be positively related to attraction and liking (Andersen, 1985). It follows that closeness will also perpetuate relational satisfaction. Therefore, as people spend time with their partner, move physically close, engage in eye contact, and touch their partner, higher levels of satisfaction are likely to result.

These findings could impact future research from a number of different perspectives. First, our findings may be applicable to long distance relationship research. Communication over the telephone does qualify as spending time with one's partner, other nonverbal behaviors are absent. Eye contact and directing oneís body towards oneís partner, two major behaviors associated with perceived understanding, are virtually impossible to accomplish. In addition, moving closer to one's partner and using gestures, two behaviors associated with relational satisfaction, would also be difficult, if not impossible to accomplish. Perhaps with the implementation of video phones, the use of multimedia, and the development of virtual reality, we will approach a world in which such nonverbal behaviors will be able to be communicated live over long distances. Currently, such technology is accessible by very few people; therefore, long distance relationships seem likely to fail given our findings.

Second, our research suggests that nonverbal communication has an important role in relational communication. Fortunately, many relational studies are incorporating nonverbal behaviors in their designs. Perhaps future research could further examine immediacyís role across various relationship types. In addition, further research needs to address the role of back-channel communication in perceived understanding at all relationship levels.

There are some limitations to the present study. The first limitation lies in our back-channel measurement. In order to obtain a more reliable test of back-channel communication, the measurement needs to be expanded to include more nonverbal behaviors. The second limitation lies in the actual survey we distributed. The sections of the survey on immediacy, and back-channel communication did not specify (unlike the perceived understanding instrument) that the subjects should describe how they perceive their partners when they try to make themselves understood. Instead they were general assessments of partner behaviors. This introduces the possibility that subjects answered questions about immediacy and back-channel communication in a more general context of communication rather than during attempts to be understood. Finally, since this study relied entirely on self-report measures, studies of actual nonverbal behavior would compliment and validate these results. Nonetheless, studies of the perceived behaviors are valid in their own right, since perceptions may have the most direct impact on perceived understanding and relational satisfaction.

References

 

Appendix A

Demographic Questions

1. My age group is: 24 and Under, 25-35, 36-46 ,47-57, 58-over

2. My gender: (Male/Female)

3. The relationship level we are in is (check one):

___We date casually

___We date seriously

___We are married-bound

___We are married

4. How long has the relationship existed?

___Days, but less than a week

___Weeks, but less than a month

___Months, but less than a year

___Years, but less than a decade

___A decade or longer

 

Appendix B

Perceived Understanding Instrument

Please indicate the extent to which each term describes how you generally feel when and immediately after trying to make yourself understood by your relational partner. As you fill in the answer sheet, respond to each statement according to the following scale:

5 - Strongly Agree, 4 - Agree, 3 - Undecided, 2 - Disagree, 1 - Strongly Disagree

When I try and make myself understood, I feel:

___ 1. Satisfied___ 9. Annoyed

___ 2. Relaxed___ 10. Uncomfortable

___ 3. Pleasant___ 11. Dissatisfied

___ 4. Good___ 12. Insecure

___ 5. Accepted___ 13. Sad

___ 6. Comfortable___ 14. A sense of failure

___ 7. Happy___ 15. Incomplete

___ 8. Important___ 16. Uninterested

 

 

Appendix C

Hendrickís Relational Assessment Scale

Please indicate your level of agreement with the following statements.

5 - Strongly Agree,

4 - Agree,

3 - Undecided,

2 - Disagree,

1 - Strongly Disagree

___1. My partner meets my needs.

___2. In general, I am satisfied with my relationship.

___3. My relationship is good, compared to most.

___4. I donít regret have gotten into this relationship.

___5. This relationship met my original expectations.

___6. I love my partner.

___7. There are few problems in my relationship?

 

Appendix D

Behavioral Indicants of Immediacy Scale and Back-Channel Scale

Please complete the following scales to indicate how you see your relationship with the person indicated above in situations requiring understanding. Please mark the following statements to indicate whether you: (1) Strongly Disagree, (2) Disagree, (3) Undecided, (4) Agree, or (5) Strongly Agree. There is no correct answer, simply record your perceptions.

___1. This person engages in more eye contact with me than most other people do.

___2. This personís body is more tense than most other people.

___3. This person gestures more than other people.

___4. This person touches me less than most other people do.

___5. This person has a more relaxed body position than most other people.

___6. This person uses more head nods more than most other people do.

___7. This person directs her or his body position more toward me than most other people usually do.

___8. This person smiles more than most other people.

___9. This person engages in less eye contact with me than most other people do.

___10. This person seems eager to spend time talking with me.

___11. This person touches me more than most other people do.

___12. This person seems more distant from me than most other people do.

___13. This person directs her or his body position less toward me than most other people usually do.

___14. This person gestures less than most other people.

___15. This person smiles less than most other people.

___16. This person seems less distant from me than most other people do.

___17. This person uses less head nods than most other people do.

___18. This person seems reluctant to spend time talking with me.

 

Response from Leonard J. Shedletsky

Yet again, intrapersonal variables play a critical role in our understanding of communication behavior, this time, relational communication. In rough, one may translate this research to tell us that the perception of supportive attention to us, interest and understanding, is a key factor in a relationship. Listening is powerful. And again, the power comes from meaning that is communicated. Are students, workers, and lovers all pointing to the phenomena studied here when they speak of others listening and caring?

 

The Pedagogical Effects of Interactive Video Instruction in Oral Communication

Michael Cronin

About the Author: Michael Cronin is a professor of Communication, Director of the Oral Communication Program at Radford University, Box 6932, Radford, VA 24142, telephone 703-831-5750. An earlier version of this article was presented at the May 1993 ICA Convention, Washington, DC. The author gratefully acknowledges the design and statistical assistance provided by Sharon Myers, James Davis, Janet Milton, and Jill Hampton of the Radford University Statistical Consulting Service. This article is a revision of a paper presented at the 1993 AECT Convention.

Abstract: This article reviews experimental research on the pedagogical effects of three IVI programs in oral communication. Formative evaluations indicate that students enjoyed the level III interactive video instruction (IVI) programs. Results indicate that the IVI program "Coping with Speech Fright" was as effective on speech fright and recall measures as lecture or linear videotape instruction by two outstanding public speaking instructors. Furthermore, students in the IVI condition achieved significantly higher immediate and delayed cognitive test scores and significantly greater reduction of speech fright over a four-week period than did students in the control group. Results of two separate studies indicate that students receiving IVI in "Constructing Speaking Outlines" and "Developing Key Ideas: The Four Sís" achieved significantly higher immediate application test scores than did students in the control group or the comparison group. Regression analyses indicated no significant effects of control variables including novelty, GPA, or nature of participation (voluntary versus required) on application test scores of the IVI treatment group in either study.

Perspective:

1. Do the reported pedagogical advantages of interactive video instruction apply to the soft skills disciplines?

2. Do the reported pedagogical advantages of interactive video instruction apply to IVI in oral communication?

3. What is the effect of control variables (voluntary versus required participation, novelty effects, and grade point average) on treatment effects of IVI in oral communication?

 

Recent reviews report significant learning outcomes associated with IVI (Bosco, 1986; DeBloois, 1988; Gayeski & Williams, 1985; Kalowski, 1987; Kearsley & Frost, 1985; McNeil, 1989; Smith, 1987). The most recent review (Fletcher, 1990), which included a meta-analysis of 47 empirical studies of IVI in defense training and related applications in industrial training and higher education, concluded that IVI is more effective and less costly across a variety of instructional settings and objectives than conventional instruction. However, these findings must be interpreted with caution because methodological weaknesses in many of the IVI studies have been identified, including: investigator bias, non-random assignment of subjects, lack of a control group, inadequate definition of instructional treatments, failure to measure the degree to which the treatments were implemented by subjects, artifact, lack of generalizability, and inadequate sampling (Bosco, 1986; Bunderson, Baillio, Olsen, Lipson, & Fisher, 1984; Cushall, Harvey, & Brovey, 1987; Reeves, 1986, 1990; Slee, 1989; Smith, 1987).

Many of the empirical studies considered in these reviews were conducted in the hard sciences. Theorists (Biglan, 1973; Kolb, 1981; Kuhn, 1962; Moses, 1990) have identified significant differences in typical teaching style and typical learning style in the hard sciences (laboratory science and mathematics) as compared to soft skill disciplines (humanities and social sciences). Do the reported pedagogical advantages of IVI apply to the soft skill disciplines? Cronin and Cronin (1992b) reviewed 32 post-1984 empirical studies in soft skill areas. They found that many of these studies avoided the methodological problems of earlier studies and concluded that "taken as a whole, these studies appear to indicate that IVI produces significantly greater cognitive and application gains than conventional methods of soft skill instruction" (p. 59). IVI produced significant cognitive or application gains in soft skill areas such as reading, management, study skills, logical reasoning, foreign language, sales training, photography, secondary and special education, economics, and art.

Theoretic Foundations of IVI1

Theoretical frameworks for research on learning outcomes from IVI have focused on the interactive, visual, and learner control capabilities of IVI. The multimedia interactivity inherent in well-designed IVI can enhance learning. Although increased interactivity may hinder learning for certain types of learners (cf. Dalton, 1986a; Milheim & Azbell, 1988; Schaffer & Hannafin, 1986), "in general, where the learner reacts to or interacts with the criterial stimulus, learning is facilitated, and that facilitation increases with the degree of learner activity or involvement" (Fleming & Levie, 1978, p. 138). Specific types of interactivity such as guided pathways for inexperienced users (Hoelscher, 1989) and instructional cues for complex interactive programs (Lee, 1989) help learners form accurate interpretations, provide practice in important concepts, and provide relevant feedback (Schaffer & Hannafin, 1986).

Visual images available in well-designed IVI can enhance learning. Theorists have explained the effectiveness of visuals in learning as a function of the increased comprehensibility of the content (Burwell, 1991); selective increases in learnersí attention (Brandt, 1987; Miller & Irving, 1988); and increased enjoyment (Sewell & Moore, 1980). Hansen (1989) postulated that the video component of IVI enhances learning at several stages of the five-stage model of skill acquisition developed by Dreyfus and Dreyfus (1986). Realistic video representations in IVI can foster recognition and understanding of plan-oriented behaviors in complex social situations (Hansen, 1989) and facilitate higher levels of skill performance through active discovery and application on the part of the learner (Hamilton & Taylor, in press). IVI allows learners to link visual images with related text or graphic information in order to compare their understanding to video representations of othersí understanding of the same material.

Appropriate use of the learner control options in well-designed IVI can enhance learning. Reigeluth and Stein (1983) contended that informed learner control by motivated learners generally increases the effectiveness and appeal of instruction. An underlying assumption of much of the IVI research is that at least some users of the technology can best determine what they want and need to know and the best way to achieve that understanding. Individualized instruction is, theoretically, best able to adapt to various learning styles (Kinzie & Berdel, 1990; Laurillard, 1989).

The ability and willingness of learners to use control options effectively (Hannafin, 1985) is of paramount importance given the fact that students "may not be the best judges of what instruction they need, how much instruction [they need], when to seek instruction, and what to attend to in an instructional segment" (Canelos, Baker, Taylor, Belland, & Dwyer, 1986, p. 67). Based on a review of research regarding learner control, Milheim and Azbell (1988) suggested that learner control is most effective when students have some expertise in the content area, are trained in the use of learner control, possess high aptitude and high inquiry, and are unlikely to skip important material or quit the lesson prematurely.

Although the interactive, visual, and learner control theories discussed above have not been integrated in a comprehensive theoretical structure specific to IVI, the assumption in these preliminary investigations is that a holistic examination of these strategies is warranted in explaining the motivational and achievement effects of IVI. This holistic approach integrates strategies for promoting motivation and achievement through IVI derived from cognitive, affective, and social learning theories. For example, the opportunities for choice and immediate, consistent, and non-judgmental feedback in IVI enhance self-efficacy and expectancy of success on the part of the user (Dalton & Hannafin, 1984). Self-regulated learning fosters both learning and motivation (Como & Mandinach, 1983) by making such learners metacognitively, motivationally, and behaviorally active participants in their own learning process.

Although IVI in soft skill areas appears to offer significant instructional benefits in cognitive achievement, transfer of learning to performance, motivation to learn, student achievement across uncontrolled student characteristics, user acceptance of the technology, and time required to achieve content mastery; do these outcomes apply for IVI in oral communication? The remainder of this article summarizes experimental studies of the pedagogical effects of IVI in "Coping with Speech Fright," "Constructing Speaking Outlines," and "Developing Key Ideas: The Four Sís."

"Coping With Speech Fright"2

Recent research indicated that approximately seventy percent of the population experiences moderately high or high communication apprehension in public speaking contexts (Richmond & McCroskey, 1989). High levels of communication apprehension have negative consequences in the speech-making process such as: communication avoidance (Beatty, 1987; Mulac & Wiemann, 1984); shorter speeches (Beatty, Forst, & Stewart, 1986); lower self-esteem, less effective public communication (Daly & Stafford, 1984); less effective preparation for public speaking, increased perceptions of failure in speaking (Kelly, 1984); increased disfluencies, less effective nonverbal communication while speaking in public (McCroskey, 1982); more frequent decision-making errors in constructing a speech (Beatty, 1988); and considerable anxiety in those with high levels of communication apprehension if they are forced to communicate (Beatty, 1987).

Most institutions have too few speech faculty competent to provide traditional instruction in coping with communication apprehension to such a large population (Ayres & Hopf, 1987). Interactive video instruction (IVI) is capable of providing effective training to large numbers of students in a cost-effective manner (Fletcher, 1990).

Research Hypotheses

This investigation involved a control group which received no formal instruction in overcoming speech fright, an IVI group, and a group which received lecture/linear videotape instruction by outstanding teachers in cognitive restructuring techniques to cope with speech fright. Because the content of the lecture/linear videotape version of the lesson was virtually identical to the IVI and because only outstanding instructors were used to present the lecture/linear videotape version in this study, no significant differences on cognitive gains between treatment groups were predicted. However, given the educational advantages of IVI discussed previously these hypotheses are proposed:

Hypothesis 1: Students using IVI will achieve significantly higher cognitive test scores immediately after treatment than will students in the control group.

Hypothesis 2: Students using IVI will achieve significantly higher cognitive test scores four weeks after treatment than will students in the control group.

After receiving IVI, students in all conditions presented at least one graded speech before completing the delayed-test instruments. Because students trained over a substantial time period in cognitive restructuring techniques and given opportunities to practice them showed significant reductions in communication apprehension (Connell & Borden, 1987; Harris, 1980), we expected:

Hypothesis 3: Students using IVI will achieve significantly greater reduction in speech fright over a four-week period than will students in the control group.

Method

Subjects. After adjusting for student absences and failures to follow directions, 138 college students enrolled in introductory public speaking classes at a middle-sized, comprehensive university in the southeast region served as subjects. Male subjects comprised 48% of the sample and female subjects made up 52%. Students received no extra course credit for participation in the study.

Instructional Materials. Two parallel forms of the lesson were developed, a lecture/linear videotape and a level III interactive videotape version. The multimedia IVI module incorporates a tutorial/simulation approach. It includes carefully designed orienting activities, questions, feedback, and review options to promote understanding. A user-friendly design facilitates student use by explaining the nature of IVI, instructing students in program use, and repeating specific instructions for use of each screen requiring student reaction.

The multimedia approach promotes interest and understanding through humorous graphics, visual memory cues, and workbook exercises designed to enable students to apply IVI learning to their specific problems of speech fright. Video adapted to the target audience includes examples of speakers experiencing speech fright, testimonies of those who suffer from speech fright, examples of studentsí negative self-evaluation, trainer-led exercises and simulations for developing positive self-evaluation, and a brief analysis of other techniques for coping with speech fright.

The IVI program provides a carefully constructed combination of learner-control of major topics to be explored (i.e., the nature of speech fright, how to overcome speech fright, and nonverbal indicators of speech fright) with program-control of IVI within each major topic. This allows learners to focus on topics that interest them and ensures that they will follow the lesson sequence developed by experts in instructional design for each topic selected (Canelos, Baker, Taylor, Belland, & Dwyer, 1986).

Procedures and Design. The independent variable consisted of the videotape-based instructional materials with two levels (interactive videotape versus lecture/linear videotape). The dependent variables included immediate and delayed cognitive test scores, immediate and delayed scores on the public speaking section of the Communication Apprehension in Generalized Contexts instrument, and immediate responses on the formative evaluations by the treatment groups.

Students in intact public speaking classes, most of whom had received no instruction in reducing speech fright, served as the control group. In an attempt to avoid a Hawthorne-type effect, students in the control group received a 45-minute placebo lecture on public speaking techniques unrelated to coping with speech fright before completing dependent measures.

Students in the treatment groups, most of whom had received no instruction in reducing speech fright, were randomly assigned to IVI or lecture/linear video (LLV) treatments. Treatments were conducted during the second week of the Spring 1991 semester. The 45-minute LLV treatment was presented in three classes by the assigned instructor to groups of eight to ten students in a classroom equipped for videotape playback. Both instructors received outstanding teaching ratings from the department chair and averaged 4.5 out of 5 on their most recent student evaluations. The instructional content of the LLV was virtually identical to the IVI version and included almost all the video included in the IVI condition. However, this video was presented as linear video.

The randomly assigned IVI users received an average of 46 minutes of individual instruction in a small private laboratory room. This condition included the opportunity to complete workbook exercises but did not require completion of these exercises, provided no opportunity for class discussion, and did not require students to be exposed to all the content instruction.

Students completed a ten-item test of cognitive recall (test-retest reliability = .693) and the CAGC-PS (test-retest reliability = .702) immediately after they received the instructional material. They completed Form B of the recall test and the CAGC-PS (Richmond & McCroskey, 1989) four weeks later.

Results

Cognitive TestñImmediate. The results supported hypothesis 1, which predicted that students using IVI would achieve significantly higher cognitive test scores immediately after treatment than would students in the control group. There was no significant difference between the mean immediate cognitive test scores of students in the IVI and the LLV conditions. The maximum possible score on the cognitive test was 10. The mean score was 8.63 for students in the IVI condition, 8.37 in the LLV condition, and 7.10 in the control group. ANOVA results showed significant differences among the mean immediate cognitive test scores of subjects in the three groups (R2 = .26, F [2, 135] = 23.92, p < .0001). The group differences accounted for 26% of the variance.

Cognitive TestñDelayed. The results supported hypothesis 2, which predicted that students using IVI would achieve significantly higher cognitive scores four weeks after the treatment than would students in the control group. There was no significant difference between the mean delayed cognitive test scores of students in the IVI and LLV conditions. The maximum possible score on this cognitive test was 10. The mean score was 7.77 for students in the IVI condition, 7.26 in the LLV condition, and 6.22 in the control group. ANOVA results showed significant differences among the mean delayed cognitive test scores of subjects in the three groups (R2 = .22, F [2, 135] = 18.70, p < .0001). The group differences accounted for 22% of the variance.

Communication Apprehension Test Scores. The results supported hypothesis 3, which predicted significantly greater reduction in speech fright over a four-week period for students in the IVI treatment than in the control group. Students in the IVI condition showed a mean reduction in speech fright of .866 points over the four-week period, while students in the control condition showed a mean increase in speech fright of .876 points. An independent t-test indicated that these mean change scores were significantly different (t [70.5] = -1.81, p < .04). There were no significant differences between the CAGC-PS scores of students in the IVI and LLV treatment groups on either immediate or delayed measures.

Educational Implications

Discussion of implications relevant to all three studies of IVI in oral communication will be presented in the final section of this article. Only implications unique to this study will be discussed here.

The finding of no significant differences between the IVI and LLV treatments may encourage educators to investigate innovative applications of IVI in "Coping with Speech Fright." The use of individualized, self-paced IVI to train large numbers of students may be at least as effective, and perhaps more effective and less costly across a wide variety of instructional settings and objectives than conventional instruction (Fletcher, 1990).

Given the demonstrated success of the IVI program in increasing cognitive learning and reducing self-reported speech fright, the CAGC-PS scores of students in the IVI and control conditions may show even more significant differences if compared over a longer period of time. Cognitive restructuring training requires substantial time and speaking practice for students to apply these techniques to reduce their specific fears. The absence of instructor-led practice in applying cognitive restructuring techniques to public speaking performance, while helpful to the integrity of the research design, reduces the likelihood of producing significant effects on speech fright. Instructors may wish to explore innovative combinations of IVI in "Coping with Speech Fright" and instructor-led practice in applying cognitive restructuring techniques to reduce communication apprehension. Such cost-effective applications may enable institutions to teach larger numbers of students to reduce their speech fright.

 

"Constructing Speaking Outlines"3

Theoretical Perspectives

Research, though limited, generally supports the conclusion that effective organization of an oral message increases recall, attitude change, and speaker credibility ratings (Daniels & Whitman, 1981; McCroskey & Mehrley, 1969; Thompson, 1967; Whitman & Timmis, 1975). The authors of current public speaking textbooks view outlining as critical to the organization process. A survey of recent public speaking texts reveals that organization is seen as a key to speaker success and outlines are seen as the key to organization (e.g., Sprague & Stuart, 1988; Sproule, 1991, Verderber, 1991).

Research Hypotheses

Given the theoretic analysis and educational advantages of IVI discussed previously:

Control Variables

It is, perhaps, more important in multimedia research to examine treatment group effects in light of control variables than to focus on comparisons of various instructional treatments (Clark, 1985; Clark & Sugrue, 1988). Thus, voluntary versus required participation, novelty effects, and reported GPAs were included as control variables for the hypothesized treatment effects.

No previous study has investigated the effects of voluntary versus required participation in IVI. High willingness-to-communicate subjects are significantly more willing to agree to participate and significantly more likely to participate in communication research studies than low willingness-to-communicate subjects (Zakahi & McCroskey, 1989). Thus, volunteer subjects in communication research are likely to exhibit higher willingness-to-communicate than students who are required to participate. Furthermore, volunteer subjects are likely to be more accepting of IVI than non-volunteer participants and thus may be more motivated to learn from the IVI.

Several theorists assert an alternative hypothesis for the reported pedagogical effects of IVI, namely; the novelty effect of IVI may produce higher initial motivation to learn (Clark & Sugrue, 1988; Hannafin, 1985; Slee, 1989). However, a critical synthesis of IVI research in soft skill areas identified several studies that controlled for novelty effects and reported significant pedagogical effects for IVI (Cronin & Cronin, 1992b). Furthermore, the only reported experimental study of IVI in speech communication found no significant novelty effect for IVI on immediate or delayed cognitive test scores. Neither preference for IVI over traditional instruction nor previous use of other IVI modules versus first-time use showed a significant association with cognitive test scores (Cronin, Grice, & Olsen, 1994).

The self-paced individualized instruction available in IVI may be more effective than traditional instruction for low-prior-achievement users. Savenye and Strand (1989) reported that the average student using IVI in physical science instruction demonstrated higher achievement levels than the average student receiving traditional physical science instruction. This difference was even more pronounced for students of lower prior achievement. Furthermore, although results are mixed, most of the research on application-intensive IVI similar to that used in this investigation found either no significant difference on immediate posttests between high-prior achievers and low-prior achievers using IVI (Dalton, 1986b; Gray, 1987) or a significant prior achievement x treatment interaction (Schaffer & Hannafin, 1986) favoring low-prior achievers. Students with high GPAs generally surpass students with low GPAs on cognitive tests on instructional material. However, IVI may help close the application-gain gap between students with lower academic achievement and students with higher GPAs.

Method

Subjects. After adjusting for student absences and failures to follow directions, 141 college students at a middle-sized, comprehensive university in the southeast region served as subjects. Male subjects comprised 51% of the sample and female subjects made up 49%.

Instructional Materials. The multimedia approach promotes interest and understanding through humorous graphics, visual memory cues, dual screen and dual channel presentations, and exercises designed to enable students to apply IVI learning to the construction of speaking outlines. Two-screen or two-channel display in IVI provides two major pedagogical advantages. First, these dual options allow the user to become the editor of the program. The user can decide whether to listen intently to the voice-over, to listen while reading the summary, to try to tie in the visuals with the text, or to integrate all of these elements. This editing option allows users to adapt the IVI program to their learning styles. Secondly, the simultaneous use of several interactive channels adapts to the modern TV, remote control, stimulus-load expectations of users. The stimulus load may be a necessary adaptation to the media expectations of users accustomed to the quick-cutting images of modern music videos, commercials, and movies, as well as routine "channel hopping."

Procedures and Design. The independent variable consisted of the presence or absence of videodisc-based IVI on "Constructing Speaking Outlines." The dependent variables included immediate application test scores and treatment group responses on the formative evaluations. Analysis of variance (ANOVA) was used to analyze the application test scores among the treatment, control, and comparison groups. Following a significant F, Duncanís Multiple Range Test was used to determine significant differences among individual group means. Regression analysis was used to analyze the application test scores of the treatment group with the nature of participation (required versus voluntary) and Sís rating of feedback in the IVI program, rating of the video in the IVI program, reported grade point average, previous use of IVI on other topics, computer skills estimate, and preference for IVI compared to traditional classroom instruction serving as control variables.

Students from non-speech classes in economics, political science, health, and marketing either volunteered or were required to undergo IVI. They were randomly assigned to the treatment or the control group. These subjects received neither instruction in constructing speaking outlines in class prior to the study nor extra credit for their participation in this study.

In an attempt to avoid a Hawthorne-type effect, students in the control group received approximately 30 minutes of placebo IVI in "Developing Key Ideas: The Four Sís." This instruction provided no information on constructing speaking outlines. The control group consisted of 25 males and 22 females, 24 volunteered and 23 were required to participate.

The treatment group received approximately 35 minutes of IVI in "Constructing Speaking Outlines." The treatment group consisted of 24 males and 20 females, 24 volunteered and 20 were required to participate. These ratios were virtually identical to those in the control group.

The comparison group consisted of 50 students from three intact performance classes in speech (two in public speaking and one in argumentation and debate) who had received "normal" instruction, practice, and feedback on constructing speaking outlines. The comparison group was included to compare learning outcomes from IVI in constructing speaking outlines with the usual instruction on this topic provided in these speech classes.

The study was conducted between November 12 and 26, 1991. Participants in both the treatment and control conditions were shown how to use the IVI program by a trained student worker and were left alone to complete the lesson. Students in both treatment and control groups completed a sixteen-item test on constructing speaking outlines (split-half reliability = .89), a sixteen-item test on developing key ideas, and a formative evaluation of the instruction immediately after they received the instructional material. Students in the comparison group completed both sixteen-item tests during a regularly scheduled class. These measures were randomly ordered and randomly assigned in all groups to control for an order effect.

Results

ANOVA results supported hypothesis 1, which predicted that students using IVI on "Constructing Speaking Outlines" would achieve significantly higher application test scores immediately after treatment than would students in the control group. ANOVA results also supported hypothesis 2, which predicted that students using IVI on "Constructing Speaking Outlines" would achieve significantly higher application test scores than would students in the comparison group. The maximum possible score on the application test was 16. The mean score was 10.72 for volunteer participants in the treatment group, 10.21 for required participants in the treatment group, 7.82 for required participants in the control group, 7.76 for volunteer participants in the control group, and 7.62 for participants in the comparison group. These scores were significantly different (R2 = .30, F = [4, 136] = 14.72, p < .0001). The group differences accounted for 30% of the variance.

Regression analysis using control variables as predictor variables indicated that volunteer participants did not achieve significantly higher application test scores immediately after treatment than did subjects required to participate. The mean application test scores in the treatment group for volunteers (M = 10.72) and required participants (M = 10.21) were not significantly different (F [2, 41] = .61, p < .4382). The addition of this predictor variable to the regression model increased the R2 by only .014 (see Table 1).

Regression analysis using control variables as predictor variables indicated that there was no significant novelty effect on application test scores in the treatment group. No significant association

was found between immediate application scores in the IVI treatment and variations in subjectsí previous use of other IVI modules (R2 = .05, F [1, 42] = 2.23, p < .1432). Likewise, there was no significant association between immediate test scores and variations in preference for IVI versus conventional instruction (F [7, 36] = .10, p < .7582). The addition of this predictor variable to the regression model increased the R2 by only .002 (see Table 1).

Regression analysis using control variables as predictor variables found no significant association between variations in GPA and application test scores immediately after treatment (F [4, 39] = .26, p < .6158). The addition of this predictor variable to the regression model increased the R2 by only .006 (see Table 1).

"Developing Key Ideas: The Four Sís"

This study was conducted concurrent with the "Constructing Speaking Outlines" study. The theoretical perspectives were similar in exploring the effects of organization in oral communication. However, this study focused on the organization of key ideas via signposting, stating, supporting, and summarizing each key idea. Grice and Skinner (1993) contend that "use of the ë4Ssí is fundamental to effective organization within the body of any speech" (p. 175). Except for substituting

 

Table 1

Constructing Speaking Outlines

Treatment Group Increments to R2 and p for

Each Predictor Variable

VariableR2IncrementFdfSEp

Required v. voluntary.0643.0141.412, 41.548.256

Video rating.0946.006.646, 37.683.693

Reported GPA.0816.006.874, 39.651.492

Computer competence.0880.006.735,38.273.602

Previous use of IVI.0503----2.231, 42.279.143

Feedback rating.0755.0111.093, 40.754.364

Preference for IVI.0970.002.557, 36.287.788

 

 

Table 2

Developing Key Ideas: The Four S's

Treatment Group Increments to R2 and p for

Each Predictor Variable

VariableR2IncrementFdfSEp

Required v. voluntary.0349----1.561, 43.569.219

Video rating.0557.0208.932, 42.568.342

Reported GPA.0665.0108.473, 41.611.495

Computer competence.0716.0051.224, 40.255.642

Previous use of IVI.0739.0023.105,39.233.756

Feedback rating.0742.0003.016,38.544.918

Preference for IVI.0742.0000.007, 37.350.979

 

"developing key ideas" for "constructing speaking outlines," the research hypotheses, control

variables, and the method were identical to those described in the "Constructing Speaking Outline" study with three exceptions. (1) The split-half reliability of the application test in this study was .891. (2) The instructional material provides a tutorial approach in this multimedia IVI module on "Developing Key Ideas." It includes carefully designed orienting activities, questions, feedback, graphics, video simulations, and review options to promote understanding. A user-friendly design facilitates student use by explaining the nature of IVI, instructing students in program use, and repeating specific instructions for use of each screen requiring student reaction. (3) The subjects in the treatment group in this study on "Developing Key Ideas" served as the subjects in the control group in the "Constructing Speaking Outlines" study and vice versa.

Results

ANOVA results supported hypothesis 1, which predicted that students using IVI on "Developing Key Ideas" would achieve significantly higher application test scores immediately after treatment than would students in the control group. ANOVA results also supported hypothesis 2, which predicted that students using IVI on "Developing Key Ideas" would achieve significantly higher application test scores than would students in the comparison group. The maximum possible score on the application test was 16. The mean score was 13.96 for volunteer participants in the treatment group, 13.14 for required participants in the treatment group, 4.95 for required participants in the control group, 4.96 for volunteer participants in the control group, and 7.27 for participants in the comparison group. These scores were significantly different (R2 = .60, F = [4, 138] = 51.75, p < .0001). The group differences accounted for 60% of the variance.

Regression analysis using control variables as predictor variables indicated that volunteer participants did not achieve significantly higher application test scores immediately after treatment than did subjects required to participate (see Table 2). The mean application test scores in the treatment group for volunteers (M = 13.96) and required participants (M = 13.14) were not significantly different (F [1, 43] = 1.56, p < .219).

Regression analysis using control variables as predictor variables indicated that there was no significant novelty effect on application test scores in the treatment group. No significant association was found between immediate application scores in the IVI treatment and variations in subjectsí previous use of other IVI modules (R2 = .07, F [5, 39] = .10, p < .756). The addition of this predictor variable to the regression model increased the R2 by only .002 (see Table 2). Likewise, there was no significant association between immediate test scores and variations in preference for IVI versus conventional instruction (F [7, 37] = .00, p < .979). The addition of this predictor variable to the regression model increased the R2 by .000 (see Table 2).

Regression analysis using control variables as predictor variables indicated that there was no significant association between variations in GPA and application test scores immediately after treatment (F [3, 41] = .47, p < .495). The addition of this predictor variable to the regression model increased the R2 by only .011 (see Table 2).

General Educational Implications

The apparent efficacy of IVI in significantly enhancing subjectsí learning of oral communication techniques in each of the three studies is consistent with research on the effects of IVI in related soft skill areas (Cronin & Cronin, 1992b). This finding appears to indicate that a well-designed IVI program can prove effective in instructing students in oral communication.

The formative evaluations conducted in each of the three studies indicate that IVI users appeared to be highly motivated to learn from the medium. Most users found the IVI programs interesting and enjoyable and wanted to learn more about oral communication after taking the instruction. IVI users felt that they would be capable of using the instruction to improve their oral communication competency. Likewise, less than one percent of the IVI subjects found it difficult to use the programs. This finding may reinforce the results of a recent study (Cennamo, Savenye, & Smith, 1991) in which undergraduate students perceived that it was significantly easier to learn from IVI than from instructional television and television.

The speaking outline and developing key ideas studies found no significant effects of GPA or nature of participation (voluntary versus required) on application test scores in the IVI treatment groups. Furthermore, the most plausible competing hypotheses did not appear to explain results in these studies. Time-on-task and instructional content were equivalent for the IVI and lecture/linear video groups in the speech fright study which reported that IVI was as effective as instruction from outstanding teachers. The novelty effect associated with IVI (Clark & Sugrue, 1988; Hannafin, 1985; Slee, 1989) did not appear to explain results in any of the three studies.

Perhaps the most important outcome of these studies is the preliminary empirical documentation of the instructional effectiveness of the three IVI programs in oral communication. Prior to these studies, very little research supporting applications of IVI in oral communication was available.4

Notes

1This section includes a review and extension of an article by Cronin and Cronin (1992a).

2Material in this section is taken from Cronin, Grice, and Olsen (1994).

3Material in this section and the "Developing Key Ideas" section is taken from Cronin (1994).

4Although space limitations preclude an extended discussion, Cronin and Myers (1993) have published an article that describes the nature and preliminary assessment of an interactive videodisc instructional program on listening. Formative evaluations indicate that students enjoyed the level III interactive multimedia instruction (IMI). Results of a dummy variable regression analysis indicate that students receiving IMI in listening achieved significantly higher immediate cognitive test scores and listening gain scores than did students in the control group. Regression analysis indicated no significant effect of reported GPA, year in school, previous listening instruction, or time-on-task on listening test gain scores of the IMI treatment group. However, significant effects on cognitive test scores were found for reported GPA, previous listening instruction, time-on-task, and senior versus non-senior standing.

 

References

Response from Leonard J. Shedletsky

Dr. Cronin, with an intrapersonal approach to learning, adds to the reminder that learning is only partly a change in observable behavior. One can take from this research the conclusion the IVI instruction is at least as good as classroom instruction in the ways measured here. Interactivity seems to be the central variable. Is that true, and if so, what exactly is the nature of interactivity?

 

Penetrating the Glass Wall: Creating and Retaining Imagined Interaction in Televised Distance Education

Gregory Gutenko

About the Author: Dr. Gregory Gutenko (Ph.D., University of MissouriñColumbia) is currently an associate professor of Communication Studies at the University of MissouriñKansas City. He has over ten years of professional experience in instructional media production, including video programming and televised distance education.

Abstract: Televised courses cannot support the same levels of interaction and interpersonal dynamics which can exist between teacher and student in conventional classroom settings. Studies examining various video production methods suggest a number of factors which may contribute to creating an effective illusion of interaction. Providing the appearance of interpersonal engagement, a performance skill commonly exercised by media professionals with the aid of teleprompters, can be valuable in developing parasocial interaction with student viewers. Constructing telecasting sites which can encourage the display of interpersonal communication cues in instructors is possible, but will require a new level of production design.

Perspective:

1. How important is personal interaction between instructors and students in learning?

2. Do you think a televised distant education course would be at least as involving as a class in a large lecture hall with over 100 students? Would the televised class be preferable in any way?

3. What obligation do educators have in designing courses and instructional methods which are supportive of students with moderate to low motivation levels?

  

Virtually every television station and video production facility has one. The teleprompter, a device used to project a speakerís text directly before the lens of a camera, has been a fundamental production tool for almost forty years. Its value in creating effective on-screen presentationsñenabling the impression of direct eye contact between speaker and viewing audienceñhas never been questioned in the broadcast and teleproduction workplace. However, the actual effective consequences of its applications have been questioned and tested by a few outside of the workplace, and these results suggest that the teleprompterís potential in teleteaching must be carefully understood and appreciated.

It must first be noted that teleprompters are not ubiquitous in telecourse environments. This is usually the result of budget priorities, not of a lack of interest. Additionally, the dominant technical challenges in televised distance education production have concerned unavoidable and crucial matters of basic engineering, transmission, distribution, and operation. Teleprompter use has aesthetic rather than technical consequences, and many telecourse activities have been compelled by limited time and resources to deal exclusively with the technical. However, the aesthetics of teleteaching production may have a significant effect on instructional achievement. 

The Video Wall

The current implementations of televised distant education technology systems and programming mark what can be considered the third widespread attempt to transcend the walls of the classroom and deliver instruction (and instructors) to remotely situated students through the video medium.1 Insufficient attention and study has been devoted, however, to many of the communication obstacles inherent in this medium. There has always been a pervasive hyperbole concerning televisionís "impact" and power to engage audiences, and many who are new to the use of video communications assume and trust that this medium will naturally endow their messages with a compelling allure.

Decades ago, television was truly captivating in and of itself, a fascinating new household presence. Televised distance education also is imbued with the aura of novelty, but this novelty often quickly evaporates as students become accustomed or even disenchanted with the experience (Libler, 1991). Television at large is a commonplace thing. Television is an idly regarded feature of everyday life that rarely stimulates much beyond viewer acquiescence. Its acknowledged powers of persuasion reside in its repetitive delivery of consistent and redundant advertising, entertainment, and cultural messages. Television is still, by nature, a medium that filters and reduces reality, and it will always transmit less than what is placed before it. Television truly mediates communication.

This is a barrier that must be contended with. It is a glass wall constructed of lenses and video screens which physically separates the observed from the observer. Images do not travel easily or uncompromised through such a dense medium. Visual resolution is lost (with or without compression), colors are diminished, and there is no peripheral vision or third dimension. Penetrating this wall successfully would necessitate a full consideration and exploitation of those attentional variables which survive transmission through this electronic filter.

To compensate for diminished stimuli (e.g., loss of visual resolution, three dimensional space, sound perspective) adroit users of the medium have developed many techniques and criteria for constructing, selecting, and presenting content. Two fundamental approaches can be noted which have relevance to intrapersonal communication processing: preserve from mediation and retain as much reality and presence as is possible, and compensate for what is unavoidably lost by exploiting and enhancing the special attributes and capacities of the medium (e.g., manipulation of time, presentation of images of far removed events and phenomena, graphic renderings)

The latter strategy is easily appreciated by educational television producers, and for over half a century, many instructional designers have excelled in the development, testing, and implementation of production techniques to achieve this end (Mitchell, 1979). Less recognized and appreciated are the techniques of capturing and conveying the visual cues and behaviors of on-camera instructors which are often lost when teaching is packaged for television. These cues and behaviors are the vital communicative indicators that are apprehended and responded to within live interpersonal settings. Important attentional variables which may contribute to the effectiveness of live classroom instructional delivery are frequently stripped away carelessly by the process of televising.

The Need for the Communicative Illusion

Disenchantment and dissatisfactions with the use of television for instructional delivery, which first came to light in the late 1950s with ETV or closed circuit video, and again in the late 1960s and early 1970s with videotaped production, often have not been systematically reviewed and addressed by many contemporary distance education proponents. (The more easily recognized and defined tasks of hardware installation, network building, and enrollment development demand and take center stage.) One early observation concerning the loss of interest and support for the earlier applications of video instruction faults the lack of interaction and interpersonal dynamics which normally exist between the teacher and student in conventional classroom settings (Technology Comes to the Classroom, 1959). This same concern is being voiced once again regarding contemporary efforts (Brown, 1988; Cambre, 1987).

Current technology enables live presentation, which is not a new capability, and live voice exchange with distant students by telephone, which is a new capability but one which is often underutilized by students (Gutenko, 1989). The promotional videos and literature which are used to recruit participants for various distance education offerings suggest that this level of interaction provides for an adequate engagement of the distant student. However, the extent and frequency of interaction with telephone tutors and actively involved local facilitators appears to be the most significant factor supporting desirable levels (equivalent to live instruction) of student motivation and persistence in learning (Keinath, 1991; Speth, Mercer, Poggio, and Glassnap, 1992; Sullivan, Andrews, Madigan, and Noseworthy, 1979). While recent research has yet to define consistent and specific tutor or facilitator attributes and practices which correlate with student success (Russell, 1991; Libler, 1991), it would appear that the success of distance education as an overall process primarily depends on both the quantity and quality of direct human interaction.

The enormous costs and technological complexities associated with the implementation of systems capable of delivering full-motion, high resolution video and audio in a two-way, fully spontaneous manner can be overwhelmingly prohibitive. Can a more affordable and obtainable illusion of interaction also motivate and engage students? Simulating interpersonal engagementña requisite performance skill possessed by successful mass media personalitiesñhas been effective in creating and maintaining viewer empathy and engagement. This perception is particularly true of those on-camera performersñnewscasters, program hostsñwhose mode of address is primarily direct into the camera.

As described by Horton and Wohl (1956), parasocial interaction is a particular manifestation of an interpersonal illusion where broadcast performers, especially at the local level, come to be regarded as familiar and approachable, and open to casual and informal conversation when encountered in person by viewers. A personís intrapersonal processing and imagined interactions give rise to the illusion of a relationship. Parasocial interaction is quite different from the isolating "star aura" that surrounds and distances major television and film performers, but it is perhaps one of the most significant factors in generating viewer loyalty to a specific broadcast station. Consider the six digit salaries that key anchors earn compared to the modest five digit salaries of the same television stationís general assignment field reporters. Clearly, the value of the principal, full-time personalities are recognized and rewarded. A popular anchor who moves from one station to another in the same market may take most of her or his audience with them, despite the fact that both stationsí newscasts may be formally and technically identical. Establishing and maintaining parasocial interaction is both possible and significant for audience attention and viewing motivation.

It is often argued that motivation, while not essential to learning, is certainly facilitative (Paluzzi, 1980). Using specific production techniques to draw and hold the attention of students to lesson content continues to be a fundamental task in instructional media design, and one which merits continuing study and testing (Mitchell, 1979). However, telecasting site configurations and production techniques that can capture and convey the interpersonal communication cues of non-media trained instructors will require a greater "hardware transparency" than is required by media performers. As is the case with cameras and microphones, the teleprompter can be an intimidating presence, and teachers may be encouraged by telecasting devices which can masquerade as teaching aids or prompting devices.

Turn On, Tune In, Drop Out?

It is not uncommon for instructors at post-secondary institutions to find fewer students accounted for on their class roster at the end of a semester than were listed at the beginning. A student may withdraw from a class for many reasons. However, the attrition rate for students enrolled in televised courses is, by several accounts, higher than it is for live courses (Carl, 1991). Some observers have estimated the dropout rate to be double that found in conventional learning (Clark & Verduin, 1989). Certainly, there have been no accounts of televised courses having fewer dropouts than comparable live attendance courses.

It is acknowledged that those students who enroll in and complete televised courses are more highly motivated than the average student (Speth, 1990; Carl, 1991). This suggests that average students are either disinterested in televised courses, or are the ones who become discouraged and drop out. If it is true that the learning needs of the average student are not being adequately met by current distance education methods, this should be a matter of particular concern for institutions which are state supported. Public funding is allocated to such institutions on behalf of all citizens, even those who can muster up only an average motivation towards study. If a state supported institution diverts significant physical and academic resources into instructional delivery systems in which only the most persevering of citizens can succeed, how fair and equitable is this institutionís support of educational opportunity?

One highly publicized outcome of televised distance education is the apparent equivalency of student achievement between the distant and in-studio sections of televised courses. Most comparisons show equal or slightly better grade-point achievement by the distant students. However, equivalency of student achievement between televised courses and conventional classroom courses has not been addressed in such comparisons.

The Illusion of Interaction

The predominant shot used in most live television courses is the "talking head" of the instructor, unless there is a considerable amount of presentation time allocated to viewing graphics or the instructorís sketchpad and/or chalkboard. The term "talking head" has a negative connotation, suggesting a boring and stilted image, as was noted in the past and present (Cambre, 1987, p.6; Paluzzi, 1986, p. 35). A talking head can be a compelling image, however, if the viewer is interested in and engaged by the subject. Bill Moyers of the Public Broadcasting System has produced a number of "head-to-head" programs, as in The Power of Myth series with Joseph Campbell. In the television news industry, the most highly paid and critically valued personnel are the "talking heads." The potential and actual importance of the human image in the mix of all images used in news and information video programming is undeniable.2

What is the difference, then, between the "talking head" that holds viewersí attention and the one that does not? What attributes of the image and staging of the shot leads a viewer to develop a parasocial bond with one subject and tune out with another? The imagined interaction between television performers and viewers can create in some members of the public a sense of camaraderie and familiarity with the media performer that requires great patience, professional aplomb, and understanding to endure (Horton & Whol, 1956; Levy, 1977; Levy, 1979; Atkins, 1984). Certainly, the physical characteristics possessed by a particular media talent, along with her or his well utilized communication performance skills, are critical variables. Personal "charisma" may also be a key factor. Beyond personal magnetism and refined professional delivery, however, are a number of image and staging factors that will affect the presentation and apprehending of any subject placed before the camera.

Most pertinent to the production of live television performance are the formal variables of unified eyelines, eye contact, and perspective. A unified eyeline exists when the participants in an interpersonal exchange are all situated at one common, shared eye level or plane. Eye contact exists when camera performers address the audience with their gaze, and perspective relates to the perceived distance between performers and camera viewpoint. Teleprompter use essentially compels unified eyeline, close-in perspective, and facilitates any amount of eye contact.

The constancy and value of these particular production variables in affecting the perception of televised subjects by viewers has been corroborated experientially by media professionals and by academics engaged in media production teaching. In experimental studies, however, the importance of these variables remains to be clearly established.

In Coldevinís reviews (1979, 1980) of studies where variables such as subject framing, eye contact, and camera position and movement were experimentally tested, inconsistent relationships were found between technique and viewer response. Baggaleyís studies on eye contact versus no eye contact video presentations (Baggaley, Ferguson, and Brooks, 1980) showed that viewers regarded the no contact presenter as being more informed and trustworthy, a result he attributed to subjectsí long-established home viewing experiences of "expert" interview subjects being recorded as they speak directly to off-camera reporters and talk show hosts. It is interesting to note how these responses validate the common advertising production practice of having actors, playing the roles of "real people" in commercial testimonies, deliver their lines off-camera with no eye contact. The presentational venue of the telecourse instructor, however, is not that of the interviewed expert but compares more closely to that of a regularly scheduled news anchor or program host. Frequent presentations by news anchors who introduce, explain, and summarize units of information (news stories) bear a strong similarity to the ongoing semester-long presentations by instructors who also introduce, explain, and summarize units of information. The experimental significance of variables such as eye contact and eyeline has been difficult to test with adequate control, and more empirical study is needed. Additionally, experimental studies to date have employed short video and film clips for viewing, so the longitudinal effects of these variables in the development of parasocial interaction over time has not been tested.

Strategies for Creating and

Maintaining the Illusion

If televised instruction appears to be having less than desired rates of student retention and performanceña debatable contentionñthen more is required in production methodology to improve parasocial interaction between teachers and students. If the receptivity of telecast instructional delivery can be improved by the positive application of specific production techniquesñeven if these techniques create mere illusionsñthen these strategies need to be made use of by distance education practitioners. The power of mere illusions should not be discounted. Sylvie Richardís account of her first experience as a telecourse student gives a good idea of the power of televisionís illusions. Her TV instructor often encouraged students to write to him with additional questions (there was no live audio interaction available). This she did several times, but received not one reply. At the end of the course, she mentioned this neglect of her inquiries to the teaching assistant at her location, who apologized on behalf of the distant professor, who had died three years earlier (Gutenko, 1991).

The production variables of unified eyeline, eye contact, and perspective are essentially independent of a media subjectís personal attributes. Achieving unified eyeline and desirable level of intimate perspective are the outcomes of staging, and involve the constructive placement of cameras, monitors, and people. Eye contact is a behavioral or performance outcome often encouraged and facilitated by staging design.

 

Unified Eyeline

The unified eyeline attribute in production is simple to describe, simple to achieve if it has been allowed for in studio or classroom design at the outset, but is not so simple to extol as an obvious factor in facilitating the illusion of visual interaction. Having a unified eyeline means assuring that all the participants in the interpersonal exchange (instructor, on-site students, and distant students) are situated so as to be at one common eye level, so that everyone is sharing the same horizontal visual plane. This is a fastidiously attended to variable in television and film, achieved by having the chairs occupied by persons on television sets placed on platforms, or by having the camera lowered or raised to different heights, so that the camera lens is on the same horizontal plane as are the eyes of the talent. Rarely are persons looked up to or down onto, unless it is to suggest some important psychological displacement or lack of equivalency between that which is being viewed and that which is viewing. Many video and film texts extensively discuss the potential emotional and psychological consequences of varying camera angle (Zettl, 1990).

Various deviations from the unified plane of vision may be found in televised teaching. The instructor may be on a stage, looking down to the on-site students. The instructor may also be looking up to a high-mounted camera, or to the side to see her or his monitor. Because of such subject and camera position disparities, both on-site and distant students may see their instructorñand themselvesñat odd and disparate angles as they are captured by cameras shooting sideways or down at them. This conflict in viewing angles is what must be avoided.

Cameras are not intentionally positioned to convey disparity between the participants of the course, but perhaps to assure a clear shot of the instructor over the audience, to allow for equipment security, ease of servicing, efficient cable runs, or to conform to the seating and podium or lectern layouts of existing rooms. Instructors and production personnel should give thought to the psychological effects of the images beyond the pragmatic setup of cameras, monitors, and people.

Maintaining unified eyeline is complicated because it involves the proper arrangement of telecasting site elements and a compatible arrangement of receive site monitors, return video cameras, and seating. At various receive sites, monitors and cameras may be positioned above viewerís heads, at eye level, or below eye level. It is difficult to control equipment configurations for locations that are outside of the producerís jurisdiction, and careful maintenance of eyeline at the origination point can be easily undone at the other end.

Perspective

Perspective is probably the most subtle factor of the three strategies under discussion. The distance from a viewer to things viewed is revealed by the sizes of the things viewed. Close things appear large and distant things appear small. In addition, perspective reveals viewing distance through the exaggeration or diminution of relative differences in size between the viewed objects themselves.

A cluster of trees at a distance may appear to be composed of similarly sized trees, but when seen up close, the nearest trees occupy a larger volume of visual space than do the trees further way in the cluster. Foreshortening, converging lines, and other compositional terms relating to the enlargement of close objects and the diminution of far objects are all references to perspective.

Like film directors and cinematographers, teachers should be particular and deliberate about specific camera-to-subject distances. A subject shot from a distance, even if enlarged with a telephoto lens, still looks like it was obtained at a distance; it looks "flat." Film directors are convinced that viewers respond to the effects of perspective and can sense the distance from the subject at which they have been placed by the director or cinematographer. This use of perspective regulates the spatial intimacy of the viewer and subject. As a "classic" manipulative device in cinematography and videography, perspective is noted and visually illustrated in film and video production texts and guidebooks.

To reiterate, perspective is the most subtle factor of the three strategies that are under discussion. An instructor who is shot with a camera mounted in the back of a room will have the flattened perspective that would be seen by a viewer seated at the back of this room. Magnifying this image with a telephoto lens negates this perception to a great extent, especially if the instructor is in a relatively empty space. Perspective is most obvious if there are other objects visible in front of and/or behind the instructor; whose variable size relationships would be revealed.

In developing a visual strategy to foster parasocial interaction , it is essential that perspective be dealt with, and that the camera dedicated to the instructor be located at an appropriate distance for the level of intimacy desired. The perception of "personal space" is recognized similarly by members of a particular culture. A white Canadian would feel the intrusiveness of a camera that is situated two feet from the instructorís face, and would apprehend the detachment of a camera that is thirty feet away from the instructor (if perspective cues are visible). As a general rule, the distance that is acceptable to most people in a live face-to-face conversation is an appropriate camera-to-instructor distance,

from three to six feet, depending on desired intimacy.

Regarding appropriate distance, an informal clue to instructional effectiveness and student engagement that might be useful to consider here could be called "migration." On the first day of the semester, most live venue instructors will find, even in a half full room, that some students have taken seats in the first row, some in the middle, and some are as far away as it is physically possible to be. Clearly, a statement about a preferred level of interactivity has been made by all. Some instructors mayñperhaps counterproductivelyñtend to assign "motivated" and "unmotivated" labels to their students on this basis. Instructors may then consider any migration of students around the room during the semester to be a useful interpersonal indicator of instructional success. Unfortunately, all students in the telecourse venue are placed in the same "row" by the camera, and migration ceases to be a possible instructor feedback cue. Motivated and unmotivated students all are assigned the same seat, and that will be the seat the camera chooses.

Eye contact

The seasoned media professional learns how to regard the black hole of the camera lens with compassion, cordiality, passion, deep respect, and any other projection of emotion or argument required by the objectives of the delivered message. The media professional learns how to hear and see "out of the corners" of their senses; television monitors and other devices are watched with peripheral vision, and verbal instructions from directors are heard without missing a beat in performance. Their eye contact with the viewer does not waver, except when their visual connection needs punctuation, for effect. These skills construct an exceptional, highly developed illusion of rapport, and such skills are expected of one whose professional career obliges them to project a precise image. Few telecourse instructors can carry off this level of on-camera comfort.

The great challenge for the television course producer or director is finding a means by which non-media professionals, the instructors, can project themselves through the glass wall with some of this illusion of connection that is delivered adeptly by the professional talent. Some have resurrected the mid-sixties "candid classroom" design, expecting that hidden cameras will somehow put the non-media person at ease. Hidden cameras make eye contact with distance students an impossibility, and the candid classroom tends to treat the remote observers with a view like a security camera watching a shoplifter in a convenience store. There is a complete visual disconnection because the instructor fails to acknowledge the television viewers. It is also debatable that a camera that is known about but hidden from sight reduces "stage-fright." Are television instructors ever televised without their being aware of their being televised? One contribution of staging is to provide simple and supportive motivations to lead the instructor to look towards a camera that is within visual range. An appropriate feature to draw the instructorís gaze is a monitor which is providing valuable guidance.

Teleprompting devices can be positioned so that a camera can shoot through the display screen from behind, allowing the on-camera talent to read text material while maintaining eye contact with the camera. Aside from script display, most prompting screens can display any video image that is desired. Script display is merely a conventional use. Teleprompters can display the graphics being shown to the class, the video images being transmitted, video being sent back from remote sites, or any other picture source that is desired. Many instructors do not read lecture notes verbatim, and so the conventional application of the script prompter would not necessarily be required.

A fixed, non-moving prompter screen, placed at an appropriate distance from the instructor, can provide the instructor with whatever monitoring function they would be comfortable with; it might display only the current time, for example. The instructorís camera could operate unseen behind the display.

Often the most thoughtlessly positioned devices in television classrooms is the instructorís television monitor. This monitor is often placed to the side of the room or concealed in the podium. As ap

consequence, the instructor is frequently looking off to the side or down into the podium as they monitor their graphics and presentation. This leads to breaks in eye contact with the instructorís camera. Again, it must be noted that professional media talent learn to view monitors using peripheral vision and maintain eye contact with their camera. Media inexperienced instructors tend to turn their heads and look directly at their monitors.

Providing the monitoring functions through a camera teleprompter keeps eyes naturally on target. While the instructor may be looking at notes, graphics, or videotape inserts on the monitor or prompter, the visual illusion for the distant participants is that the instructor is regarding them directly and including them throughout the progress of the lesson. The kinds of images displayed on the prompter would be determined by the instructor, and she or he could select different monitor displays on a switcher at the podium.

Over-the-shoulder or other direct monitors provided for the benefit of on-site (studio) students should be dedicated to show only graphics and other instructional illustrations. When on-site students view themselves, especially if they are speaking, the experience can be disconcerting or provocative. Self-observation may interfere with the cognitive formulation of a response or question as they unavoidably respond to and evaluate their appearance on camera.

An instructor may require the presence of students in the television studio or classroom. If so, placement of on-site students should reinforce the staging of the production; they should be few in number and grouped around the prompter so that the instructorís sight will remain directed towards the vicinity of the prompter and camera. It is counterproductive to try and make one facility work well both as a conventional, large volume classroom and as a telecasting environment because both objectives will be compromised.

A Telecasting Environment for Creating and Maintaining Visual Interaction

Applying the foregoing considerations to the design and construction of a television teaching origination space will result in an environment that is neither a conventional broadcasting studio nor a simple variation of a classroom or lecture hall. Most telecourse environments still look like either classrooms or television studios, or some Frankenstein creation of stitched-together components and concepts derived from these two preceding communication venues. To work effectively and appropriately, a telecourse production environment must have its own unique form, one that has evolved and grown in response to a conceptual synthesis, just as an organism develops from the imperatives of its genetic code. It wonít work to just bolt parts together from a Mack truck and a Mercedes. To accommodate the requisite technology and production variables, something that is specifically appropriate will emerge; form must follow function.

The illustrated floorplan for a hypothetical teleteaching environment suggests a general physical layout that is designed to support the creation and maintenance of the illusion of visual interaction. This plan anticipates the use of small remote controlled CCD cameras at all positions, including the camera dedicated to the instructor. This camera would be located in an enclosure at the front of the room in the middle of the first row of on-site student seating. The enclosure also would house the prompter monitor, which can be larger and much heavier than conventional camera mounted units because it is stationary and floor mounted. This position will provide the desired perspective, and this particular camera would also have controls which would allow direct control by an operator. The instructor-dedicated cameraís focus, focal length, and axial movements could be guided by an RF transmitter included in the instructorís talkback earphone, which would provide a target reference for a "shotbox" based camera tracking system. There are several variations on tracking systems, and the most appropriate choice would depend on factors outside the scope of this discussion. The enclosure would be large enough to accommodate a seated videographer in case of a tracking system malfunction, and access to the enclosure would be from the video control room and out of sight of the classroom.

Situating teacher and studentsí eyelines, camera lenses, and monitors on the same visual plane requires careful rather than casual placement of the latter hardware items and necessitates limitations in seating since this configuration is designed to optimize the visual access of the distant student. Although this site plan may appear to be demanding in terms of construction and investment, it is no more so than many far less appropriate and workable alternatives. It would, however, be less conveniently converted back into a video studio or classroom if the distance education enterprise at hand is abandoned.

A First Course in Televised Distance Education

How many people responsible for the production, design, politicking, and promotion of televised distance education have taken a complete television course themselves? Many have sampled snippets and excerpts of selected highlights from promotional videos and hardware demonstrations, tuned in for short workshops and the like, but few have completed a full course in order to understand as a consumer this educational product that is being so skillfully packaged, so aggressively marketed, and so widely distributed. All teachers learn to teach by having been taught, by having been students themselves. There may be a considerable infusion of theory and guided practice, but we must admit that we have adapted our own personal criteria and techniques from the many models we have encountered as consumers of education. In teleteaching, however, there are no effective life-long models. Where have we learned, cognitively, experientially, and intuitively, what we need to know about being taught at a distance?

The technological determinists in the televised distance education field point towards newer and more expensive hardware solutions to counteract and amend the shortcomings of this "new" approach to education. They should remember that sophistication and quantity of their tools is less important than the skill and intelligence with which they are used to create an effective product. As Tony Bates of the Open University stated, "television is not just a delivery system . . . itís a whole different way of instructing" (Bates, 1979). No recommendation has been made here for the acquisition of a new "state-of-the-art" technology, but only for a careful application and integration of existing tools, and an awareness and appreciation of the formidable wall that stands between the teacher and the distant student.

Notes

1Identifying the widespread efforts to adapt the television medium into a significant teaching system is somewhat arbitrary. The first wave of development was certainly in the late 1950s and early 1960s, when, in the United States, National Educational Television (NET) and many associated educators utilized their UHF allocations to deliver live courses. The high cost of production (there were no industrial or "prosumer" grades of equipment then; it was all enormously expensive and complex), and the encroachments made by vociferous commercial broadcasting interests with designs on license allocations and frequency reservations doomed these efforts. There was hardly time to fully evaluate the instructional design and methods being used. The second wave started when videotape recording of an affordable nature was marketed (first 1/2 inch reel monochrome and then color Umatic) and media departments turned this technology towards classroom recording and proprietary instructional materials production.

2KSHB 41, an independent television station in Kansas City, Missouri, experimented with a newscast format where the anchors and reporters were never seen on camera. 41 Express screened only news event footage and graphics. It never developed an adequate, ratings-measurable audience before it was canceled.

Response from Leonard J. Shedletsky

While Gutenko's insistence upon eye contact, unified eyeline, and perspective (which presumably enhanced the illusion of connection, rapport, or interaction) is persuasive, the actual evidence presented is far more uncertain. Gutenko's question, "What is the difference, then between the ëtalking headí that holds viewersí attention and the one that does not?" remains to be answered. Moreover, the illusion of interaction (or, imagined interaction) is critical here. How does the illusion of interaction operate as an intrapersonal experience

Homepages: Communication Resources (Aitken) | Communicating (Berko) | Intimate Communication | Intrapersonal | Public Speaker/Public Listener (Berko) | "Speech Communication Teacher" (NCA)