Donohue et al.’s (1983) Hidden Gem: Research on the Interchange of Verbal and Nonverbal Cues to Immediacy Unlocked Relational Communication, Offline and Online

The Roles of Nonverbal and Verbal Communication

One focus within interpersonal and nonverbal communication research has been called the study of channel reliance. It seeks to determine and explain which among the variety of nonverbal and verbal cues that might be used, do people tend to use, when and for what communicative purposes. This literature has been strongly affected by such classic positions as Watzlawick et al. (1967) The Pragmatics of Communication, that asserted there are report and command functions that accompany each other in any interpersonal exchange. Report is the content or substance or topic of an exchange, whereas the command is the relational aspect by which communicators signal dominance, affection, etc. According to Watzlawick et al. (and many others), report is communicated by verbal content, whereas the commend is signaled through nonverbal means. This position is often echoed, couched in different terms such as the task and maintenance dimensions of communication, task and relational dimensions, etc., and the intuitive assessment that nonverbal communication is best, most natural, or generally relied upon, for non-task relational messages.

The literature generally characterizes nonverbal communication as both innately and uniquely capable of conveying relational messages, whereas verbal communication was not. Nonverbal cues are “implicitly seen as natural or even ‘sole’ carriers of relational information, (and) subtle verbal variations that also carry relational information have been neglected,” according to Donohue et al. (1983/2023, p. 3). Empirical research seemed to support the notion that nonverbal communication had a monopoly on relational communication: Mehrabian and Ferris (1967) concluded (without recognizing a methodological error) that 93% of the social meaning of spoken utterances derive from nonverbal variations in kinesics and vocalics; only 7%, they claimed, emanated from words. Argyle and Dean’s (1965) equilibrium theory experiments showed that, when one stranger (an experimental confederate) came close to naïve participants, participants backed up. When they could not back up, participants averted eye contact in response to the invasion of space. This showed that nonverbal cue systems—proxemics and gaze in this case—were functionally interchangeable immediacy cues. Language was not part of that equation.

Research on language, at the time, mostly focused on content and fluency effects, according to Donohue et al. (1983/2023). However, their literature review revealed some countervailing positions asserting language variations’ connotation of affinity and other aspects of relational communication. These included Bradac et al.’s (1979) review of primary studies examining the lexical variables intensity, immediacy, and diversity; and Weiner and Mehrabian’s (1968) proposal that certain linguistic variations constitute verbal immediacy.

The Challenge and its Demonstration

Donohue et al.’s (1983/2023) paper challenged both the equilibrium theoretic principle as well as its incorporation of nonverbal channels exclusively. Burgoon (1978) had by this time articulated her Communicative Model of Violations of Distancing Expectations, (Burgoon et al., 1979) later to become known as nonverbal expectancy violations theory. Burgoon’s theory accounts for discrepancies that equilibrium theory, and other immediacy exchange theories, could not: Why it is that we back off from people who come very close sometimes, but, when someone is attracted to the too-close individual, they like it and reciprocate it. (If this exception to equilibrium theory’s predicted retreat from any proxemic invasion did not exist, after all, the human species could not procreate!) Burgoon’s elegant solution was to conceptualize proxemics as being governed by normative expectations, and that responses to proxemic expectancy violations are moderated by one’s appraisal of the violator (see Burgoon, 2016 for a review).

The Donohue et al. (1983/2023) experiment extended Burgoon’s theory beyond proxemics, to incorporate another form of immediacy. Not only that: The specific other form, in this case, was verbal rather than nonverbal. The experiment examined whether spatial violations toward (or away from) an individual who could not retreat from (or approach) the violator would prompt commensurate “psychological distance displayed in verbal cues” (Donohue et al., 1983/2023, p. 7). Hypotheses predicted that if a violator came unexpectedly close, and was rewarding, target persons would reciprocate immediacy verbally; but if the violator was unrewarding, the target would express nonimmediacy verbally. If a rewarding individually violated proxemic expectations retreating, the target would chase them through greater verbal immediacy.

By and large, the results of the experiment supported the most central hypothesized patterns, with one exception. Using student subjects, the experiment attempted to implement systematic variation in violators’ rewardingness by employing either a real friend of the target, or a random stranger, to enact distance violations. This manipulation was unsuccessful. Students, apparently, like other students, at least at first, and all targets responded to all violators in the hypothesized high reward patterns: More verbal immediacy when the violator either approached or retreated, compared to no-violation control conditions. Two of the four dimensions of verbal immediacy reflected hypothesized effects, and two others did not.

Impact in Face to Face and Mediated Communication Research

Despite the mixed support, these findings were breakthroughs. As many readers will know, expectancy violations theory would grow in scope and precision. Its approach to conceptualizing and operationalizing violator’s reward value would be refined by Burgoon and colleagues and many others. Most importantly, for present purposes, this study demonstrated the theory’s incredible heuristic value. It opened the door to consider all forms of immediacy behavior—verbal, nonverbal, and otherwise (later formalized in Burgoon & Hale, 1988). It established the principle that immediacy takes many forms in many channels and that these forms and channels can be functionally interchangeable in their expression of relational communication.

Not only was this benchmark important in the domain of face to face interpersonal interaction. The conceptual and empirical precedent set by Donohue et al. (1983/2023) provided a cornerstone for the development of a particular second-generation theory in computer-mediated communication (CMC), as well. CMC at that time supported only typewritten verbal messages. One of the central questions in the field was how the medium’s occlusion of nonverbal cues affected communication. Several first-generation CMC theories addressed this question drawing directly from the channel reliance literature in nonverbal communication research. Citing psychological studies of nonverbal interaction (e.g., Ekman et al., 1980) they proposed that, because of its absence of nonverbal cues, CMC would be inferior to face to face communication in its positive relational qualities. Numerous experiments, from the 1970s through the 1990s and beyond, supported these assertions (although the validity of many such studies were challenged in the ‘90s; see Walther, 2011 for a review).

An alternative, second-generation CMC theory drew directly on the principle of interchangeability between nonverbal and verbal cues of affinity that Donohue et al. (1983/2023) established. The social information processing theory of CMC (Walther, 1992) makes that principle an explicit assumption within its own theoretical framework: “Relational messages are transmitted (i.e., encoded and decoded) by nonverbal and/or verbal, linguistic, and textual manipulations” (p. 69; Assumption #4), including “the words and other written matter that appear as typed characters transmitted in CMC … verbal content, lexical variation, syntactic usage, or other feature of language that may be conveyed in written communication” (p. 82). The theory contends that individuals, in whatever setting, exploit whatever cues are available to communicate content and relational communication; but when (as in the case of text-based CMC) there are no nonverbal cues available, communicators translate the relational messages (that may otherwise appear nonverbally), into verbal content and style variations (see Westerman et al., 2008). The theory can be thought of as a thought experiment applying Donohue et al. to CMC: If one student increases verbal immediacy when another presumably likeable student physically retreats, as Donohue et al. found, will the same thing happen when that second student retreats all the way behind a computer? Given ample opportunity to exchange of immediacy cues verbally, tests of the new theory found, the answer was yes.

Social information processing theoretic research addressed a variety of cue systems and their application a variety of online phenomena. Cue systems came to include chronemics (Kalman & Rafaeli, 2010), irony (Hancock, 2004), emoticons (Derks et al., 2007), and self-disclosure and personal questions (Tidwell & Walther, 2002), among others. One study in particular demonstrated the translatability between nonverbal and verbal expressions of liking between parallel dyadic face -to-face and computer chat conversations (Walther et al., 2005): Face to face partners relied primarily on a specific group of kinesic cues and secondarily on certain vocalic cues, whereas CMC partners relied on a set of verbal cues and strategies (and a couple emoticons) to express liking. Although the sets of cues were different, there was no significant difference in the amount of affinity detected between the two settings. The principles helped illuminate how groups using plain-text electronic messages developed trust in distributed collaborations (Wilson et al., 2006), to online romances and internet date finding (Gibbs et al., 2006).

It is essentially because of its subsumption of the interchangeability heuristic from Donohue et al. (1983/2023) that the social information processing theory of CMC generated its own influence, and years after its publication it “stood the test of time and has become a stimulus for new conceptualizations of communication phenomena” (National Communication Association, n.d.).

The Donohue at al.’s (1983/2023) paper comes from a time when the field was actively exploring and debating questions about how communication works; how alternative symbols create messages. It tested a scientifically creative and courageous model, taking a counterintuitive position that was at odds with the dominant literature, identifying the logic that allowed it (or required it) nevertheless to be true, and demonstrating it in a clear and replicable way. The resonance and magnitude of its simple principle of symbolic multiplicity and interchangeability expanded the way researchers could see, hear, and read how people influence one another’s fundamental relational dynamics across contexts as well as across technologies unforeseen by the paper’s authors.

I was fortunate that Professor Burgoon gave me a copy of this paper in the 1980s when I studied nonverbal communication under her tutelage. Professor Gavriel Salomon also gave me empirical-based articles about computer-mediated communication to read, saying “Here’s what happens to interpersonal communication when there’s no nonverbal cues.” They demonstrated significantly less affinity in CMC because, they asserted, the expression of affinity requires nonverbal communication. Their ontology and that of Donohue et al. (1983/2023) seemed mutually exclusive. The inconsistency demanded articulation of theoretical and methodological explanations that accounted for when and why each competing position could appear to be correct. Social information processing theory was an answer to that question, allowed entirely by the principle of interchangeability established by the heuristic vision of Donohue et al.’s pioneering study.