The object of research on message production is to answer the question, “Why do people say what they do?” The academic sources on this topic come from across the academy, notably cognitive science, discourse studies, and artificial intelligence, and from several specializations in communication, especially persuasion, interpersonal communication, and argumentation. Unconnected research was done by many investigators for decades, but the project solidified with the publication of Greene’s (1997b) Message production. From that point on, message production became an explicit topic for research and graduate seminars.
This work is now mainly done in the discipline of communication, which was originally framed by the classical tradition of rhetoric. The first canon of rhetoric is invention, which refers to the generation of content for a speech. Message production research is mainly relevant to this canon, although it makes little use of the ancient doctrines of topoi (the “places” from which orators can draw material) or stases (the logically required “resting places” of a controversy). For the most part, message production work is only beginning to address the remaining canons of rhetoric. Organization (the second canon) has received no attention at all, style (the third canon) is just starting to be a focus of research, and delivery (the fifth canon) has been treated to a small degree by nonverbal scholars. Memory (the fourth canon) is the chief exception because it is a continuing theoretical basis for message production theory.
The Goals-Plans-Action (GPA) Model Of Message Production
The goals-plans-action model, first articulated for the message production community by Dillard, is straightforward at its kernel. The idea is that, upon encountering a situation that calls for communication, people spontaneously form communication goals, which stimulate message plans, and the message (or action) involves acting out the final plan. The most basic hypothesis, robustly supported, is that the message will be in service of the goals, and this relationship is mediated by the plan. As the research has grown more sophisticated, however, so has the GPA model.
In the original formulation, the GPA model’s initial processual focus was on goals. These were understood to have been activated by a match between a subjective situation and the person’s cognitive stores. Little effort was made to theorize situations, although some work was done to identify dimensions of persuasive situations. Experimenters supplied respondents with an instruction (e.g., “you need to persuade a friend to give you a ride”), and simply assumed that the respondents understood the situation in the same way as the experimenter.
A notable advance was Dillard and Solomon’s (2000) conceptualization of situations as “social densities.” Just as the universe is mainly vacuum interrupted by densities of gravity and mass, so our social world contains recognizable clumps of social significance. An exploratory space vessel might aim at a particular mass, and similarly, people orient to repetitive kinds of interaction. These social densities are stored in long-term memory as nodes, which may well be labeled with a particular name (e.g., persuade, acquaint, etc.). Dillard and Solomon continue to formulate these social densities in terms of their goal structures.
Some other points of view appear in the literature. Hample (2005) suggests these omissions: obstacles, barriers, climates, and perception of both self and other. He believes that messages will be produced in respect to these situational elements, as well as to goals. O’Keefe (1988) objects to viewing goals only as subjective. Her idea is that situations have objective constraints, regardless of whether they are subjectively noticed. Persuasion, for instance, is inherently threatening to the face of the target, and a persuader who is oblivious to this will be punished, perhaps with ineffectiveness or disdain. And other scholars notice that not all communication-relevant situations actually generate messages (Berger 2004; Dillard 2004; Hample 2005).
Message production is said to be under personal control, and therefore responsive to subjective goals. Dillard’s initial formulation of the GPA specified that two sorts of goals would be activated. The first is what he calls the primary goal, which frames the situation, or defines it as a particular type (e.g., as influencing, comforting, joking, etc.). Secondary goals modify that frame, bringing to bear various other motivational issues, such as anxiety, protection of personal resources, and, most importantly, politeness. The secondary goals may be so immediately important that they overwhelm the primary goal (e.g., a person worried about giving offense might actually decline to try to persuade), but the situation remains subjectively defined by the primary goal (e.g., influence). No typological theory of specific goals is in wide use, although several have been suggested. Influence is the primary goal most often studied, because communication research on message production is a direct outgrowth of work on compliance seeking (Wilson 2002). The most common set of secondary goals concerns politeness.
Several investigators have explored the consequences of having multiple goals. Greene and his associates have shown in several studies (e.g., Lindsey et al. 1995) that people are less articulate and more hesitant when producing messages that respond to two goals, as compared to single goal instructions. In Greene’s work, the second goal is in some way inconsistent with the first (e.g., to give unwelcome information while ensuring that the other person is not offended). Samp and Solomon (2005) assess goal complexity (the number of activated goals) and goal strain (whether both pro-social and antisocial goals are activated). They find that both complexity and strain increase the number of clauses in messages. Thus, multiplicity of goals increases cognitive load, resulting in both immediate nonfluency and more eventual effort.
A plan is a projected sequence of actions that is intended to achieve a goal. Messages are held to emerge from plans, which may be conscious but are much more likely to consist of unconscious assemblies and intentions (e.g., Greene & Graves in press). Berger (1997) is the leading scholar on this topic.
The plan is the point in the message production process at which content comes into play. For a message, many of the plan steps consist of things one will (or may) say. These items are understood to pre-exist in one’s cognitive stores, and constitute a repertoire of possible messages or message elements. They might be understood as personal topoi, that is, places or resources. A plan, perhaps conceptualized as a memory organization packet (Kellermann 1991), assembles repertorial elements into sequences prior to utterance. Berger explains that one may have a simple sequence in mind, but may also have imagined a more complicated pattern that might even have various branch points at which choices must be made. Complicated situations may well require highly textured plans, but Berger has shown that complex plans reduce the fluidity of message production. He has also demonstrated that a given plan exists simultaneously at several hierarchically organized levels of abstraction, ranging from general intention down to the physical requirements of pronunciation and performance.
Berger is aware that plans change during the course of invention and interaction. The planning stage is therefore also the site of message editing (Hample 2005). Berger shows that the least effortful available changes are those first attempted. Meyer (1997) says that cognitive stores contain two key kinds of associations: those between the situation and the message, and those between messages and their outcomes. The situation–message association system generates an initial draft of a message. Once a potential message is assembled and activated, it may in turn stimulate notice of various likely consequences (e.g., cursing causes social disapproval). These consequences reflect the presence of secondary goals, particularly politeness issues. Should the consequences have sufficient activation levels, they may stimulate message revision or abandonment. Rehearsing a message plan improves the fluidity and quality of the resulting message.
Most message production research to date has a limited theoretical domain, mainly going either from one element of the GPA model to the next, or from one early component directly to the message. The main findings, which are robust, are that situationally activated goals predict messages’ apparent objectives, and that plan content predicts message content. These results are rather general, however.
Some more specific work has generated useful, if partial, descriptions of messages. Scholars have explored the fluidity of message production, focusing on onset latencies, pauses, and semantic variety. They relate these matters mainly to plan or goal complexity, respectively. Either sort of complexity increases cognitive load, interfering with ease of expression. Samp and Solomon’s (2005) research program explores message embellishment, length, and focus (e.g., on self or other), linking these outcomes to goal strain and goal complexity. Both strain and complexity lead to longer and more elaborate messages. The obstacle hypothesis (Francik & Clark 1985) connects message content and phrasing all the way back to situation, so that requests acknowledge the situational feature most likely to impede a favorable reply (e.g., “If you’re not too busy, could you find me a map?”). Repertoire features (e.g., degree of attention to face or argumentative issues) reappear as the parallel message features across topics. Not one of these research projects, however, spans every key component of the GPA model, and collectively their message characterizations are somewhat unconnected.
The message production project is essentially a cognitive one, developing theories that are intended to apply with equal force to any competent person with an ordinary cognitive system. Distinguishing among people on the grounds of personality or cognitive differences has not been a central priority. People are certainly understood to have different goals in similar situations, for instance, but the theoretical impulse has been to say that whatever the goal is, it will be subjected to essentially the same cognitive dynamics as any other goal, by any other person. Nonetheless, some work has used individual difference measures as ways of understanding which goals, which plans, which revisions, and which messages will appear.
When people react to a situation, they act and think in response to the subjective situation, not the objective one. Wilson (2002) shows that people exposed to what was apparently the same situation form different interaction goals, and that this difference is traceable to their interpersonal differentiation levels. The usual theory of situation-to-goal process says that goals will be activated according to how well they fit situational features, how strongly associated the goal and situational features are, and how recently the goal has been activated. Several researchers have explored how stable individual differences affect these matters, especially fit and strength.
Personality traits are understood as indicating what interaction goals are chronically accessible (i.e., strong). Introverted people will commonly activate avoidance goals, and other-oriented people will readily form relational and identity goals. Meyer (2005), for instance, shows that people high in verbal aggressiveness are less likely to reject potentially offensive messages, because relational goals are not chronically accessible to such people. Berger (1997) reports that high self-monitors produce more complex plans, because they are more oriented to their interactional partners. Research connects message behavior to argumentativeness and verbal aggressiveness (Rancer & Avtgis 2006). Other trait measures have been associated with editorial behavior, predicting whether people will edit their messages at all, and if so, on what grounds (Hample 2005).
Individual differences other than personality traits have also been investigated. Interpersonal construct differentiation is a key variable. Other non-trait measures that have garnered some attention by message production researchers include cognitive speed and capacity, age, creativity, and academic ability. Several research programs have determined what beliefs or naïve theories people have about communication, and have shown that those belief sets predict various features of message production.
The bulk of message production work has studied solitary communicators, who generate either a written message or an oral monologue. The lack of attention to dyadic or group interaction has been criticized by Waldron (1997), and his objections remain forceful.
Applying the theory and results bearing on a solitary communicator to individuals engaged in interaction may not be simple (Greene & Graves in press). Conversation has its own structure, to which people almost inevitably respond. So a question creates an answer slot, a challenge calls out a defensive reply or a concession, and so forth. One’s plans must be adaptable in the instant, because every action by the interlocutor alters the situation, and so may stimulate new goal configurations or plans. The finding that the power of individual difference variables evaporates once an interaction is under way is not uncommon. Once engaged in conversation, people’s messages appear to respond mainly to what the other person says and does, rather than to private, pre-existing goals and plans. A few studies do show that people persist when the other person is uncooperative, forming more aggressive goals, and revising their plans in predictable ways. However, joint development of communicative action is not yet a substantial research topic, and neither is adaptation to topic changes. A fruitful approach to these difficulties may be that of Bruce and Newman (1978).
Prospects For Message Production Research
This article has featured the GPA model, which summarizes (or at least is compatible with) nearly all the current research. Several other theories either do not connect with the GPA very well, or go into remarkable detail about matters that are not otherwise prominent in the GPA. These include the local management of meaning theory (O’Keefe & Lambert 1995), action assembly theory (Greene 1997a), and O’Keefe’s (1988) message design logic theory. This article also omits mention of biological work on message production because such work is only beginning. The prominence of the GPA has perhaps unfairly limited scholarly attention to these interesting approaches. Many scholars and teachers also integrate material from lines of research that are not explicitly about message production, for instance work on social support or self-disclosure. This sort of research is very helpful in moving message production work beyond its roots in persuasion.
In the last decade, message production has become an important topic among communication scholars. How people generate their utterances and writings is obviously a fundamental matter, and it should not have been neglected for as long as it was. Potentially, the communication discipline could well be reorganized into three main divisions: message production, messages, and message reception. Academic tradition and inertia will probably prevent this, but the fact that it is sensible should immediately display the importance of this topic.
- Berger, C. R. (1997). Planning strategic interaction: Attaining goals through communicative action. Mahwah, NJ: Lawrence Erlbaum.
- Berger, C. R. (2004). Speechlessness: Causal attributions, emotional features and social consequences. Journal of Language and Social Psychology, 23, 147–179.
- Bruce, B., & Newman, D. (1978). Interacting plans. Cognitive Science, 2, 195 –233.
- Dillard, J. P. (1990). A goal-driven model of interpersonal influence. In J. P. Dillard (ed.), Seeking compliance: The production of interpersonal influence messages. Scottsdale, AZ: Gorsuch Scarisbrick, pp. 41–56.
- Dillard, J. P. (2004). The goals-plans-action model of interpersonal influence. In J. S. Seiter & R. H. Gass (eds.), Perspective on persuasion, social influence, and compliance gaining. Boston, MA: Allyn & Bacon, pp. 185 –206.
- Dillard, J. P., & Solomon, D. H. (2000). Conceptualizing context in message-production research. Communication Theory, 10, 167–175.
- Francik, E. P., & Clark, H. H. (1985). How to make requests that overcome obstacles to compliance. Journal of Memory and Language, 24, 560 –568.
- Greene, J. O. (1997a). A second generation action assembly theory. In J. O. Greene (ed.), Message production: Advances in communication theory. Mahwah, NJ: Lawrence Erlbaum, pp. 151–170.
- Greene, J. O. (ed.) (1997b). Message production: Advances in communication theory. Mahwah, NJ: Lawrence Erlbaum.
- Greene, J. O., & Graves, A. R. (in press). Cognitive models of message production. In D. R. RoskosEwoldsen & J. L. Monahan (eds.), Communication and social cognition. Mahwah, NJ: Lawrence Erlbaum.
- Hample, D. (2005). Arguing: Exchanging reasons face to face. Mahwah, NJ: Lawrence Erlbaum.
- Kellermann, K. (1991). The conversation MOP: Progression through scenes in discourse. Human Communication Research, 17, 385 – 414.
- Lindsey, A. E., Greene, J. O., Parker, R. G., & Sassi, M. (1995). Effects of advance message formulation on message encoding: Evidence of cognitively based hesitation in the production of multiple-goal messages. Communication Quarterly, 43, 320 –331.
- Meyer, J. R. (1997). Cognitive influences on the ability to address interaction goals. In J. O. Greene (ed.), Message production: Advances in communication theory. Mahwah, NJ: Lawrence Erlbaum, pp. 71– 90.
- Meyer, J. R. (2005). Effect of verbal aggressiveness on the perceived importance of secondary goals in messages. Communication Studies, 55, 168 –184.
- O’Keefe, B. J. (1988). The logic of message design: Individual differences in reasoning about communication. Communication Monographs, 55, 80 –103.
- O’Keefe, B. J., & Lambert, B. L. (1995). Managing the flow of ideas: A local management approach to message design. In B. Burleson (ed.), Communication yearbook 18. Thousand Oaks, CA: Sage, pp. 54 – 82.
- Rancer, A. S., & Avtgis, T. A. (2006). Argumentative and aggressive communication: Theory, research, and application. Thousand Oaks, CA: Sage.
- Samp, J. A., & Solomon, D. H. (2005). Toward a theoretical account of goal characteristics in microlevel message features. Communication Monographs, 72, 22 – 45.
- Waldron, V. R. (1997). Toward a theory of interactive conversational planning. In J. O. Greene (ed.), Message production: Advances in communication theory. Mahwah, NJ: Lawrence Erlbaum, pp. 195 –220.
- Wilson, S. R. (2002). Seeking and resisting compliance: Why people say what they do when trying to influence others. Thousand Oaks, CA: Sage.