ISSA Proceedings 2014 ~ Dialogue Types And Argumentative Behaviors

Abstract: Empirical tests of the dialogue types developed by informal logicians have been conducted recently. In this paper, we further advance this line of research by connecting dialogue types with several well-established measures in argumentation research: argument frames, argument beliefs, argument competence, argumentativeness, and verbal aggressiveness. Results indicate participants prefer the persuasive dialogue to the other types, and dialogues are well predicted by argument competence as well as the pro-social component of verbal aggressiveness.

Keywords: dialogue types, interpersonal arguments, Walton.

1. Introduction
The study of dialogues as normative frameworks has primarily been undertaken by informal logicians (e.g., Walton, 1998; Walton & Krabbe, 1995). Walton (1998) proposed a new approach to studying propositional commitments and turn-taking moves that occur during a dialogue. He argued that the concept of dialogue must be tailored so that it can accommodate (and explain) how individuals argue in their everyday exchanges. It should prescribe how arguments ought to occur and develop and it should provide criteria for assessing whether an argument has been used correctly (Walton, 1998).

Walton and Krabbe (1995) and Walton (1998) developed and detailed six main types of dialogues: persuasion, inquiry, information seeking, negotiation, deliberation and eristic. These dialogues differ depending on the initial situation that sparks argumentation and the main goal of engaging in a specific type. Persuasion stems from an open conflict that parties seek to resolve. Negotiation and the eristic dialogue also stem from an open conflict, but their goals are different; parties seek a practical settlement in a negotiation but only a provisional accommodation in an eristic dialogue. Inquiry and deliberation both stem from an open problem but differ in their main goal: inquiry seeks a stable resolution, whereas deliberation seeks a practical settlement. Finally, information seeking stems from an unsatisfactory spread of information and seeks to reach a stable resolution of the situation[i].

There is little research that examines these dialogues empirically. Cionea (2011) made a case that examining these dialogue types in interpersonal relationships can enhance our understanding of how, when, or why people employ each dialogue in their argumentative exchanges. Later, Cionea (2013) developed self-report measurement scales and tested four of the dialogues in the context of romantic relationships. In this paper, we propose developing measurement scales for the remaining two dialogue types and examining the associations (if any) that dialogue types may have with other argumentation variables. In addition, we propose that a seventh dialogue type may be feasibly added to the list developed by Walton (1998) and Walton and Krabbe (1995): information giving. Dialogues are a give and take process in which arguers seek information but also give information to the other party. Thus, we conceptualize this dialogue type as the reverse of information seeking; instead of trying to seek information from the other person, the arguer offers information to the other person. The goal of the dialogue and the initial situation that triggers it are the same as for information seeking.

In what follows we present the results of two studies examining dialogues as individual preferences that people tend to adopt in their arguments. We describe the goals of each study, the method we have employed, and our results. We conclude with a general discussion of what our research unveils about dialogue types and the potential future directions in which this line of research can be expanded.

2. Study 1
The goal of this study was to develop measures for the two dialogue types (inquiry and deliberation) not previously examined by Cionea (2013) and for the dialogue type that we propose should be added – information giving. To accomplish this goal, we created items for these three dialogues and assessed their reliability and factor structure.

2.1 Participants
Participants in the study were 189 individuals recruited from Amazon Mechanical Turk in the United States. One hundred and twenty one participants were male and 68 were female, with ages between 18 and 62 years old (M = 31.66, SD = 10.41). Participants were mostly White (n = 134), followed by Asian (n = 26), African American (n = 13), Hispanic and/or Latino/Latina (n = 11), and other ethnicities or combinations of the previous ones. Participants came from all regions of the United States, with the highest numbers from the Pacific (n = 42), Middle (n = 32) and South Atlantic (n = 34) regions. Most participants had a college degree (n = 59) or some college (n = 69).

2.2 Procedures
Participants completed an online questionnaire in which they provided consent and demographic information. They were then asked to think about what they do when they discuss, argue, or have any dialogue with another person and rate statements measuring dialogue types on a scale from 0 (absolute disagreement) to 100 (absolute agreement). Participants were compensated 50 US cents for their participation. The research was approved by the Institutional Review Board at a West South Central university in the United States.

2.3 Measures
Persuasion, negotiation, information seeking, and the eristic dialogue were measured with items developed by Cionea (2013). Persuasion dialogue was measured with six items (e.g., “I try to explain my position to the other person” or “I try to give the other person reasons for my position”), as was the negotiation dialogue (e.g., “I try to make a deal with the other person” or “I try to come up with an agreement that we can both live with”). Information seeking dialogue was measured with four items, such as “I try to find out more information from the other person” and “I try to ask the other person for the whole story.” The eristic dialogue was measured with six items, too (e.g., “I try to vent” or “I try to take the opposite position from the other person”).

Seven items for inquiry and seven items for deliberation were developed for this study. Examples include “I try to decide with the other person how we should proceed,” “I try to analyze with the other person the consequences of our plan(s),” and “I try to weigh the options with the other person” for deliberation and “I try to find the truth,” “I try to insist that we draw logical conclusions” and “I try to analyze how we move from facts to the conclusion(s)” for inquiry. Four items for information giving were rephrased from the items for information seeking (e.g., “I try to let the other person know more information” or “I try to offer the other person the whole story”).

2.4 Results
The reliability of each scale was examined based on Cronbach’s alpha. The factor structure for each scale was examined based on confirmatory factor analyses. We relied on LISREL 9.10 and the maximum likelihood estimation method to assess the model fit for each dialogue type. We also examined an overall measurement model of all seven dialogues together. Results are presented in Table 1 below. Based on the corroborated evidence from these analyses, we eliminated two items: one from the deliberation dialogue (“I try to deliberate with the other person to reach a decision”) and one from the inquiry dialogue (“I try to scrutinize all available evidence prior to drawing any conclusions”) which had lower reliability and in which their respective latent factors did not explain as much variance as they did in the other items.

The results of Study 1 indicate that the scales proposed for measuring people’s orientation towards the seven dialogue types are reliable and unidimensional. Therefore, we conducted a second study in which we examined these dialogue orientations in connection with other argumentative inclinations and behaviors widely used in previous argumentation literature.

3. Study 2
Our main goal in this paper was to examine the dialogue orientations in more depth and situate them in the argumentation literature. First, we were interested in whether people show preference for any of the dialogue types. Cionea (2013) found that individuals who argued about a relational transgression in their romantic relationships tended not to use two of the dialogues: deliberation and inquiry. Is that the case in other contexts? Additionally, the eristic dialogue may elicit different behavioral responses than persuasion or negotiation. Cionea, Hopârtean, Hoelscher, Ileş, and Straub (2013) found that people perceived persuasion could be accomplished by discussing things with the other person, not by quarrelling with others. However, individuals did not perceived debates and quarrels as significantly different in respect to their roles in people’s lives and in American society. They also engaged in both when addressing a variety of topics, such as socio-political issues or entertainment, and they indicated both forms could be appropriate when arguing with others. These results suggest that people may prefer one dialogue orientation to another depending on what function they perceive arguing serves in their lives. So, we investigate this possibility by asking,

RQ1: Do people prefer a dialogue type to others?

A second aspect we were interested in is the relationship between dialogue types. Cionea’s (2013) studies revealed that persuasion, negotiation, and information seeking tended to be associated with more positive goals, whereas the eristic dialogue was used to give voice to frustrations and dominance. Cionea, Hample, and Fink (2014) pointed out the high correlations between persuasion, negotiation, and the information seeking dialogue, questioning whether people are able to distinguish them in everyday arguments. Thus, we ask the following:

RQ2: What is the relationship between the seven dialogues?

Finally, our third and main area of interest was to examine the relationship between dialogue types and other variables studied in the argumentation literature. We decided to focus on four main areas we believe are pertinent to dialogues. The first one is argument competence. Initially operationalized by Trapp, Yingling, and Warner (1987), argument competence captures whether arguers have the appropriate knowledge and skills to engage others in interpersonal exchanges successfully. The concept has two dimensions: an effectiveness dimension and an (in)appropriateness dimension. Competence could be a good indicator of what dialogue type an arguer may choose. Competent and appropriate arguers are likely to rely on constructive dialogues, such as persuasion and negotiation, whereas incompetent arguers may rely more on eristic approaches in which they could enact inappropriate argumentative moves, such as ad hominem attacks or fallacious reasoning.

Table 1 – Study 1 Reliabilities and Confirmatory Factor Analyses Fit Indices Note: N = 305
a. Revised model without items 5 and 6 and with an error covariance permitted between items 1 and 2.
b. Revised model with errors covariances permitted between items 1 and 6 and 2 and 4.
c. Revised model with an error covariance permitted between items 1 and 2.
d. Revised model with an error covariance permitted between items 2 and 3.
e. Revised model with errors covariances permitted between items 2 and 4 and items 5 and 6.
f. Revised model without item 2.
g. Revised model without item 1 and with error covariances permitted between items 2 and 3 and items 5 and 6.
h. Measurement model with all dialogue types and previously implemented modifications for each scale included.

The second area we focused on is argument beliefs, initially operationalized by Rancer, Kosberg, and Baukus, (1992) and further refined by Johnson (2002). Beliefs about arguing represent cognitive representations of the attitudes and predispositions that people have in respect to arguing (Rancer, Baukus, & Infante, 1985). For example, if arguing is believed to be a means of solving conflict, individuals may engage in arguments with others when trying to address incompatible goals. We propose that beliefs about arguing offer useful information about people’s tendencies to select specific dialogue types when arguing with others; what one believes about arguing can predict what strategies one will adopt when arguing. For example, if arguing is believed to have dysfunctional outcomes, then individuals may be tempted to rely on an equally destructive dialogue approach, engaging in the eristic dialogue. We examine the list of beliefs that Johnson (2002) refined: pragmatic outcomes (i.e., arguing has pragmatic outcomes, such as resolving conflict), dysfunctional outcomes (i.e., arguing has dysfunctional outcomes, such as increasing tension), enjoyment (i.e., arguing is a fun experience), self-concept (i.e., arguing enhances one’s self concept, making a person feel positive), and ego-involvement (i.e., one argues because the topic is important to the person).

A third area we believed would be relevant to predicting what dialogue orientation people may take is argument frames (Hample, 2003). Frames are somewhat similar to beliefs; they reveal what people believe they are doing when they argue with others. Hample (2005) explained that frames are the initial expectations people have about arguing and, therefore, they affect the beginning stages of arguing (changes being possible as an argument progresses). We argue here that these beginning stages are captured by the dialogue type one is inclined to choose. In other words, frames capture expectations about arguing that are translated into a specific dialogue orientation to be enacted in the actual dialogue. We rely here on a revised version of the frames measure from Hample and Irions (2014) that identifies seven aspects:

– identity (i.e., arguing permits displaying one’s identity)
– play (i.e., arguing is a way to have fun with others)
– dominance (i.e., arguing is used to enact dominance or gain power)
– cooperation (i.e., arguing is a collaborative enterprise)
– utility (i.e., arguing serves a utilitarian purpose, allowing one to achieve what one wants)
– blurting (i.e., arguing permits people to say what is on their mind, without filters) and
– civility (i.e., arguing is a calm, civilized exchange).

Finally, a fourth area we propose can shed some light on people’s reliance on specific dialogue orientations consists of two trait variables that have been studied extensively in argumentation: argumentativeness (Infante & Rancer, 1982) and verbal aggressiveness (Infante & Wigley, 1986). Argumentativeness is the positive trait, indicating one’s ability to attack others’ ideas, whereas verbal aggressiveness is the negative trait, indicating one’s tendency to attack other people’s self-concept. Our reasoning here is that the tendency to approach arguments may lead people to engage in dialogues that enable them to cultivate this appreciation for arguments, such as persuasion, whereas the tendency to avoid arguments will be reflected by less arguing, perhaps even reliance on degenerated forms of arguing, such as quarrels. In terms of verbal aggressiveness, the pro-social dimension may connect to dialogues that enable this supportive communication style – negotiation or information giving – whereas the anti-social dimension may lead individuals to rely on the eristic dialogue. In light of all the considerations explained, we ask:

RQ3: Do competence, beliefs about arguing, argument frames, argumentativeness, and verbal aggressiveness predict each of the dialogues?

In what follows, we describe the method of our study and the answers to each of these three research questions.

3.1. Method

3.1.1 Participants.
Participants in the study were 286 undergraduate students at a large West South Central university in the United States. Participants ranged in age from 18 to 33 years old, M = 19.71, SD = 1.96. One hundred and three of them were male and 183 were female. Most participants were White (n = 223), followed by Hispanic or Latina/Latina (n = 19), African-American (n = 14), American-Indian or Alaska Native (n = 11), and some other ethnicities (n = 19). Most participants were freshmen (n = 101), followed by sophomores (n = 90), juniors (n = 52), and seniors (n = 40), while three individuals indicated another class standing. Students came from a variety of majors, including accounting, business, communication, energy management, health and exercise science, marketing, and public relations.

3.1.2 Procedures.
Participants were recruited from the departmental research pool, completed an online questionnaire, and received extra credit for their participation. The online questionnaire asked participants to provide consent for the research, provide demographic information, and then answer questions measuring dialogue orientations, argument competence, argument frames, beliefs about arguing, and argument traits. The research was approved by the Institutional Review Board of a West South Central university in the United States.

3.1.3 Measures.
The variables of interest were measured using a scale from 0 (absolute disagreement) to 100 (absolute agreement). Dialogue orientations were measured using the same scales as in Study 1. Argument competence was measured with 20 items (ten items measuring effectiveness and ten items measuring inappropriateness) from Trapp et al. (1987). Beliefs about arguing were measured with 24 items from Johnson (2002): four items measured pragmatic items, six items measured dysfunctional outcomes, six items measured enjoyment, four items measured self-concept, and four items measured ego-involvement[ii]. Argument frames were measured with 54 items from Hample and Irions (2014): eight items for identity, four items for play, six items for dominance, eight items for competition-cooperation, eight items for utility, ten items for blurt, and ten items for civility. Argumentativeness was measured with 20 items (ten items measuring the tendency to approach arguments and ten items measuring the tendency to avoid arguments) from Infante and Rancer (1982). Finally, verbal aggressiveness was measured with 20 items as well (ten items measuring the pro-social dimension and ten items measuring the anti-social dimension) from Infante and Wigley (1986). Reliabilities for all scales are presented below.

Table 2 – Study 2 Means, Standard Deviations, and Final Cronbach Reliability Estimates
Notes: N = 286.
Decision to omit items made after confirmatory factor analyses were conducted on each scale.

3.2 Results and Discussion
Our initial interest was to assess whether our respondents preferred one dialogue orientation to others (RQ1). We conducted a series of within-sample t-tests to compare adjacent means. Persuasion dialogue, with a mean of 82.39, was the clear preference, differing from the orientation with the next highest mean at p < .001. That dialogue type, information giving, was in turn significantly higher (p < .05) than interest in deliberation dialogues. The deliberation, inquiry, negotiation, and information seeking dialogues were not different from one another. The lowest mean of these (for information seeking) was significantly higher than that for the eristic dialogue (p < .001). So, our respondents clearly preferred to take a persuasion orientation; followed by information giving; followed by deliberation, inquiry, negotiation, and information-seeking; and the least preferred was eristic dialogue. This result provides some support for Walton and Krabbe’s (1995) claim that “the critical discussion (what we call persuasive dialogue) is the most fundamental context of dialogue needed as a normative structure” for analyzing arguments (p. 7). We also notice that this order roughly corresponds to the order one might supply if ranking the orientations on the basis of social desirability in Western cultures.

A second obvious matter of interest is the relationship among the dialogue types. To answer RQ2, we correlated the dialogue orientations. The eristic dialogue was essentially uncorrelated with the other orientations except for deliberation. This suggests that eristic and deliberative dialogues may not have been sharply distinct for our respondents, or perhaps that they saw the differences but assumed that deliberation leads to intemperate confrontation. Information seeking and information giving were substantially associated (r = .49), indicating that these were conceptually paired for respondents, as they ought to have been, given that one of them is simply a rephrased form of the other. Information seeking and information giving were both strongly associated with negotiation, deliberation, and inquiry. This is a reasonably perceptive understanding of the importance of evidence (information) to these constructive sorts of interactions. The relationship of the two informational orientations to persuasion was also positive but noticeably weaker than for the other constructive dialogues. A possible implication is that respondents felt that persuasion might also be undertaken by means of non-evidential resources (although we have no data on this point, such resources might include power, status, forcefulness, and so forth).

The final key issue concerns the relationships among dialogue orientations and the other variables that we believed might be explanatory. To answer RQ3, we conducted multiple regressions in which we predicted each dialogue orientation by the other variables in Table 2. Here we report only the statistically significant predictors in equation form, using standardized regression weights. All the multiple regression models were statistically significant at p < .001.

* Persuasion dialogue = .20 Competence effectiveness – .20 Argumentativeness avoid + .16 Verbal aggressiveness pro-social (adj. R2 = .10)
* Negotiation dialogue = .18 Competence effectiveness + .22 Verbal aggressiveness pro-social – .21 Play + .14 Cooperation (adj. R2 = .27)
* Information-seeking dialogue = .27 Competence effectiveness + .23 Verbal aggressiveness pro-social -.22 Dominance (adj. R2 = .22)
* Information-giving dialogue = .27 Competence effectiveness + .26 Cooperation -.14 Blurting (adj. R2 = .20)
* Eristic dialogue = .47 Competence inappropriateness + .15 Argumentativeness avoid + .22 Verbal aggressiveness anti-social -.18 Dominance + .19 Blurting (adj. R2 = .43)
* Inquiry dialogue = .37 Competence effectiveness + .17 Verbal aggressiveness pro-social (adj. R2 = .31)
* Deliberation dialogue = .26 Competence effectiveness + .31 Verbal aggressiveness pro-social + .13 Cooperation + .15 Utility (adj. R2 = .36)

The predictions varied in the degree to which the dialogue orientations were predicted, ranging from adjusted R2s of .10 to .43. Even 10% of the variance in a dialogue orientation was a substantial result, and some of the other adjusted R2s were very encouraging, considering that no correction for measurement unreliability was made.

Competence and the pro-social dimension of verbal aggressiveness significantly predicted several dialogue types (persuasion, negotiation, information seeking, inquiry, and deliberation). Some other variables added to the individual predictions for each dialogue. For example, the tendency to avoid arguments negatively affected one’s intent to engage in persuasion, which is a reasonable result given that persuasion would involve actually engaging with the other person. Negotiation presupposed cooperation, working with the other person as the frame of mind with which arguers approached it, again a reasonable expectation. So did deliberation, which suggests this dialogue is also perceived as a cooperative endeavour, and information giving, which implies a desire to work with the other person if one is to provide information. In addition, deliberation has a utilitarian frame associated with it, potentially due to its desired outcome of reaching a settlement. Interestingly, information giving was positively associated with blurting, suggesting some information sharing may be spontaneous, unfiltered, and unplanned. These results point to the importance that other-oriented variables (such as effectiveness, inappropriateness, or cooperation) have in the selection of dialogue types that involve the other person as well, such as negotiation or deliberation.

The eristic dialogue was strongly predicted by a self-report of inappropriateness in arguing, a preference to avoid arguing, an interest in being antisocial, and a willingness to blurt. It was contraindicated by an interest in asserting dominance. The avoidance impulse might be explained by a recent finding of Wright and Roloff (2014) that defensiveness and rumination about conflict were associated with both avoidance impulses and the desire to exact revenge on the other person.

4. Conclusions
This paper examined dialogue types in an effort to expand knowledge about the ways in which individuals use these argumentative strategies in their everyday exchanges. We tested self-report measures for each of the seven dialogues, and establish some needed connections with other argumentation variables. We conclude that dialogue types can be reliably measured based on the scales proposed by Cionea (2013) and the scales we have proposed here. More important though, we have found interesting associations with other variables that can help predict what dialogue orientation(s) people may prefer or rely on when they argue with others.

In general, individuals seem to prefer some dialogue types over others, with persuasion being the clearly preferred one. Several argumentation views and behaviors are important in predicting constructive dialogues. People’s self-report of their effectiveness in argumentation was a positive predictor for every dialogue type except the eristic one. The pro-social subscale of the verbal aggressiveness instrument also contributed positively to people’s attraction towards most of the constructive dialogue types. These two findings suggest that self-confidence and a set of appropriate argumentative intentions were fundamental to preference for the constructive dialogue types. The negative regression weights for argument avoidance, playfulness, dominance, and blurting reinforce this conclusion, as do the positive weights for cooperation and utility. The eristic orientation was predicted by a contrasting set of variables, one that is a conceptual fit to eristic interaction: it is inappropriate, antisocial, and undisciplined. Thus, our results identify suggest clear patterns exist in individuals’ argumentative behaviors, patterns that consist of related variables and inclinations.

Our study is not without limitations. First, our Study 2 population consisted of undergraduate students, which means results should be interpreted with this sample in mind. The relationships identified may or may not be replicable with other populations, but that is an area of research that future studies should pursue. Second, we asked participants to indicate what dialogue orientations they adopted in general when arguing. Participants’ responses may reflect general tendencies that people develop, but there may also be differences in the immediate orientations that people adopt in a specific circumstance, depending on a variety of factors, such as the topic of argument, the other person, and the environment in which arguers are. Such possibilities should be examined further. Finally, these dialogue orientations may constitute only the initial approaches that individuals have but that change as an argumentative exchange evolves. Future research should specifically focus on actual interactions between people and mapping out not only opening moves, but also shifts in dialogues and mixed dialogues.

NOTES
i. We have used here the exact terms that Walton and Krabbe (1995) use when describing the initial situation and main goal of each of the six dialogue types.
ii. Due to poor reliability and factor structure problems, self-concept and ego-involvement were dropped from further analyses.

References
Cionea, I. A. (2011). Dialogue and interpersonal communication: How informal logic can enhance our understanding of the dynamics of close relationships. Cogency, 3, 93-105.
Cionea, I. A. (2013). A dual perspective on the management of relational transgressions in romantic relationships. Retrieved from ProQuest Dissertation and Theses database (UMI No. 3587414).
Cionea, I. A., Hample, D., & Fink, E. L. (2014). Dialogue types: A scale development study. In D. Mohammed & M. Lewiński (Eds.), Virtues of Argumentation: Proceedings of the 10th International Conference of the Ontario Society for the Study of Argumentation (OSSA), Windsor, ON: OSSA.
Hample, D. (2005). Arguing: Exchanging reasons face to face. Mahwah, NJ: Erlbaum.
Hample, D., & Irions, A. (2014). Arguing to display identity. Manuscript submitted for publication.
Infante, D. A., & Rancer, A. S. (1982). A conceptualization and measure of argumentativeness. Journal of Personality Assessment, 46, 72-80. doi: 10.1207/s15327752jpa4601_13
Infante, D. A., & Wigley, C. J. (1986). Verbal aggressiveness: An interpersonal model and measure. Communication Monographs, 53, 61-69. doi: 10.1080/03637758609376126
Johnson, A. J. (2002). Beliefs about arguing: A comparison of public-issue and personal-issue arguments. Communication Reports, 15, 99-112.doi: 10.1080/08934210209367757
Rancer, A. S., Baukus, R. A., & Infante, D. A. (1985). Relations between argumentativeness and belief structures about arguing. Communication Education, 34, 37-47. doi: 10.1080/03634528509378581
Rancer, A. S., Kosberg, R. L., & Baukus, R. A. (1992). Beliefs about arguing as predictors of trait argumentativeness: Implications for training in argument and conflict management. Communication Education, 41, 375-387. doi: 10.1080/03634529209378899
Trapp, R., Yingling, J., & Warner, J. (1987). Measuring argumentative competence. In F. H. van Eemeren, R. Grootendorst, J. A. Blair, & C. A. Willard (Eds.), Argumentation: Across the lines of discipline (pp. 253-261). Dordrecht, The Netherlands: Foris.
Walton, D. (1998). The new dialectic: Conversational contexts of argument. Toronto, Canada: University of Toronto Press.
Walton, D. N., & Krabbe, E. C. W. (1995). Commitment in dialogue: Basic concepts of interpersonal reasoning. Albany: State University of New York Press.
Wright, C. W., & Roloff, M. E. (2014). When hurt continues: Taking conflict personally leads to rumination, residual hurt and negative motivations toward someone who hurt us. Communication Quarterly, 62, 193-213. doi: 0.1080/01463373.2014.890118