Billet publié sur Twitter/X le 19/05/2021 03:56

Résumé

Have you ever been in a conversation that lasted too long or ended too soon? According to recent findings from Mastroianni et al., conversations rarely end when people want them to end. I propose a framework for studying conversation and outline new questions that follow from Mastroianni and colleagues’ generative studies. Have you ever been in a conversation that lasted too long or ended too soon? According to recent findings from Mastroianni et al., conversations rarely end when people want them to end. I propose a framework for studying conversation and outline new questions that follow from Mastroianni and colleagues’ generative studies. Imagine talking with someone you find utterly boring but you cannot seem to extricate yourself from the conversation. Or imagine being engaged in a fascinating conversation and the other person says, much to your chagrin, ‘Well, I’d better go’. If these imagined situations feel familiar to you, you are apparently not alone: recently published research from Mastroianni, Gilbert, Cooney, and Wilson [1.Mastroianni A.M. et al.Do conversations end when people want them to?.Proc. Natl. Acad. Sci. 2021; 118e2011809118Crossref PubMed Scopus (8) Google Scholar] provides evidence that conversations almost never end when people want. Across two studies, participants reflecting on their last in-person conversation or a conversation that had just occurred in the laboratory reported that only 15.6% of the time did the conversation end when they desired it to end (n = 1173 pre-exclusions; https://osf.io/8k4aj/). The other 84.4% of the time, participants’ desired conversation duration deviated from the actual duration by nearly half of the actual length of the conversation. At least on the metric of achieving one’s desired end time, people appear to be shockingly bad at ending conversations. On the one hand, conversations’ undesirable endings are puzzling in light of the substantial experience humans have socializing with one another. People spend almost a third of their waking hours talking with or listening to others [2.Milek A. et al.“Eavesdropping on happiness” revisited: a pooled, multi-sample replication of the association between life satisfaction and observed daily conversation quantity and quality.Psychol. Sci. 2018; 29: 1451-1462Crossref PubMed Scopus (21) Google Scholar]. In the entertaining words of Mastroianni et al.: ‘an alien observer could be forgiven for concluding that human beings were mainly designed to eat, sleep, and vibrate their vocal cords in each other’s presence’. Indeed, scholars have noted the ease with which two conversation partners align their linguistic representations, with one paper proposing that ‘humans are “designed” for dialogue’ [3.Garrod S. Pickering M.R. Why is conversation so easy?.Trends Cogn. Sci. 2004; 8: 8-11Abstract Full Text Full Text PDF PubMed Scopus (405) Google Scholar]. On the other hand, conversation is undeniably complex. Consider, in detail, how a conversation works. A conversation is an attempted exchange of mental content, thoughts, feelings, beliefs, and opinions, between two or more people. One person decides to start talking to another. To do so, they must translate what is on their mind into language (such as words, gestures, and/or facial expressions) intended to be understood by the other person who then (typically) attempts to interpret their partner’s communication and respond accordingly. This imperfect translation between two minds through the medium of language, with one person trying to express thoughts and the other trying to interpret, may continue for several turns, until one person, both people, or an external event end(s) it. Put this way, challenges arise at every step. First, how do people determine whether others really want to talk with them (e.g., [4.Epley N. Schroeder J. Mistakenly seeking solitude.J. Exp. Psychol. Gen. 2014; 143: 1980-1999Crossref PubMed Scopus (100) Google Scholar])? How often do people ‘misspeak’, failing to perfectly translate their clear thoughts into equally clear words, or, even when their language precisely captures their intended meaning, how often is it still misinterpreted by the conversation partner (e.g., [5.Chang J.P. et al.Don’t Let Me Be Misunderstood: Comparing Intentions and Perceptions in Online Discussions.in: WWW ’20: Proceedings of the Web Conference. 2020: 2066-2077Crossref Scopus (7) Google Scholar])? How does a conversation partner know to respond: whether to reciprocate one joke with another or be merely appreciative (e.g., Boothby et al.’s [6.Boothby E.J. et al.The liking gap in conversations: do people like us more than we think?.Psychol. Sci. 2018; 29: 1742-1756Crossref PubMed Scopus (27) Google Scholar] proposal that conversations are ‘conspiracies of politeness’)? How does a person attentively listen while simultaneously plotting witty rejoinders? How do people manage any number of tricky conversation dynamics, from getting sufficient airtime [7.Cooney G. et al.The many minds problem: disclosure in dyadic versus group conversation.Curr. Opin. Psychol. 2020; 31: 22-27Crossref PubMed Scopus (9) Google Scholar], to appearing receptive [8.Yeomans M. et al.Conversational receptiveness: improving engagement with opposing views.Organ. Behav. Hum. Decis. Process. 2020; 160: 131-148Crossref Scopus (7) Google Scholar] and being likable [9.Huang K. et al.It doesn’t hurt to ask: question-asking increases liking.J. Pers. Soc. Psychol. 2017; 113: 430-452Crossref PubMed Scopus (50) Google Scholar], to switching topics? Most relevant to Mastroianni et al., how and when does a conversation wind down? That the preceding paragraph is composed almost entirely of questions is a testament to how deeply generative Mastroianni et al. is. It also highlights the perturbing truth that, although many papers in social psychology focus on outcomes that involve conversation (e.g., relationship satisfaction), few papers actually study the dynamics of conversation itself. Those papers that do examine conversation experiences typically measure retrospective evaluations only after the conversation ends (e.g., [4.Epley N. Schroeder J. Mistakenly seeking solitude.J. Exp. Psychol. Gen. 2014; 143: 1980-1999Crossref PubMed Scopus (100) Google Scholar,6.Boothby E.J. et al.The liking gap in conversations: do people like us more than we think?.Psychol. Sci. 2018; 29: 1742-1756Crossref PubMed Scopus (27) Google Scholar,8.Yeomans M. et al.Conversational receptiveness: improving engagement with opposing views.Organ. Behav. Hum. Decis. Process. 2020; 160: 131-148Crossref Scopus (7) Google Scholar,9.Huang K. et al.It doesn’t hurt to ask: question-asking increases liking.J. Pers. Soc. Psychol. 2017; 113: 430-452Crossref PubMed Scopus (50) Google Scholar]), providing little insight into the ‘black box’ of what happens in the conversation itself. With advents in natural language processing and conversations becoming increasingly common in virtual environments, where they can be more easily recorded, opportunities for prying open the black box of conversation abound. By studying the end of conversations, Mastroianni et al. thus make their mark on a topic that is largely unexamined in social psychology, but popular among lay audiences. A quick Google search reveals multiple websites that offer advice on dozens of unique ways to end conversations, ranging from nonverbal cues like checking one’s watch or looking toward the door, to statements like ‘I’d love to keep talking but I’m sure you have other things to do’. Many questions remain. At the risk of writing another paragraph replete with questions, I select only a few. For one, what does it mean to want a conversation to end? Conversation ebbs and flows. The desire to end could be transient; one moment of boredom could quickly switch to interest. Moreover, choosing to end a conversation prematurely could be a ‘mistake’ in the sense that people who leave conversations may have gained more utility if they had stayed. Relatedly, conversations that end too early are likely to be psychologically different in several ways than those that last too long. Last, if ending conversations is a ‘coordination problem’ with both partners trying to read the other’s preferences, as Mastroianni et al. propose, are conversationalists in closer relationships better at solving the problem? Table 1 summarizes these questions and others worth exploring.Table 1A Nonexhaustive List of Potential Research Questions for Each Stage of ConversationConversation stageResearch questionsStart of conversation•How does a person know if another person wants to talk? How do people gage the likelihood of social rejection? What are the nonverbal cues that signal openness to engagement?•How does technology (e.g., smartphones, networking apps) influence people’s willingness to engage?•Which cues do people use to predict the success of a conversation before it even begins (e.g., purpose of conversation being instrumental or social, relationship with conversation partner, amount of time to converse)?•Conditional on deciding to start, how does a person do so? Which topics are most common for starting conversations and which are more successful for maintaining them?Middle of conversation•How do people strive to keep each other’s attention and interest? How does question-asking influence interest? When do people cater more to their own interests or the other person’s interests? More broadly, what makes a conversation interesting?•How much do people disclose, when do they disclose, and what do they disclose?•How do people decide whether and when to switch topics in the conversation?•Which cues do people use to take ‘turns’ in a conversation: how do they know when it is their turn to speak and signal to the other person that they should speak?•How does alignment in conversationalists’ goals (e.g., coordination, explanation, hedonism, connection) influence their enjoyment of a conversation?End of conversation•How do people’s desires to continue or end a conversation change over the course of the conversation?•To what extent do concerns about the other person’s feelings influence people’s desire to end a conversation? In general, what considerations do people have when determining whether to end a conversation?•Do people have ‘stock’ ways of ending conversations? Are some methods of ending conversations more common or more effective than others?•How does the way a conversation ends influence beliefs about how the rest of the conversation went?•To what extent do people end conversations ‘too’ early or late, in the sense that they would have gained more utility if they had stayed in the conversation longer or left the conversation earlier? Open table in a new tab Conversation is one of the most ubiquitous activities in which humans engage and for good reason. It is the foundation upon which people build the relationships necessary for their psychological well-being and physical health [10.Holt-Lunstad J. et al.Loneliness and social isolation as risk factors for mortality: a meta-analytic review.Perspect. Psychol. Sci. 2015; 10: 227-237Crossref PubMed Scopus (1793) Google Scholar]. Yet social psychologists know little about conversational dynamics. Though each conversation phase, the beginning, middle, and end, deserves further study, Mastroianni et al. make valuable inroads toward understanding conversation endings in particular. Their work indicates that people who strive for a satisfying conclusion to their conversation are likely to be unsatisfied with the timing of the conclusion. Like the final line of a paper, the final sentence of a conversation has the potential to be most impactful, but might be hardest to get exactly right. No interests are declared.