UC Santa Cruz

Humor in face-to-face interactions regularly exploits multimodal cues such as emotional facial expressions as social cognitive tools that help us coordinate our beliefs with others’. The encryption theory of humor (Flamson & Barrett, 2008) states that humorous acts are encrypted and we need a common “key” to understand what’s funny and humor acts a marker of internal similarity. We communicate this shared knowledge with laughter, which denotes affiliative interactions and shared underlying assumptions about the world (e.g., “This is socially okay, so I can believe it’s funny”; Flamson & Bryant, 2013) while cringing away indicates an affective stance of “this is not okay”.

These interactions are prevalent in every day face-to-face conversations, but are difficult to express in computer-mediated communication (CMC). However, the regular use and presence of GIFs in online mediums such as text messaging newly affords a similar style of multi-party interactions where one’s commentary on another person’s jokes may change someone’s opinion of the joke and interaction itself. If these multimodal cues exist in face-to-face group interactions, can CMC afford the same socioemotional cuing? GIFs, which re-present looped animations of embodied actions and are commonly used as reactions to everyday occurrences (Herring, 2015) have been found to act as embodied depictions that demonstrate one’s own current reactions in text-messaged conversations (Tolins & Samermit, 2016).

With the affordances, the possibilities for action that an agent directly perceives in an object or an environment (Gibson, 1979), provided with GIFs, this dissertation aims to answer the following broad question: How does the presence of a reaction GIF change how funny people perceive “overheard” text-messaged jokes? And does the way the GIF is understood as ironic (a combination of different dimensions of irony, such as how hyperbolic, understated, sarcastic, playful, performative, and humorous people find something; Gibbs & Samermit, 2017) have a predictive relationship with how funny they might find the joke itself?

This dissertation addressed these broad question in two experiments. In Experiment 1, participants acted as “overhearers” in a text-messaged group chat where they received a joke from one member of the group, and viewed either another member’s positive or negative GIF response, another member’s positive or negative Text response, or no response at all, before rating how funny they found the joke, and how ironic (sarcastic, hyperbolic, critical, and playful) they found the response. In Experiment 2, participants did the same as Experiment 1, but also saw static or animated GIF or Text responses. The two experiments tested two hypotheses: the passive observer hypothesis, where overhearers do not integrate valence with ironic information into their judgments of humor, and the active negotiator hypothesis, where they do. Results across both experiments support the active negotiator hypothesis, indicating overhearers in text messaged conversations conduct metarepresentational reasoning and integrate valence with ironic information about responses to make judgments of humor. This finding supports the encryption theory of laughter and humor.