unconventional communication

From a lantern in a church steeple to a cryptic emoji in a group chat, people can press nearly any space of actions into communicative service. But how do people convey meaning in the absence of prior agreement? If communicative acts are special cases of rational action, as Grice argued, then recognizing communicative intentions is the corresponding inference problem. I suggest this inference is made tractable by three ingredients: pre-existing priors, online social inference about one’s partner, and interactive coordination, which allows ad hoc conventions to be collaboratively established in a short period of time. On this view, conventional modes of communication are the amortized traces of many unconventional communication problems across partners, communities, and generations past.

Only humans create pictures, so why is there no consolidated field for studying graphics in Cognitive Science? A primary reason may be because graphics are conceived as reflections of percepts or direct representations of visual concepts. However, this notion falls apart in the face of simple observations about graphics, and I show how a simple embedding of graphics in a cognitive/linguistic paradigm better accounts for the structure of pictures and multimodal communication.

Standard accounts of social cognition focus on how people read observable agents, inferring others’ goals, beliefs, and preferences from their faces, bodies, and behavior. I argue that this picture is incomplete. Human behavior leaves traces in the physical world, and from early in life, children use these traces to extract rich social information from the inanimate environment — a capacity I call intuitive archaeology. By combining intuitive physics with social-causal reasoning, people can reconstruct past actions, detect others’ prior presence, and infer mental states from physical evidence alone. This system also supports communication, using objects to send targeted social messages to others (social identity, wealth, etc).

In this talk, I will briefly review three lines of research examining the scope of children’s social reasoning from the inanimate world. First, I provide evidence that children as young as four years use causal inference to detect when one person has copied another’s design, reasoning about the events that caused objects’ features. Second, I show that by age 6, children “read the room”: They understand how objects’ locations in a living space jointly reflect their owners’ preferences and social contexts. Third, I ask if this reasoning occurs on a larger scale, and show that children and adults use protective architectural features of towns (e.g. fences, walls) to make systematic inferences about residents’ traits, mental states, and societal structures. Together, these findings suggest that the inanimate world serves as a rich source of social information that humans intuitively both write and read.

Emblematic gestures, such as thumbs-up or waving, are learned signals used to transmit culturally-codified messages. Although they show striking diversity in their forms and features across cultures, all emblematic gestures are fundamentally movements meant to transmit a message between a sender and a recipient. As such, we hypothesized that emblematic gestures may be shaped to make it easier for recipients to recognize that the movements are meant to communicate. To study the shared features of emblematic gestures, we assembled a cross-cultural dataset of over 300 gestures. We found that the large majority are visually distinct from ordinary goal-directed movement, often through repetition. Our work provides evidence that emblematic gestures across cultures are shaped to readily reveal their communicative nature.

Even when they seem trivially simple, or when their content actually is trivially simple, sounds can convey quite a bit of meaning. In this talk I will present research on four types of meaning in sound: (1) attractiveness information in animal vocalizations; (2) individual-identity information in speech and song acoustics; (3) group-identity information in music; and (4) functional/behavioural information in song. These acoustic contexts, along with others that have not yet been studied experimentally, suggest that sound is a modality that can uniquely communicate ideas, emotions, and memories.