Cognitive mechanisms for inferring the meaning of novel signals during symbolisation.

Justin Sulik,Mitchell Rabinowitz

doi:10.1371/journal.pone.0189540

Justin Sulik, Mitchell Rabinowitz

Open Access

https://doi.org/10.1371/journal.pone.0189540

Copy DOI

Abstract

As participants repeatedly interact using graphical signals (as in a game of Pictionary), the signals gradually shift from being iconic (or motivated) to being symbolic (or arbitrary). The aim here is to test experimentally whether this change in the form of the signal implies a concomitant shift in the inferential mechanisms needed to understand it. The results show that, during early, iconic stages, there is more reliance on creative inferential processes associated with insight problem solving, and that the recruitment of these cognitive mechanisms decreases over time. The variation in inferential mechanism is not predicted by the sign’s visual complexity or iconicity, but by its familiarity, and by the complexity of the relevant mental representations. The discussion explores implications for pragmatics, language evolution, and iconicity research.

Highlights

Humans are readily able to generate hypotheses about the meanings of novel signals in communicative games such as charades or Pictionary, even when the perceptual form of these signals changes over time
There is a significant negative effect of turn on perimetric complexity (β = −0.199, SE = 0.033t = −6.029, p < 0.001, bootstrapped 95% confidence intervals (CIs) [-0.263, -0.134])
The data shows a significant decrease in subjective reporting of an ‘Aha!’ experience over time in a graphical novel signalling game. This constitutes a predictable behavioural difference along the time-course of symbolisation. This behavioural measure is diagnostic of a difference in cognitive processing, and that recruitment of these cognitive mechanisms has previously been observed in novel signalling tasks and in relevance-deciding tasks

Summary

Introduction

Humans are readily able to generate hypotheses about the meanings of novel signals in communicative games such as charades or Pictionary (hereafter, ‘novel signalling tasks’), even when the perceptual form of these signals changes over time. Garrod, Fay and colleagues [1, 2] investigate repeated Pictionary-like games, where participants draw graphical representations of cues which a partner has to guess, and where this process is repeated over several rounds with the same items. They show that, if the participants are able to interact while playing, the initially iconic signs become less iconic or more conventionalised, resulting in symbolic signs (iconic signs share perceptual properties with their referents, whereas symbolic signs are arbitrary, lacking such a link, or are those for which convention or habit play a role in their interpretation [3]). After all items in a round were guessed, the participants swapped roles and repeated the game with the same items

Methods

Results

Discussion

Conclusion