Submission Summary
Share...

Direct link:

Chunking - the cognitive basis of a dynamic grammar

Sun, March 19, 1:50 to 2:20pm, Marriott Portland Downtown Waterfront, Willamette

Session Submission Type: Paper

Summary

Humans make sense of incoming linear speech flow by intuitively breaking it down into manageable chunks. We test hypotheses based on the Linear Unit Grammar (Sinclair & Mauranen (2006) and report findings from a test where participants marked chunks on a web-based application while listening to audio clips.

Abstract

We hypothesise that humans make sense of incoming linear speech flow by intuitively breaking it down into manageable chunks. Our point of departure is the model of Linear Unit Grammar (LUG) developed by Sinclair and Mauranen (2006; see also Mauranen 2012, 2016) and in this presentation we discuss recent findings from testing the feasibility of the model. Unlike previous research based on LUG (Cameron 2013; Carey 2013), which is essentially corpus-based, our aim is to test its cognitive hypotheses. LUG relates to other grammars seeking to describe dynamism in online speech (Brazil 1995; W. O’Grady 2005; G. O’Grady 2010), but makes more radical assumptions about both temporality and processing. Related research also comes from cognitive processing of speech (e.g. Christiansen & Charter 2015). The properties of on-line chunking should be primarily determined by (1) the linearity of text; (2) constraints of human cognition and (3) mechanisms of meaning construction. To examine them, we conducted an experiment where participants were asked to listen to short audio clips of natural language interaction and follow them from transcripts. Their task was to mark boundaries between chunks as they listened by putting a boundary where they felt a chunk ends. Each extract was followed by a comprehension question to correlate chunking behaviour with understanding. The chunking task was designed as a web-based application for tablets, which records all boundaries marked by participants. These boundaries were then analysed both individually and in the aggregate to see (1) whether they correspond to predictions of Linear Unit Grammar; (2) which boundaries are perceptually more salient being most commonly marked; (3) which chunk types are perceived as least breakable in that participants never insert a boundary within them. Insofar as chunking is a basis for language processing, our models of learning and teaching require a radical overhaul.

Strand

Presenters