Formulaic Language: How Can It Be Assessed?

Laura Vilkaite, Henrik Gyllstad

Research output: Contribution to conferenceAbstractpeer-review

Original languageEnglish
Publication statusUnpublished - 2014
EventAAAL, 2014 - Portland, Oregon, United States
Duration: 2014 Mar 222014 Mar 25

Conference

ConferenceAAAL, 2014
Country/TerritoryUnited States
CityPortland, Oregon
Period2014/03/222014/03/25

Bibliographical note

Abstract

Formulaic Language (FL) is very widespread in language and fulfils a number of crucial communication roles. But as fascinating and pervasive as FL is, its complex nature (particularly the various categories of FL and their different characteristics) raises a number of challenges for researchers and teachers alike. One fundamental issue is how to assess FL in meaningful and robust ways. Unfortunately, rigorous discussion of FL assessment is often lacking, e.g. assessment was not included in last year’s ARAL 32 issue on formulaic language. This presentation will address this gap, and draw together disparate research strands and discuss their implications for best practice for FL assessment.

The first requirement for good assessment is a clear and widely-accepted definition of what FL is, and the presentation will discuss various proposed definitions and their advantages and disadvantages for assessment. Other requirements include the presence of reliability and validity, and this presentation will highlight these measurement linchpins when discussing how FL is and could be assessed. The presentation will discuss how FL has been defined, operationalized and measured, focusing on selected examples from published research from the last 10 years. These will cover both more lab-based, online psycholinguistic studies and studies employing more traditional offline, paper-and-pencil approaches. This will illustrate how the context for the assessment has an effect on aspects like item sampling, selection of test format, and test procedures. The discussion will pay particular attention to the question of how to capture one of the key assumptions of FL, namely that of holistic storage: how do we know if the formulaic language we observe is stored and retrieved whole from memory? The presentation will finish with suggestions for improved FL testing procedures, and predictions of potential future trends in the assessment of FL.

Subject classification (UKÄ)

  • Languages and Literature

Cite this