To What Extent Does Text Simplification Entail a More Optimized Comprehension in Human-Oriented CNLs?

The aim of this study is to develop a new cockpit controlled language for future Airbus aircraft by using psycholinguistic testing to optimize pilot comprehension. Pilots are aided by cockpit messages in order to deal with different situations during airc

  • PDF / 1,755,566 Bytes
  • 12 Pages / 439.37 x 666.142 pts Page_size
  • 86 Downloads / 215 Views

DOWNLOAD

REPORT


CLLE, University of Toulouse, CNRS, Toulouse, France [email protected] 2 Airbus Operations SAS, Toulouse, France {nataly.jahchan,emmanuelle.cannesson}@airbus.com

Abstract. The aim of this study is to develop a new cockpit controlled language for future Airbus aircraft by using psycholinguistic testing to optimize pilot comprehension. Pilots are aided by cockpit messages in order to deal with different situations during aircraft operations. The current controlled languages used on the Airbus aircraft have been carefully constructed to avoid ambiguity, inaccuracy, inconsistency, and inadequacy [21] in order to ensure the safety of the navigation, operational needs, and the adaptability of the human-computer interaction to different situations in the cockpit. However, this controlled language has several limitations, mostly due to small screen sizes (limited number of words and sentences) and is highly codified (non-conforming to natural language syntax, color-coded, etc.) so that it requires prior pilot training in order to achieve fluency. As future cockpit design is under construction, we might be looking at a different flexibility margin: less limitations, different screen sizes, less coding, etc. Keywords: Comprehension-oriented CNL  Controlled language  Airbus cockpit alarms  Human factors  Psycholinguistics  Text comprehension  Comprehension optimization

1 Introduction Going back to the origins of controlled languages, we would find that the main goal of the first CNLs was to facilitate communication among humans – such is the case with BASIC English, created by Charles Ogden in 1930. Readability research and controlled language production (BASIC English, PLAIN English, AECMA SE, etc.) have constantly been criticized for lack of empirical research that justify their rules and existence [8, 9, 13, 19]. Rudolf Flesch in his article “How Basic is Basic English?” [9] claims that Basic English “is neither basic nor English” and starts off with an example “If I were Mr. Churchill, I would not like being reduced to calling Hitler “a very bad man” or a bomber “an air plane sending down hollow balls full of substance with a tendency to go off with a loud noise”, in reference © Springer International Publishing Switzerland 2016 B. Davis et al. (Eds.): CNL 2016, LNAI 9767, pp. 69–80, 2016. DOI: 10.1007/978-3-319-41498-0_7

70

N. Jahchan et al.

to Basic English’s arbitrarily selected 850 word vocabulary. He criticizes Ogden for “deliberately avoid[ing] the scientific approach and not [being] lucky enough to find the key to simplicity by accident”. With time, the very nature of our modern day communication has pushed researchers to find different usages (i.e. other than CNLs for facilitating communication, mutual comprehension and ease of use) for controlled/simplified/processable/etc. languages such as automatic translation or formal notations in different domains (Industry, Academia, Government, etc.). See Kuhn [16] for a complete survey of the available CNLs and their usage. AECMA SE, more recently known as