Aviation English

  • View
    82

  • Download
    1

Embed Size (px)

DESCRIPTION

An analysis of aviation English

Text of Aviation English

  • Berns Prelim 1

    An Analysis of Aviation Communications,

    Aviation English,

    and Methodology

    Kitty Campbell Laird

    Purdue University

  • Berns Prelim 2

    What research methods have been used in analyzing aviationcommunications and phraseologies? What approaches to text/discourseanalysis do they represent? In their analysis, which coding schemes areused? Review these schemes, and then assess which, if any, would beappropriate and why for the project you propose to undertake for yourthesis study.

    IntroductionThis paper will address the research methodologies which have been used in

    analyzing aviation English and communications. I will address the relevant taxonomies

    and coding schemes implemented and propose possible applications for my research.

    Research in Aviation CommunicationsThe majority of research conducted regarding aviation communications utilized

    one of several methodologies. Varieties of discourse analysis have been used to analyze

    corpora of pilot and Air Traffic Control recordings and codify the phraseology and

    structure of aviation English, survey data has been collected by self report mechanisms

    such as NASA's Aviation Safety Reporting System (ASRS) forms detailed below, and

    field studies and experimental simulations have attempted to study phenomena of

    aviation communications. While nearly ten years old, the following meta analysis

    provides a good introduction to the types of studies commonly occurring in the field of

    aviation.

  • Berns Prelim 3

    Prinzo & Britton (1993)

    This study analyzed 43 reports regarding the literature to date for pilot and

    controller communications. The make up of the reports was 45% survey data, 41% field

    studies, and 14% laboratory studies. The review addressed three major questions: 1) What

    is known about ATC/pilot voice communications and the issues pertaining to

    miscommunications? 2) What approaches have been used to study miscommunications?

    and 3) What research needs to be performed so that real solutions can be offered to the

    aviation community? The emergent hypothesis was that miscommunications occur more

    often when air traffic controllers experienced overload due to heavy traffic, frequency

    congestion, and lengthy messages. Data in the studies presented was collected from

    NASA Aviation Safety Reporting System (ASRS) forms, audiotape analysis, and

    laboratory studies. The meta analysis compared and contrasted various taxonomies

    including Kanki & Foushee's speech act coding scheme, Morrow, Rodvold, & Lee's

    taxonomy of routine vs. non-routine transactions, and Cardosi & Boole's time

    components in ATC/pilot verbal communications. The analysis concluded that audio

    taped communications data bases could be helpful in post-communication analysis and a

    globally implemented taxonomy for aviation speech acts should be established to report

    findings in a uniform and systematic way for better comparison between studies.

    Prinzo and Britton collaborated on the report; however, their analysis and

    synthesis methods were not described. The study was descriptive in nature and did not

    contribute any original research, as it only provided a synthesis of existing literature. It

    did give an accurate picture of the research that had been conducted previous to the date

    of publication regarding pilot/controller voice communications. No previous study had

  • Berns Prelim 4

    offered such a comprehensive cross-study analysis.

    Philips (1991)

    Philips explored how official phraseology of air traffic communications differed

    from natural English. He studied the official phraseology of the Civil Aviation Authority

    Radiotelephony manual in addition to research from the Ecole Nationale de l'Aviation

    Civile, Toulouse, France. The emergent hypothesis of the study was the International

    Civil Aviation Organization (ICAO) phraseology has a special purpose sub-grammar. The

    corpus consisted of 541 phraseological utterances and 36 structural modifications. Philips

    taxonomized and coded the phraseology and compared the sub-grammar to natural

    English. He determined that aviation phraseology has two sub systems: 1) English as a

    sub-grammar and 2) a context and domain dependent speech community.

    The corpus samples were limited to European standards and thus do not

    generalize to air traffic communications conducted in the United States. Although ICAO

    proposes to be an international flight organization, it does not have any regulatory power

    as an arm of the United Nations; therefore, international communications standards are

    only advisory in nature. Standardization and usage of official phraseology was

    determined via regulatory aeronautical manuals. Actual practical usage was not evaluated

    and as such the applications for real-world communications were not addressed. The

    study did highlight the need for a global topography or taxonomy of aviation phraseology

    in order to better understand flight communications.

  • Berns Prelim 5

    Morrow, Rodvold & Lee (1994)

    The study used Clark & Schaefer's collaborative scheme (1987) as a framework to

    organize what they termed "routine" and "non-routine" communications. The context and

    participants consisted of a corpus of recordings of pilot and controller communications

    gathered Three regions of the United States. Communication transactions were gathered

    from West Coast, Midwest, and Southeast, level 5 Terminal Radar Control (TRACON)

    centers. The emergent hypotheses was that non-routine transactions often lead to

    miscommunications. While 42 hours of data were collected, only 12 hours were

    "randomly selected" for analysis. Of this 12 hours, six were from approach frequencies

    and six were from departure frequencies. A taxonomy based on the collaborative scheme

    was used to code the transmissions. Analysis found 163 non-routine transactions with

    understanding problems and 120 with information problems. It was concluded that non-

    routine transactions decreased efficiency by focusing on resolving problems rather than

    allowing for the presentation of new information.

    The scope of the study was limited. While three regions of the country were

    represented, the East Coast was not. I question why the Southeast was selected over the

    East. The East has some of the busiest air space in the nation including New York City

    and Washington, D.C. Without representation from this region, the study is not

    generalizable to the entire U.S. Inter-rater reliability was accounted for by gaining

    agreement (77% - 91%) on 15 transactions containing 149 speech acts; however, only two

    raters coded the data. I would have preferred at least three raters if not more. The study

    did demonstrate the applicability of the collaborative scheme for analysis of Air Traffic

    Control communications.

  • Berns Prelim 6

    Prinzo (1998)

    Prinzo analyzed voice communications in a simulated approach control

    environment to determine if workload effected the performance of air traffic controllers.

    The study involved 24 full performance level controllers from two terminal radar

    approach control (TRACON) facilities. Her emergent hypothesis was that controllers

    under high stress or workload display higher vocal pitch, louder voice volume, and

    increased rate of speech. Data was collected from both simulations and the field. The

    simulation environment produced 13,900 transmission consisting of 33,000

    communication elements. Field recordings consisted of 1900 transactions and 5,336

    communication elements. VERBEX voice recognition software was used for initial

    transcription of the recordings. The Aviation Topics Speech Acts Taxonomy (ATSAT)

    was utilized to analyze communication elements. The study found similarities in

    communication style between simulation and field recordings.

    Participants in the study were primarily male with a male:female ratio of 9:1 at

    TRACON 1 and 10:2 at TRACON 2. Sex of the controllers was not considered as a

    covariate. There were limitations in the VERBEX voice recognition software, as the

    system was unable to accommodate non-standard language beyond the restricted

    phraseology of air traffic control. Natural language was therefore, not accounted for in the

    study. The study did demonstrate that simulations can provide relatively equivalent

    environments to real life TRACON situations. Simulators could be utilized for future

    experimentation and training.

  • Berns Prelim 7

    Morrow & Prinzo (1999)

    Researchers hoped to gain insight into the presentation of information, specifically

    looking at the effect of grouping or chunking information on memory capacity. The study

    hypothesized that "grouped" information would reduce memory load of pilots. "Paid

    volunteers" consisting of 21 males and 3 females were randomly assigned to "grouping"

    or "control" groups. Other variables included instructional type and mission sequence for

    simulation flights. Two days were devoted to the experiment with the first day including

    pretesting and familiarization training. Pretesting consisted of a demographic

    questionnaire and the administration of the Weschler Adult Intelligence Scales-Revised

    sections including Forward Digit Span score and Backward Digit Span score. On the

    second day, the missions were flown in a simulator which