Running Head: AVIATION ENGLISH CORPUS LINGUISTICS

Aviation English Corpus Linguistics: Using the Right Phraseology? Nicholas Swinehart Ohio University

1

AVIATION ENGLISH CORPUS LINGUSTICS

2

Introduction On January 25, 1990, Colombian airline Avianca Flight 52 crashed on Long Island after running out of fuel, killing 73 people on board. One of the pilots had told the air traffic controller (ATC) “we're running out of fuel sir” and “we need priority please,” but never mentioned the words emergency or Mayday. A bombing at a nearby airport and dense fog contributed to the deadliest plane crash in history, when an American flight and a Dutch flight collided on Spanish soil in the Canary Islands in 1977, but miscommunication was deemed the ultimate cause. According to the Netherlands Aviation Safety Board's final report on the accident, the Dutch flight “has [sic] taken off without take-off clearance, in the absolute conviction that this clearance had been obtained, which was the result of a misunderstanding between the tower and the KLM (Royal Dutch Airlines) aircraft” (p. 67). As a result, the word take-off is now only to be used by ATC when an aircraft is cleared to depart (“The Tenerife Airport Disaster”, n.d.). English has been the official language of the international aviation industry since 1951, but only within the last two years has the International Civil Aviation Organization (ICAO) enforced language requirements on aviation personnel. Along with a standard phraseology to be used in all radio communications, the ICAO in 2011 began requiring pilots and ATCs to demonstrate English proficiency to a specified level. The present study compares standard and non-standard usage of a particular lexical item, right, in cockpit voice recorder (CVR) transcripts to investigate sources of ambiguity in aviation English discourse. Literature Review This literature review examines the requirements set forth by the ICAO and attempts by researchers using corpus linguistics to analyze discourse in aviation English. Two documents issued by the ICAO in recent years have had a substantial impact on the aviation industry. The

AVIATION ENGLISH CORPUS LINGUSTICS

3

Manual on the Implementation of ICAO Language Proficiency Requirements (2004) established minimum English skill level requirements for flight crews and air traffic controllers. Every individual working in aviation radiotelephony must achieve a score of four or higher on a sixpoint scale determined by the ICAO. While the requirements may have been brought about by a number of accidents involving miscommunication and non-native English speakers, the document made it clear that native and non-native speakers alike are responsible for maintaining clear and concise discourse. Aviation personnel are instructed to follow ICAO standard phraseologies as much as possible and refrain from using jargon or idioms. The Manual of Radiotelephony (2007) outlined these phraseologies to be used in all radio communications to provide “maximum clarity, brevity, and unambiguity” (p. 3-2). Only one sense of the word right appears in ICAO standard phraseology: the directional sense, as in the opposite of left. Mendez-Naya (2006) investigates the evolution of right over time, but her article is useful here to examine the different ways it is used in English today. She labels the sense approved in ICAO standard phraseology mentioned above as an adjunct of direction. She also cites right’s use as a manner adjunct, meaning “correct” or “in a direct course or manner”. When used as a focusing modifier, right means “exactly, just” and modifies locative or time expressions, adverbs, prepositional phrases, or clauses (right here, right after, right where I left it). It can also be used as an intensifier to modify adjectives or adverbs (Right Honorable). Finally, right functions as a discourse marker when it is a response indicating understanding or agreement or a question tag (sentence final right?). All right is also used as a discourse marker or manner adjunct (meaning “acceptable”); it is not mentioned by Mendez-Naya, but it does appear in the corpus, either as all right or alright. Mell (1991) found extensive non-standard usage when analyzing a corpus of radio

AVIATION ENGLISH CORPUS LINGUSTICS

4

communications between flight crews and air traffic controllers. Particularly striking is his finding that only one-third of air-ground discourse related to the movement and management of the aircraft, while two-thirds related to the management of the communication itself. Mell points out that so much metalinguistic speech “reflects the tenuous nature of the links between aircraft and ground stations, and the very real risks of confusion” (p. E-9). Pilots were responsible for 43 of the 48 indirect speech acts found in the corpus, creating possible ambiguity if the intended meaning was not interpreted by the ATC. Sixty-eight percent of pilots' most common speech act, giving information about flight level, used non-standard phraseology. Mell gives the example of using the preposition to directly before a number sequence, as in “climb to 3 5 0”, which could easily be interpreted as “climb 2 3 5 0”; standard phraseology dictates that FL, or flight level, be inserted between the preposition and the number sequence. This mirrors the problem examined in the present study of using a lexical item in non-standard, ambiguous ways. A more recent use of corpus linguistics to analyze aviation English appears in Tiewtrakul & Fletcher's (2010) study that found radiotelephony misunderstandings arising most often between non-native English speakers with different L1s. 312 international flights were recorded at Bangkok International Airport, Thailand, with researchers looking for communication errors among three discourse groups: Thai ATC-Thai pilot, Thai ATC-native English speaking pilot, and Thai ATC-foreign pilot who is a non-native English speaker and does not speak Thai. In the corpus data, the non-Thai or English L1 speakers had the most communication failures, which the researchers attributed to ATC local accents. Most misunderstanding came from messages containing numbers and specific names. Since linguistic differences play such a large role in aviation miscommunication, the researchers call for closer investigation of how linguistic, cultural, and regional differences affect pilot-ATC communication at airports around the world.

AVIATION ENGLISH CORPUS LINGUSTICS

5

The use of corpus linguistics to analyze aviation English is “almost nonexistent”, according to an educator and researcher in the field (M. Prada, personal communication, March 11, 2013). This leaves many interesting avenues open to investigation. The documents issued by the ICAO are important on two fronts: they aim to ensure flight safety and they place strict requirements on all airlines that wish to fly internationally. Mell (1991) examined non-standard usage in aviation radiotelephony, but that was 15 years before the ICAO ratcheted up their language requirements. Tiewtrakul & Fletcher (2010) chose one particular airport, Bangkok International, and used corpus linguistics to provide a comprehensive understanding of communication breakdowns between flight crews and ATCs. The present study takes a similar approach, choosing a particular lexical item and examining how it is used in standard and nonstandard ways, to identify sources of ambiguity by investigating the following questions: 1. What percentage of the tokens of right in CVR transcripts are used in a non-standard way? 2. Which non-standard functions are used? Methodology Corpus The corpus for this paper consists of three CVR transcripts acquired from the US National Transportation Safety Board (NTSB) website. This paper investigates how aviation discourse deviates from standard, prescribed phraseology, so only transcriptions of flight discourse that occurred since the publication of the ICAO's Manual of Radiotelephony are included. This corpus consists of 66,818 words. Materials/Instruments This paper uses ICAO documents issued by the ICAO to establish standard aviation

AVIATION ENGLISH CORPUS LINGUSTICS

6

phraseology. The NTSB website was used to collect transcripts, and a concordancer (MonoConc Pro) was used to organize lexical items and their co-text into concordance lines. Procedure To build my corpus, I went to the NTSB website and searched through its database of accident reports. Only reports that had CVR transcripts and occurred since 2007 were chosen. The objective is to examine causes of ambiguity in aviation discourse; I am focusing on the lexical item right since it has a standardized usage (the opposite of left) but is used frequently in non-standard ways (e.g., right behind, right now, or as a synonym of correct). The token right appears 128 times, along with a 47 tokens of alright. The different uses of right were identified, tallied, and divided by the total of 175 to calculate the percent of total use. Data Collection The first type of data collected was the total number of different senses of right found in the corpus by analyzing concordance lines. Then, frequency counts were conducted for each sense. Finally, the number of standard uses was divided by the total number of tokens to find the percentage of standard usage. Results and Discussion Use

Tokens

% of Total (175)

adjunct of direction

32

18.2%

focusing modifier

38

21.7%

manner adjunct

2

1.1%

discourse marker

36

20.6%

ambiguous

3

1.7%

all right/alright

64

36.6%

There were 128 tokens for right and 47 tokens for alright for a total of 175. The focusing modifier, discourse marker, and all right/alright each appeared more frequently than the standard

AVIATION ENGLISH CORPUS LINGUSTICS

7

use, the adjunct of direction, which appeared 32 times for 18.2% of the total. All right or alright made up one-third of the tokens, appearing twice as often as the standard usage. The use of right in non-standard ways creates opportunities for misinterpretation. One example many people have experienced is someone saying “turn right here”, where it is unclear whether right is meant to be an adjunct of direction or a focusing modifier. In high pressure situations in the cockpit, if someone says “do X right now” the pilot might mistakenly take right as most salient and perform an inappropriate action. Three tokens from the corpus could not be categorized, even when looking at the surrounding context. One instance was in the phrase “that's far enough right?”; it is unclear whether the adjunct of direction or question tag is intended. Another example was “too right,” without enough surrounding context to determine whether it functions as an adjunct of direction or a manner adjunct. The phrase “right off the left wing” uses the focusing modifier of right and the directional adjunct of left in close proximity. It is again unclear in “two five right?” whether the question tag or adjunct of direction is intended. Standard phraseology set forth in the Manual of Radiotelephony avoids this ambiguity by using acknowledge, affirm, approved, confirm, correct, request, roger, and WILCO (will comply) to function in the non-standard ways right is used in the corpus. One important question when studying non-standard phraseology is the parameters of discourse within the cockpit. Since this study uses CVR transcripts to gather data, the focal point is the cockpit rather than radio transmissions. Are pilots expected to have no personal conversation with their co-workers on flights that can be upwards of twenty hours? Or are pilots able to suddenly and seamlessly switch back and forth between personal topics and flight management using non-standard and standard phraseology, respectively? Even if this is the case, problems

AVIATION ENGLISH CORPUS LINGUSTICS

8

could easily arise between pilots of different cultures and linguistic backgrounds. Stopping to ask “did you mean the standard or a non-standard use of right?” is not only ridiculous but could cost vital seconds in emergency situations. The Manual of Radiotelephony allows for the use of “additional phraseologies”, “appropriate subsidiary phraseologies” and “plain language” (p. 3-1), but clearly states that plain language should only be used to make messages clearer, not instead of ICAO phraseology and never in ways that creates ambiguity. Conclusion Pilots play a crucial role in the aviation industry; they are sometimes a big liability for airlines, but they are always valuable assets. Rather than expecting pilots to be robots, my amateur opinion is that the human element should be celebrated and nurtured in the cockpit, especially on longer flights. Part of this means allowing pilots to use language in a natural way and build rapport with coworkers. If I continue to investigate these questions, I would like to know more about how far standard phraseology extends. Is it limited to radiotelephony, or should it be used the moment pilots enter the cockpit? Further research could look at how frequently non-standard phraseology occurs in conversations between pilots within a given cockpit compared to radio transmissions between pilots and control towers. For now, it can be noted that only 18.2% of occurrences of right in CVR data are used in standard ways. This clearly goes against the ICAO's efforts to provide “maximum clarity, brevity, and unambiguity” (p. 3-2), creating ambiguity in a field of discourse where clarity of communication is absolutely vital.

AVIATION ENGLISH CORPUS LINGUSTICS

9

References International Civil Aviation Organization. (2004). Manual on the implementation of ICAO language proficiency requirements. (Document 9835 AN/453, 1st ed.) International Civil Aviation Organization. (2007). Manual of radiotelephony. (Document 9432 AN/925, 4th ed.) Netherlands Aviation Safety Board. (1977). Final Report and comments of the Netherlands Aviation Safety Board of the investigation into the accident with the collision of KLM Flight 4805, Boeing 747-206B, PH-BUF and Pan American Flight 1736, Boeing 747121, N736PA at Tenerife Airport, Spain on 27 March 1977. Netherlands. (ICAO Circular 153-AN/56) Mell, J. (1991). What is not standard in real radiotelephony? 4th International Civil Aviation English Association Forum. Paris. Reprinted in ICAO Document 9835. Mendez-Naya, B. (2006). Adjunct, modifier, discourse marker: On the various functions of right in the history of English. Folia Linguistica Historica 27(1-2), 141-195. The Tenerife Airport disaster: The worst in aviation history. (n.d.). Retrieved from http://www.tenerife-information-centre.com/tenerife-airport-disaster.html Tiewtrakul, T. & Fletcher, S. R. (2010). The challenge of regional accents for aviation English language proficiency standards: A study of difficulties in understanding air traffic control-pilot communications. Ergonomics, 53(2), 229-239. U. S. National Transportation Safety Board. (1991). Aircraft Accident Report: Avianca, the airline Columbia, Boeing 707-321B, HK 2016. Washington, D.C. (NTSB/AAR-91/04)