Running head: ENGLISH IN FLIGHT COMMUNICATIONS
The Impact of English Comprehension
The impact of English comprehension on flight communications
My project researched the historical and current status of the English language in Air Traffic Control (ATC) flight communications. While English does not have an official status in international flight, its use has been widely implemented. English-only in ATC communications is not mandated on a global scale. This lack of standardization and regulation has led to much variation of English proficiency in the air transport industry. In some cases, this lack of proficiency has resulted in fatal accidents which might have been avoided if a common language had been available.
In exploring this topic, I first looked at English as a global language in general and in the area of international flight. The first three articles address the historical reasons that English has become the predominant global lingua franca in international communications and in ATC radiotelephony for international flight. The use of a register of aeronautical English with restricted phraseology is detailed and referred to as Airspeak. The proficiency of Airspeak as a subset of English is deemed crucial to clear ATC communications between pilots and controllers; however, the implementation of a global standard of this aeronautical English has not been enforced.
Articles four through seven highlight concerns regarding the lack of an international regulatory agency with power to enforce standards for an official common language and English proficiency for pilots and controllers. In order to substantiate the need for improved global communications, accidents involving language miscommunications are cited. The most frequently mentioned accidents occurred at Tenerife in the Canary Islands, JFK, and the mountainous terrain near Cali, Colombia. In each of these instances, a better understanding of the English language could have possibly prevented the accident.
Finally, articles eight through ten give some insight into current thoughts for improving ATC communications. These suggestions include the adoption of standardized instruments for measuring English proficiency, implementation of computer-based data-link communication systems, and implementation of speech recognition systems.
This project highlighted the need for further research in the area in order to help the aviation industry reach a consensus regarding ATC communication policies. The air transport industry is at a crossroads where technology could play a vital role in the future management of the air traffic system, flight communications, and global navigation. If this technology is not wisely implemented and standardized, the air transport industry may face increased communications obstacles rather than enhanced global understanding.
Crystal, D. (1997).Why English? The cultural legacy. In English as a global language (pp. 78-112). Cambridge, UK: Cambridge University Press.
This chapter provided historical background for many reasons why English has become the predominant global language over the past 50 years. Crystal noted that while English does not have the largest number of speakers, it is the power of those speakers which has caused English to be the unchallenged lingua franca for business, science, and technology. Chapter four focused on areas such as international relations, the media, international travel, education, and communications. Crystal describes registers such as "Seaspeak" for maritime use and limited phraseology for Air Traffic Control called "Airspeak." The formation of the International Civil Aviation Organization (ICAO) at the 1994 Chicago Convention is mentioned as well as the advisory rather than regulatory power of the organization. English is recommended (but not mandated) by ICAO for communication between pilots and controllers who speak different languages. Differences are noted in standards between the Federal Aviation Administration (FAA) and ICAO. The need for bilingual ATC communications is debated as it relates to cultural identity as in Montreal and accident prevention as in the fatal aviation accidents at Tenerife and Cali, Columbia. International communications in general acknowledge English as the language of computers and the Internet.
I chose to include this chapter to provide a broader framework for the basis of my specific study of English in flight communications. Crystal offers a fairly concise description of the current status of English in international travel, especially since the topic was only allowed 5 pages in the chapter. It is probably for this reason that an overview rather than in depth description of the subject is given. Some more details might have been helpful, such as mentioning ICAO's status as a branch of the United Nations. It is apparent in Crystal's writing that he is not a pilot when he refers to "other pilots in their neighbourhood" rather than just saying airspace. I disagree with his statement that "most pilots' level of English is far greater than Airspeak norms." This is unsubstantiated by any facts or statistics, not to mention any definition of who "most pilots" refers to. I have overheard some pilots and controllers on ATC frequencies with a very limited command of English, even when restricted to the limited phraseology of Airspeak.
I question Crystal's fact checking in general since Bill Gates is mentioned as the developer of DOS, when it is widely known in the technical world that Gates merely purchased the operating system and was responsible for its distribution rather than creation. However, I did find the flight data accurate and consistent with other readings.
Morrow, D., & Rodvold, M. (1998). Communications issues in Air Traffic Control. In M. K. Smolensky & E. S. Stein (Eds.), Human factors in Air Traffic Control (pp. 421-456). San Diego, CA: Academic Press.
Morrow and Rodvold provide an exceptionally detailed account of the current status of communication in ATC operations. Research data in the literature involved NASA Air Safety Reporting System (ASRS) forms, direct observation of routine operations of randomly sampled ATC communications, full-mission simulation studies, and modeling communication processes. Factors in ATC communications were categorized as perceptual, linguistic, and collaborative. ATC facility communications were compared and contrasted to controller-to-pilot interactions. Messages could be misinterpreted due to the expectations, level of fatigue, experience, and age of the receiver.
Recommendations for improving ATC communications focused on improving the fidelity of radio systems, refining the measurement of communication performance, and implementing training for ATC controllers much like the Crew Resource Management training provided for flight crews. Data-link communication systems which reduce the amount of ATC controller voice communications by visually transmitting messages via flight computers are another option for improving the clarity of communications . Text messages might initially overload work capacity and timeliness of responses for pilots and controllers, but the system does provide a better record of transmissions.
I was impressed by the broad scope of the literature and research shared in the article. I felt that both the technical and communication aspects of ATC transmissions were intelligently addressed. I especially appreciated the breakdown of communications into perceptual, linguistic, and collaborative categories. As a pilot, I have found perception of messages difficult at times due to poor transmission, pronunciation, rapid speech, and background noise. Linguistic factors have been a concern when controllers deviate from standard phraseology making message interpretation confusing. For instance, "240" without the proper context could refer to an altitude, airspeed, or heading. Collaboration comes into play when message transmission is not confirmed by readback. In the air I do not hesitate to ask a controller to "say again" or "confirm altitude assigned" but other pilots, including my husband, are sometimes hesitant to bother the controller and therefore rely on their own memory and interpretations.
The discussion of data-link computers was interesting, but I am not yet convinced these systems will serve as a solution for all ATC communication problems. If controllers and pilots only communicate by voice in the event of an emergency, voice communications may become even more unreliable. Pilots and controllers could learn to rely on text messages and ignore improving English proficiency. Communications training may be a better alternative.
Varantola, K. (1989). Natural language vs. purpose-built languages: The human factor. Neuphilologische Mitteilungen, 90(2), 173-183.
Varantola offers a comparison of the use of purpose-built languages in maritime and aviation navigation. The historical use of English and special varieties of English are discussed. International Civil Aviation Organization (ICAO) recommendations are detailed. Two options are suggested for improvement in ATC communications. One is the implementation of computer-based data-link devices. The second option prefers to refine the code of voice communications rather than switch to another mode of communications altogether. Problems with the current code are addressed such as restricted phraseology which although aiming to be unambiguous can lead to confusing communication, especially if read backs are a question rather than a confirmation or a phrase can mean more than one thing depending on the context. The possible conflict between brevity and clarity is posed as a concern. While text messages are seen as more clear and concise, Varantola argues that context is still necessary for correct interpretation.
The 1977 Tenerife aviation accident is mentioned with a brief transcript of the misleading communications just prior to the collision of a KLM 747 on take off into a Pan Am 747 which had not yet cleared the active runway. A maritime accident involving two ships in Swedish waters is also detailed. VHF communications of the incident were in contrast to what one of the ship's crew claimed to have heard and said when relaying the incident in court. In both accidents, seemingly minor words might have played a large role in miscommunications.
Although written over ten years ago, I decided to include this article because of its contrast of Airspeak to Seaspeak as well as the thorough discussion of natural vs. restricted code languages. I disagree with Varantola's position that non-native English speakers might communicate better with restricted phraseology than native speakers because they would be less likely to use extraneous wording. I believe non-native English speakers may use less words outside of the code of Airspeak or Seaspeak due to a lack of proficiency which could diminish clear communications in the long run. I agree with Varantola that context is necessary for clear message understanding.
I do not have another account of the Swedish maritime accident, but other readings regarding the Tenerife accident were more detailed. The transcript noted the KLM pilot as stating "We are now ready on (or at) takeoff." Other literature supports the "ready at takeoff" quote and also states that in Dutch syntax "at takeoff" would be the same thing as "taking off." Most native speakers of English would not know this distinction. It is understandable that the native Spanish speaking controller was equally unaware of the meaning of the Dutch KLM pilot.
Feldman, J. M. (1998). Speaking with one voice. Air Transport World, 35(11), 42-51.
Feldman laments the inability of the International Civil Aviation Organization (ICAO) to enforce mandatory English communications. The International Air Transport Association (IATA) and International Federation of Airline Pilots Association (IFALPA) are pushing for a stripped down aeronautical English to become the standard for communications. IFALPA and IATA have had English language training programs for pilots and controllers. English is mandated on a country by country basis. Russia does not mandate English only, but China, Germany, and France do. Africa was noted as a big concern as the continent is multilingual and much of the vast area is in uncontrolled airspace where pilots are responsible for communicating to each other regarding position and intentions in order to avoid mid-air collisions. In countries where English-only is mandated, the policy is often not enforced allowing pilots and controllers to converse in another more familiar language. This multilingual communication eliminates the ability of other pilots in the airspace to listen in and gain situational awareness. Language related accidents are briefly mentioned including Tenerife, Urumqi, Avianca at JFK, Dehli and Korean Air, and Cali, Columbia. Computer data-link communications are seen as a way to unburden the overloaded infrastructure of ATC voice communications; however, voice communications are the proposed back up for the data-link system and aeronautical English language proficiency will be necessary for this communication.
While a call for English-only is loud and clear in the industry, I do not believe that ICAO as a branch of the United Nations will ever have the power to mandate policy. ICAO is an advisory organization and not a regulatory agency. It will take international aviation unions such as IFALPA and IATA to encourage airlines to come to a consensus regarding standardization of aeronautical English, also referred to as Airspeak. Even if a mandate does come to pass, I am not sure a universal policy can be enforced.
Feldman boasted the Chinese government's mandate of English-only in flight, and yet the Urumqi crash of a China Northern MD-82 was unofficially reported as due to misunderstanding of an English warning emanating from the Ground Position Warning System (GPWS). I disagree with Feldman's argument that the Cali, Columbia crash was not due to poor English communication. He felt the cause was solely due to the flight crew's Flight Management Systems FMS programming errors. While this may have been the root cause, if the Colombian controller had been better equipped to communicate the urgency of the impending Controlled Flight Into Terrain (CFIT) to the flight crew in English, the crew might have regained situational awareness and the accident may have been avoided.
Nordwall, B. D. (1997, September 29). FAA: English ATC standards needed. Aviation Week & Space Technology, 147(13), 46-51.
The article details the position of the National Transportation Safety Board (NTSB) and the U.S. General Accounting Office (GAO) which encourages Congress to mandate English proficiency for pilots and ATC personnel. Nordwall states, "there are no ICAO [International Civil Aviation Organization] standards of language proficiency for pilots or controllers and no FAA [Federal Aviation Administration] benchmarks to measure the adequacy of English to fly in U.S. airspace." The Cali, Colombia crash is detailed as an example of a controller's inability to communicate effectively in English. W. Frank Price, manager of air traffic services international staff for the FAA said, "Had he [the controller] been able to do so [communicate the crew's position in English], it could have contributed to the crew's situational awareness - a factor that might have prevented the accident." Avianca at JFK, another language related accident is related in which pilots were unable to use the proper terminology for a fuel emergency. The copilot told ATC that he was running out of fuel, but never used the words "minimum fuel" or "emergency." Airline sponsored English language training at Embry-Riddle Aeronautical University and University of North Dakota as well as use of a standardized Test of English for International Communication (TOEIC) for foreign controllers was implemented on a small scale. Agreement between the FAA and ICAO regarding English standardization is needed.
Nordwall seems to think that if the U.S. under the auspices of the FAA takes the lead on English standardization for ATC communications that the world will follow. He established early in the article that ICAO has no regulatory power, yet calls for the FAA and ICAO to implement an official language policy. While I feel that an English-only policy for ATC communications might alleviate many language related accidents, I am not sure how this policy could be mandated. Airline sponsored English language programs are a step in the right direction; however, the program at Embry-Riddle Language Institute consists of brief English as a Second Language (ESL) modules. It is unclear to me that a condensed intensive training program can truly prepare pilots and controllers to effectively communicate in English, especially in emergency situations. I do agree with the article's interpretation of the Cali, Colombia crash, as I stated in my analysis of the previous article. Regarding Avianca at JFK, it is difficult for me to imagine that if a Colombian pilot said he was "running out of fuel" in a foreign accent, that a controller would not probe the statement further to determine the severity of the situation; however, the airspace in the New York City area is extremely congested with LaGuardia and Newark in nearby proximity sharing ATC center controllers.
Verhaegen, B. (2001). Safety issue related to language use have come under scrutiny. ICAO Journal, 56(2), 15-17, 30.
The article starts by relating the Cali, Colombia crash and then detailing the 1996 crash near Dehli, India. The Dehli crash, involving a Boeing 747 and an Ilyushin Il-76, was due to miscommunication amongst two foreign flight crews. Neither crew was reported to have a high level of proficiency in English. The accident inquiry revealed that there was confusion about the level to which the Il-76 was cleared to descend. International Civil Aviation Organization (ICAO) secretariat and author of the article, Verhaegen continues by stating Annex 1, the Procedures for Air Navigation Services - Rules for Air Traffic Services (PANS-RAC), and Annex 10. The policies are respectively regarding ability to communicate in the local language(s), knowledge of radiotelephony procedures and phraseology, and the provision of English for universal use in aeronautical radiotelephony communications. ICAO's Trainair language training program is mentioned as is the 1980 ICAO publication, Aviation English for Air Traffic Controllers. Collaboration with the Federal Aviation Administration (FAA) and the Multi-Agency Air Traffic Services Procedures Coordination Group (MAPCOG) is anticipated by ICAO to improve the use of standard phraseology internationally. Another task force spearheaded by the ICAO Secretariat is the Proficiency Requirements in Common English (PRICE) Study Group. Standards are needed in minimum English proficiency and aeronautical phraseology.
This article appears in the ICAO Journal, so it might overstate the span of control of the organization. Verhaegen does mention the vagueness of the ICAO policies and does note the need for cooperation amongst other countries' regulatory agencies such as the FAA and MAPCOG. At times the article seems to read more like a press release or a "State of the Union" address than an objective look at the status of current ATC communications. Verhaegen does make a plea for the "need to establish requirements enhancing the minimum performance standards for radiotelephony phraseology and use of the English language," but unfortunately his closing sentence merely states, "ICAO is actively studying this important safety-related matter." While study is necessary to ensure that the best measures are sought for clearer ATC communications, the best that ICAO can do following the outcome of these studies is hope that other entities with regulatory power will agree with ICAO's recommendations and choose to implement them. I do not have an easy solution in mind for remedying ICAO's lack of jurisdiction, but without a global regulatory agency it may be hard for the international aviation industry to come to a consensus on policy, let alone all agree on English as the official common language for ATC Communications.
Mathews, E. (2001). Provisions for proficiency in common aviation language to be strengthened. ICAO Journal, 56(3), 24-26, 41.
Mathews strongly states, "Within aviation, English has become firmly entrenched as the defacto common language of international communication. It is also the only practical choice at this time for designation as the official first language of international radiotelephony communications." The article goes on to list Annex 1, Annex 10, and current ICAO initiatives. Mathews explains her impressions of linguistics and language and how each affects communication. She stresses the importance of context and culture in the understanding of language. In this vein, she defends natural languages over artificial languages due to the complexity of language. She feels, "Poor communication, or lack of a mutual proficiency in a shared common language can be a causal factor in an accident." Brief accounts of the Tenerife, Avianca at JFK, and Cali, Colombia accidents are described. Flight deck and ATC translators are suggested as an impractical alternative to language proficiency. Computer data-link systems are posed as another option, but Mathews fails to view the implementation of such infrastructure as a replacement for standardization of English proficiency.
I agree with Mathews assertion that English is the "defacto common language" and perhaps the only "practical choice" for an official language in aviation; however, it is still a bold comment to print in an international journal such as ICAO Journal. The non-regulatory nature of ICAO puts the organization between a rock and a hard place when it comes to making recommendations vs. realizing political sensitivities. ICAO Journal notes Mathews' affiliation with Embry-Riddle Aeronautical University at Daytona, Florida; however, she is not listed amongst the Embry-Riddle Language Institute staff, evidently since she is on leave from the institution. It is not clear to me what credentials Mathews has regarding the study of languages or linguistics, but she does seem to state her case regarding the nature of each succinctly in all of the detail that a brief article on the topic can lend. Mathews does advocate English proficiency testing, even if the implementation of computer data-link communications comes to fruition. I agree with this position, because the back up for the text-based ATC messaging system is voice communications. Text based systems are most effective if based on a common language such as aeronautical English or Airspeak. To rely on translation systems in order to accommodate multiple languages might only add to confusion and miscommunication. Syntax errors, such as in the Tenerife accident where "at takeoff" in Dutch is commonly an active phrase synonymous with "taking off," between languages would only multiply the possibilities of more miscommunications.
Noble, C. E. (1997). The aviation English problem in America: Can a real-time based flight simulator help? Paper presented at the Excellence in Aviation Award Centers of Excellence Program Office, Atlantic City: NJ.
The article highlights the growing problem of a lack of English proficiency amongst Japanese flight students who seek training and certification in the U.S. Noble posits that a majority of flight training can occur in uncontrolled airspace; therefore, Japanese student pilots may train and receive private pilot certification without experiencing rapid English speech communications from ATC controllers. Without this exposure, pilots could reach "cognitive overload" and miscommunicate in a heavy traffic environments. There are currently no standardized test for English proficiency in the Federal Aviation Administration (FAA) certification of pilots. In order to determine English proficiency, an FAA examiner is only responsible for having a student pilot read and show understanding of the English language. It is recommended for the examiner to have the student read printed text such as in a magazine or newspaper and display comprehension by paraphrasing the article. This exercise is not a good indicator of a student pilot's ability to understand spoken communications in an ATC environment. Noble proposes a computerized flight simulator capable of automated measurement of language skills to address this concern.
I believe that a better instrument for determining English proficiency is needed for FAA certification of pilots. I am not convinced by this article that the simulator described provides this solution. The flight simulator interprets English comprehension based on flight control inputs made by the student pilot in response to a data bank of standard ATC communications. The proposed simulator is computer-based and utilizes sensory inputs like a computer mouse which requires much different responses than controls found in a more realistic cockpit flight simulator. Even realistic simulators which included a control yoke, rudder pedals, and flight instruments much like those found in an airplane cockpit have some fidelity issues where response is measured. The flight simulator I trained in at Lafayette Aviation is much more sensitive to control inputs than a traditional airplane. Simulator studies in general have shown that the level of fidelity is directly proportionate to the level of transferability in real life situations. Perhaps scripted mock ATC communications between instructors or flight examiners and student pilots would be more realistic in determining English proficiency when training in uncontrolled airspace. In lieu of a better FAA solution, I favor the implementation standardized English proficiency tests such as the Test of English for International Communication (TOEIC) in combination with prescribed communicative evaluations such as the Speaking Test of English for Pilots (STEP) utilized for evaluation at the Embry-Riddle Language Institute.
Clinch, P. (2000). Existing systems provide essential communications while development of data link carries on. ICAO Journal, 55(7), 16-17, 29-30.
The article describes the Future Air Navigation System (FANS), Communications, Navigation, and Surveillance (CNS) and Aircraft Communications Addressing and Reporting System (ACARS). The purpose of the FANS concept is to provide pilot-controller data-linked communications. Rather than wait for the International Civil Aviation Organization (ICAO) to decide on a data link standard, much of the airline industry has already implemented ACARS to provide aeronautical operational control (AOC) to aircraft. AOC communications involve transmission of routine airplane dispatch and weather information. Air Traffic Services (ATS) are provided by satellite via Inmarsat and SITA, a global telecommunications company. While ACARS is widely in use in air transport, the text only formatted system is not being considered for ICAO standardization. ICAO's Aeronautical Telecommunications Network (ATN) will support Automatic Dependent Surveillance (ADS), Controller Pilot Data Link Communications (CPDLC), and Dynamic Airborne Rerouting Procedures (DARPs). Unlike the text-based constraints of ACARS, ATN will support transmissions in any format. In response to this inflexibility, VHF Digital Link (VDL) avionics have been designed to expand transmission abilities of ACARS to comply with ICAO ATN standards. Future communications may be less restricted by closed systems and rely on Internet Protocol (IP).
While ICAO is great at research, implementation in the aviation industry continues to be a problem. Once again, the aviation industry is forging ahead and spending billions of dollars on equipment that may very soon be obsolete. I do not blame the industry for seeking reliable solutions to communication problems when they are available, but it seems that companies like SITA are guiding the industry more than ICAO. I can't fault the airlines, however, because they need to run a business and do it now, not ten years from now. It is unfortunate that the industry standard of ACARS is currently limited to text-based transmissions and needs a band-aid solution of adding yet another proprietary system such as VDL in order to get with the ATN program. At this stage in the game and with the current financial woes of the airline industry in general, I would think everyone would be ready to abandon ACARS and adopt an open source IP system. While money has already been spent on ACARS, how much more will have to be spent to keep the antiquated system up to speed? Then again, the aviation industry has been operating with archaic radiotelephony systems for years and doesn't seem to be in a big hurry to upgrade. I just hope that ACARS vs. IP doesn't become another Beta vs. VHS battle where an inferior system wins out due to wide spread industry adoption.
Churcher, G. E., Atwell, E. S. & Souter, C. (1997). The semantic/pragmatic annotation of an Air Traffic Control corpus for use in speech recognition. In Ljung & Magnus (Eds.), Corpus-based studies in English: Papers from the seventeenth international conference on English language research on computerized corpora (ICAME 17) Stockholm, May 15-19, 1996 (pp. 353-373). Amsterdam: Rodopi.
This study was conducted utilizing speech recognition systems in order to analyze Air Traffic Control (ATC) communications. The purpose of the system was for transcription of key data. Key information areas of messages were categorized as change of altitude, change of pressure settings, change of transponder codes which track individual aircraft, change of pilot's radio frequency, and change of aircraft heading. The model used for language was based on the Civil Aviation Authority's (CAA, the U.K. equivalent of the U. S. FAA) Radiotelephony Manual. In addition, 90 minutes of actual ATC transmissions from Leeds Bradford Airport on the approach frequency were transcribed and analyzed. The speech recognizer used for the project was SSI's Phonetic Engine 500. A limitation of the software was the model of American speakers rather than British speakers. Different parameters for lexicon, grammar, and syntax were devised and tested. The experiment utilized six speakers (three male and three female) who recorded 19 sentences with a noise cancelling microphone in both a quiet environment and normal office setting.
While the article concluded that speech recognition had a long way to go before it could be applicable in an ATC environment, I found other factors in the study incomplete. The study only monitored approach frequency communications from one facility for a period of 90 minutes. There is activity in ATC communications on ground, departure, and center frequencies that was not included in the study. The use of the CAA's phraseology makes the study only generalizable to the U.K. and excludes many other English speaking countries such as the U.S. Use of CAA Airspeak as a model for the lexicon made free speech outside of that register more difficult to analyze. It was noted in the article that transcriptionists not familiar with Airspeak interpreted phraseology incorrectly. The study does not detail how the six speakers were selected or if any of the speakers had an accent. The recordings were conducted in a relatively sterile environment in contrast to ATC facilities or the cockpit of an airplane. Background noise is much more prevalent in ATC communications even in comparison to the normal office setting used to simulate a higher noise level environment. Limitations in the voice modeling of the software (American standard vs. British) used in the study may have confounded the results of the experiments. The study did provide limited evidence that further refinement of speech recognizers is necessary before implementation in an ATC environment.
owned by: kitty
URL: http://lairds.org/Kitty/Aviation_Communications/papers/annotated_bibliography_aviation_communications