The prevalence of startle during unexpected critical events has been shown to adversely affect flight safety in a number of high profile accidents over recent years. Additionally, flight simulator startle experiments conducted by the authors, showed approximately one-third of pilots were significantly affected by startle during an instrument approach, leading to undesired aircraft states in a critical phase of flight. Training interventions for managing startle have been largely uncoordinated to date, however a holistic program of prevention and recovery training is proposed. Prevention strategies include improved training and attention in situational awareness skill sets, and particularly pilot monitoring skills, developing greater expectation and efficacy for managing unexpected critical events, and greater awareness of startle effects. Recovery strategies include more focus on evidence based training, improved training on avoidance, recognition and management of undesired aircraft states, and exposure to unexpected critical events during training. Adopting holistic training interventions for managing startle will have other benefits including improved threat and error management, and improved prevention of, and recovery from, undesired aircraft states.
Introduction A number of recent aircraft accidents have directly implicated or may have involved startle as a contributory causal factor. Colgan Air Flight 3407, Air France Flight 447, Turkish Airlines Flight 1951 and West Caribbean Airlines Flight 708 are examples of accidents over recent years where pilots were suddenly surprised by stall warnings, and subsequently took either ineffective, or inappropriate action, which failed to remedy or even exacerbated, the critical undesired aircraft state they had encountered (BEA, 2006, 2012; Dutch Safety Board, 2010; NTSB, 2010).
When considering why highly trained aviation professionals, who have had stall recovery training reinforced throughout their careers, would suddenly make exactly the opposite control response to the desired response, or at least an ineffective one, then indications point directly to some rudimentary breakdown in normal information processing.
Modern jet aircraft are generally fitted with similar stall warning systems. These mostly consist of both an auditory and tactile warning of impending stall, which is often accompanied by visual indications on the Flight Instruments. In each of the accidents mentioned above, and in others, it appears that the pilots have been genuinely surprised by the stall warning stimuli, which clearly have strong connotations of impending threat. Research shows that when people are suddenly startled under conditions of persistent threat then their level of startle is significantly worse. This ‘fear potentiated’ startle has been shown to rapidly activate the sympathetic nervous system, with a significant effect on the body’s systems. This process often includes the ‘fight or flight’ response, which introduces adrenaline to the bloodstream and increases heart rate, and is further escalated into a full stress response, where up to thirty hormones are introduced (Stratakis & Chrousos, 1995). This full stress response, which is likely where real threat exists, or is perceived to exist, can cause significant deterioration in psychomotor, working memory and other cognitive processes (Thackray, 1988), all at a time when full cognitive ability would be most welcome in diagnosing, problem solving or implementing effective recovery strategies.
While being exposed to startling critical events may mean disaster for some pilots, there are such events every day which are adequately recovered from by pilots throughout the world. Examining the root cause of fear potentiated startle may help us understand why different pilots perform in different ways when exposed to startling stimuli, which may in turn assist in developing strategies for improving overall pilot performance during surprise critical events.
Having discussed the cognitive causes and effects of startle, this paper will discuss some training strategies which may help pilots to mitigate the negative effects of startle and therefore improve performance during unexpected critical events.
Fear-potentiated Startle All humans and most animals startle. Research has shown that this startle is worse when it is unanticipated, when people are tired, or when they have elevated levels of arousal. Laboratory research has shown that where people often recover very quickly from ‘false alarm’ startles, when they are exposed to a startling stimulus which is perceived as threatening, then their level of startle is significantly worse and is likely to include the arousal of the sympathetic arm of the autonomic nervous system in a full stress response. This enhanced reaction is known as ‘fear-potentiated’ startle (Davis, 1992, 2001; Grillon, Ameli, Foot & Davis,1993; Lang, Bradley & Cuthbert, 1990).
While startle itself is technically just a reflex physical reaction (which intuitively moves the recipient away from the stimulus while at the same time aligning their attentional resources towards its source), it is generally the full stress response which accompanies startle under threat, which produces a potentially pathological ‘startle reaction’.
At the heart of this reaction is the Amygdala (Davis, 1997; Le Doux, 2000). This pair of almond shaped structures in the limbic region of the brain are responsible for storing either directly, or through associations with other memory structures, a catalogue of all those events in our lives with some emotional valence. These ‘emotional memories’ may be pleasant ones, or they may be associations we have stored of distasteful or distressing encounters. Such encounters do not necessarily need to have been experienced directly. ‘Virtual experience’ where we perhaps read about someone else’s episode, or see pictures of it, can be enough to generate negative connotations which are then stored away as ‘arousing’ emotional memories.
Certainly where people have directly experienced distressing or fearful events in real life, then they often retain very strong emotional memory for such events, and the mere hint of a recurrence of a similar event, can generate a significant stress response in them.
It is not hard to see how a pilot could make direct connotations between a stall warning and conditions of real threat. Exposure to stall recovery training in pilot careers may not be a pleasant experience for some. While for most pilots the challenge of a successful stall recovery may generate enjoyable memories, for some pilots, who perhaps lack quality training, or simply perform below optimum, exposure to stalling can be a negative experience. Coupled with media exposure of aircraft accidents and other autobiographical memories, the whole association of aircraft stalling may induce strongly negative emotional memory. When suddenly exposed to an unexpected stall warning then, such pilots may experience feelings of real threat associated with the stall situation, which in turn exacerbate their reaction to that of a ‘fear potentiated’ startle.
Where this enhanced startle occurs, and the sympathetic nervous system is fully aroused, significant deterioration is common in cognitive processes. Specifically, the attentional system tends to become very narrowly focused, and not necessarily on the most important information. Very salient, but irrelevant information may be attended to, or even fixated upon, at the expense of other, more critical information. Processing irrelevant information in a bottom-up manner at a time when critical information is ignored, could severely inhibit analysis, decision making and recovery.
Additionally, the working memory is severely impaired by acute stress (Diamond, Fleshner, Ingersoll, & Rose, 1996; Eysenck, Derakshan, Santos, & Calvo, 2007; Matthews, Davies, Westerman, & Stammers, 2008). Where stored mental schemas for recovery processes would ideally be activated into working memory from implicit long term memory, the working memory rather becomes preoccupied with task-irrelevant, anxious thoughts. With very limited capacity anyway, this preoccupation with things which won’t help to analyse or fix a situation, severely restricts the ability to recover from it. Where the situation is an undesired aircraft state then this can delay or impair recovery enough to have severe consequences.
Situational awareness, which relies heavily on the working memory to retain a ‘mental picture of the situation’, is likely to suffer major disruption with any degradation in either attentional processes or working memory function (Eysenck, Derakshan, Santos, & Calvo, 2007). This in turn has potentially serious consequences for decision making, team work, leadership, and other important processes during non-normal events.
Simulator experiments conducted by the authors, which exposed pilots to a startling stimulus at a critical stage of flight, showed that pilot reactions varied significantly. Approximately one third performed nominally, one third showed minor impairment, and one third showed substantial impairment. This last group showed significant effects of both psychomotor and cognitive disruption, with breakdowns in situational awareness and decision making, leading to highly unstable approaches in some cases (Martin, Murray & Bates, 2012).
The Appraisal Process At the heart of the fear-potentiated startle and the associated stress response, is the appraisal process which occurs in the Amygdala (Davis, 1997; Le Doux, 2000). Sensory information from four of the five senses (not smell – it is processed directly elsewhere), arrives at the sensory area of the thalamus and is then projected to the Basolateral nuclei in the Amygdala (Le Doux, 2000). The incoming information then undergoes a very rapid pattern matching with stored emotional memory. From this process an assessment is made as to whether the information is benign, irrelevant, challenging or involves some threat or potential loss (Lazarus & Folkman, 1984).
Where the information was both surprising and considered to be associated with harm, threat loss, or perhaps even challenge, then both the startle circuits (through regions at the top of the brainstem), and the fight or flight/stress response circuits (ie. the sympathetic nervous system) are activated. This process is incredibly quick, with some human experiments showing physical results of startle in as little as 14 milliseconds (Davis, 1984).
While this process is occurring, a signal is also sent to the pre-frontal cortex for cortical processing. This process, which is commonly known as perception, examines the new information in more depth, and with the assistance perhaps of more contextual information, to determine more fully, the emotional relevance and potential threat of this information. This cortical processing can however take some time, with research showing that at least 500 msec is required, and possibly more (Asli & Flaten, 2012).
This disconnect presents opportunities for false alarms. On the one hand we may make a coarse appraisal that something is threatening and suffer a ‘fright’, with the ensuing fight or flight/stress response, only to find that when we have had a chance to analyse it properly, this information wasn’t threatening at all. This would then activate ‘extinction’ circuits which in turn activate the parasympathetic nervous system, which lowers arousal and returns the body to a state of homeostasis (Myers & Davis, 2002).
Where the cortical processing suggests that the threat is in fact genuine however, then a reinforcing signal is sent back to the amygdala, which further enhances the stress response and the level of startle.
Prevention Strategies For Avoiding Startling Situations Often, startling events such as stall warnings, come after a period of reducing airspeed which would have been obvious if pilots were monitoring the airspeed. Accidents such as Turkish Airlines Flight 1951 (Dutch Safety Board, 2010) and Colgan Air Flight 3407 (NTSB, 2010) are examples where decreasing airspeed was either predictable or obvious over a sustained period, and it was only poor monitoring which prevented this being noticed.
Pilot monitoring involves the comparison of environmental cues to a master mental schema which is continuously updated for local variations on the day. A framework of SOP’s form expectations which are reinforced through repetition. On any given day this continuously updated ‘mental model’ of what should happen is compared by both Pilots to actual conditions, and disparities are either noticed and addressed, noticed and ignored, or not noticed.
While the Pilot Flying devotes considerable mental resource to managing the aircraft, it is the Pilot Monitoring who is principally responsible for noting deviations from expectations and then alerting the Pilot Flying to these deviations. The design of late generation aircraft incorporates a lot of the monitoring functions which were traditionally done by pilots, and unfortunately this skill of ‘monitoring’ is being continually eroded as a result.
Situational awareness involves a complex and extensive set of individual social and cognitive skills. These skills include: communicating effectively; planning; learning and knowledge retrieval; temporal awareness; vigilance, workload assignment and management; reviewing and modifying and inquiry (Murray & Martin, 2012). Further attention to situational awareness skill sets, and to specific monitoring skills, would increase the likelihood that trends such as decreasing airspeed would be picked up and rectified prior to a startling event.
An additional mediator of startle is knowledge. By arming Pilots with a greater understanding of the possible detrimental effects of startle, there is likely to be a greater motivation for having plans to cope with startling events and also an expectation for impaired performance during startle. This expectation should result in less likelihood for impulsive behaviours following startle, such as those in the Colgan accident, and others.
Prevention By Mitigating the Effects of Startle The crucial element in the negative effects of ‘fear potentiated’ startle is the appraisal of threat, which is based primarily on emotional memory.
Where two pilots may be presented with exactly the same unexpected critical event, one may appraise the situation as life threatening, while the other may simply view it as a challenging, but manageable event. If we compare this to the analogy of two ‘bungee jumpers’, it would not be uncommon for a ‘first timer’ to be scared to death, while an experienced jumper may actually relish the challenge as ‘exciting’ and an emotional buzz. The distinction is clearly that one has serious fear connotations associated with jumping from a great height, while the other has previously mastered those emotions and simply appraises the event as thrilling.
In training pilots to avoid, or to at least reduce the effects of fear potentiated startle therefore, the key must be in changing their appraisal of what constitutes threat.
Where a pilot has negative emotional memory for stall situations, then their performance is likely to be worse than someone with positive emotional memory. To improve the performance of the fearful pilot then, some interventions which allowed them to develop a sense of mastery for stall recoveries, would be most likely to provide improvement. This can be achieved holistically in a number of ways. Firstly, in a training environment the language used and the constructive manner in which training is conducted, can assist greatly. By creating a perception of a ‘challenging but fun’ exercise with repetitions to competence, which are verbally reinforced and praised, stall warning events downstream can possibly be appraised more positively. This may already be the case for most of the pilot population, but certainly isn’t for some.
A further method for gaining a sense of efficacy in unexpected critical events, is through personal reflection. This ‘virtual experience’ allows pilots the opportunity to develop a set of mental schemas for managing certain events (Shea & Wulf, 2005). Having a stored plan of action for generic critical events means that should such a situation occur, or a similar situation for that matter, then it is far simpler to activate this schema into working memory than it is to try and generate a plan from scratch, particularly under severe working memory impairment.
This process is common with professional, intrinsically motivated pilots, however it must be suggested that not all pilots have a level of motivation which would engender this process. Conditioned expectation of normality borne out of years of event-free flying can often dull the perceived need for continual review of critical event management outside widely spaced simulator exposure.
This development and maintenance of critical event schemas is also enhanced by group reflection. The ubiquitous question ‘what would you do if….?’ (Martin, Murray & Bates, 2011) as a means of developing critical event strategies, is a commonly used technique in military operations; however it seems to be rarely used outside designated training operations in the airline environment. Developing a culture in airline operations, where this process is normalized, would likely generate significant benefits in the development and maintenance of schemas for managing unexpected events.
Strategies For Enhancing Recovery From Unexpected Critical Events Exposure to a range of critical events in a constructive learning environment may also pay dividends in terms of threat appraisal, which then assists in effective recovery. Encouraging pilots to have their own rudimentary ‘rules of thumb’ for managing critical events followed by sufficient practice to feel comfortable in recovering from such events, is the most likely way to engender improvements in UAS recovery. A mental checklist for this may be as simple as the ubiquitous ‘Aviate, Navigate, Communicate’ or a little more detailed:
State that you have a problem
Fly the aircraft. This would be type dependent but could be something like:
- Ensure the aircraft wings are level and the attitude is close to level flight, or is appropriate, where terrain escape is a factor.
- Ensure thrust is sufficient to maintain level flight or to climb if necessary.
3. Assess the problem.
4. Make a plan and execute it.
Enunciate your actions and intentions if possible.
While highly desirable, enunciating your actions and intentions may not be possible. Given the impairment likely in working memory function and the high resource requirements needed during communication, the likelihood of having spare capacity for forming communication is slim. There are two benefits of communicating intentions if it is possible however. Firstly, it tells the support pilot what inputs you are making, and secondly, it allows, or enhances, the continued shared mental model of what is happening, between both pilots. Had the Flying Pilots in the Colgan Air Flt 3407 Accident or the Air France Flt 447 Accident enunciated their actions, then it is conceivable that the support pilot may have had sufficient wherewithal to recognise that the response was inappropriate. Regardless, two impaired working memories may be better than one in determining appropriate responses.
Finally, continued exposure to recovering from upsets will create both a sense of efficacy, and a modicum of autonomic skill response (Ericsson, 2006). Like the bungee jumper analogy used earlier, repeated exposure engenders a sense of mastery which changes the appraisal of such events in a far more positive way. Some regulators require regular upset recovery training, however the value of airlines incorporating this routine exposure are easily quantified from a psychological viewpoint. Practice in coarse manipulative recovery techniques (which are outside the normal flight skill-set), coupled with an enhanced sense of efficacy, are likely to show significant benefits, particularly in unusual attitude events.
Conclusion The effects of strong startle have been shown to have adverse effects on human performance. This startle has been shown to be particularly exacerbated during events where there is a strong connotation with threat. The effects of this ‘fear potentiated’ startle have been shown by experiment to create significant changes in the human nervous system, which in turn may impair cognitive processes in the brain. This has major implications for the handling of unexpected events, especially where those events have an element of criticality.
At the heart of this impairment is an appraisal of fearfulness, borne out of perceived threat. Suddenly being confronted by a situation which is appraised as life threatening, produces significant arousal of the sympathetic nervous system, with associated reductions in working memory function, at a time when reduced cognitive capacity is highly undesirable in dealing with abnormal events.
Changing pilots’ appraisal of what is actually life threatening during unexpected events, is not a simple fix. Holistic training programs, which target both prevention and recovery, and which enhance pilots’ sense of self-efficacy for dealing with critical events, are likely to be the best method for universal improvement.
Prevention strategies include better situational awareness and monitoring skills, better understanding of the effects of startle, self-reflection, and group reflection for managing unexpected critical events. Thinking through ‘what would I do if….?’ type scenarios, or openly discussing them in briefings, or at quiet times in flight, are likely to generate effective mental schemas for managing such events. Having such schemas stored in memory enhance the chances of these processes being accessible in working memory during time critical, or highly arousing events.
Recovery from critical events is likely to be enhanced by both upset recovery practice, and through the use of ‘go to’ strategies. A generic fundamental plan for dealing with any unexpected event, particularly if it is refreshed in memory frequently, may be an effective means of initial recovery from almost all critical events, but will always be subject to working memory limitations..
Greater aircraft sophistication, with a continued increase in reliability, will make it easier for pilots to develop a conditioned expectation for normality, which in turn is likely to enhance the level of surprise during unexpected critical events. Regular exposure to critical events in training, coupled with the development of a culture where self and group reflection for such events during normal operations occurs, will reduce surprise, enhance self-efficacy, and improve pilot performance.
Aligning airline training with evidence based, event management strategies, is likely to show dividends going forward. Less emphasis on legacy regulatory requirements and more emphasis on training and assessing contemporary skill-sets, coupled with appropriate knowledge levels, will offer more likelihood of successful management of critical situations in the future.
References Asli, O. & Flaten, M.A. (2012). In the blink of an eye: Investigating the role of awareness in fear responding by measuring the latency of startle potentiation. Brain Science, 2, 61-84
BEA (2006). Accident in Venezuela on 16 August 2005. Accident report summary in English. Retrieved from http://www.bea.aero/en/enquetes/venezuela/summary.en.php
BEA (2012). Final Report on the accident on 1st June 2009 to the Airbus A330-203 registered F-GZCP operated by Air France flight AF 447 Rio de Janeiro - ParisRetrieved from: http://www.bea.aero/docspa/2009/f-cp090601.en/pdf/f-cp090601.en.pdf
Davis, M. (1984). The mammalian startle response. In Neural Mechanisms of Startle Behavior (R.C. Eaton, ed.) pp. 287-342. Plenum Press, New York
Davis, M. (1992). The role of the amygdala in conditioned fear. InJ. P. Aggleton (Ed.), The amygdala: Neurobiological aspectsof emotion, memory, and mental dysfunction (pp. 255- 305). New York: Wiley-Liss.
Davis, M. (1997). Neurobiology of fear responses: the role of the amygdala. Journal of. Neuropsychological and Clinical Neuroscience9, 382–402.
Davis, M. (2001). Fear-potentiated startle in rats. Current Protocols in Neuroscience, 8, (8), 11-15
Diamond, D.M., Fleshner, M., Ingersoll, N. & Rose, G.M. (1996). Psychological stress impairs spatial working memory: Relevance to electrophysiological studies of hippocampal function. Behavioral Neuroscience, 110, 661–72
Dutch Safety Board (2010). Crashed during approach, Boeing 737-800, near Amsterdam Schipol Airport, 25 February, 2009 (Aircraft Accident Report). Retrieved from: http://www.onderzoeksraad.nl/docs/rapporten/Rapport_TA_ENG_web.pdf
Ericsson, K.A. (2006). The influence of experience and deliberate practice on the development of superior expert performance. In K.A. Ericsson, N. Charness, P.J. Feltovich, & R.R. Hoffman (Eds.), The Cambridge handbook of expertise and expert performance (pp. 683–703). Cambridge, England: Cambridge University Press
Eysenck, M. W., Derakshan, N., Santos, R., & Calvo, M. G. (2007). Anxiety and cognitive performance: Attentional control theory. Emotion, 7, 336–353.
Grillon, C., Ameli, R., Foot, M. & Davis, M. (1993). Fear-potentiated startle: relationship to the level of state/trait anxiety in healthy subjects. Biological Psychiatry, 33, 566
Lang, P.J., Bradley, M.M., & Cuthbert, B.N. (1990) Emotion, attention, and the startle reflex. Psychological Review, 97, 377–395
Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer.
LeDoux, J. E. (2000). Emotion circuits in the brain. Annual Review of Neuroscience, 23,155–184.
Martin, W.L., Murray, P.S., & Bates, P.R. (2011). What would you do if.? Aeronautica 1, (1), 1-16.
Martin, W.L., Murray, P.S., & Bates, P.R. (2012). The effects of startle on pilots during critical events: A case study analysis. In Proceedings of the 30th European Aviation Psychology Association Conference, Villasimius, Sardinia, September 2012 Matthews, G., Davies, D. R., Westerman, S. J., & Stammers, R. B. (2008). Human performance. Cognition, stress and individual differences. Hove, UK: Psychology Press
Murray, P.S. & Martin, W.L. (2012). Beyond situational awareness: A skill set analysis for situational control. In Proceedings of the Tenth Australian Aviation Psychology Association Symposium, Sydney, Australia, November 2012.
Myers, K. M. & Davis, M. (2002). Behavioral and neural analysis of extinction,. Neuron, 36, (4), 567–584
NTSB (2010). Loss of Control on Approach, Colgan Air, Inc., Operating as Continental Connection Flight 3407, Bombardier DHC-8-400, N200WQ, Clarence Center, New York, February 12, 2009.Report NTSB/AAR-10/01. Retrieved from: http://www.ntsb.gov/doclib/reports/2010/aar1001.pdf
Qin, S., Hermans, E.J., van Marle, H.J.F., Lou, J. & Fernandez, G. (2009). Acute psychological stress reduces working memory-related activity in the dorsolateral prefrontal cortex. Biological Psychiatry, 66, (1), 25-32
Shea, C.H. & Wulf, G. (2005). Schema theory: A critical appraisal and re-evaluation, Journal of Motor Behavior, 37,(2), 85-102
Stratakis, C.A. & Chrousos, G.P. (1995). Neuroendocrinology and pathophysiology of the stress system. Annals of the New York Academy of Sciences, 771, 1-18
Thackray, R. I. (1988). Performance recovery following startle: A laboratory approach to the study of behavioural response to sudden aircraft emergencies. FAA Technical Report No. DOT/FAA/AM-88/4, Civil Aeromedical Institute, Federal Aviation Administration, Oklahoma City, USA.