US Court of Appeals
District of Columbia
OSHA's Legal Brief vs SeaWorld
Beginning 12 November 2013
Incidents Occurred Regularly for Twenty Years Despite SeaWorld’s Efforts to Prevent Them
SeaWorld’s argument that its safety protocols and operant conditioning program provide sufficient protection to its employees is disproven by the 600 pages of incident reports documenting unanticipated and undesirable killer whale behavior with trainers. The reports describe some 100 occurrences of killer whales biting, hitting, lunging toward, pulling on, pinning, dragging, and aggressively swimming over SeaWorld trainers. SeaWorld claims the frequency of such incidents has tapered off over time, but there have been incidents every year but two since 1988, culminating in trainer deaths in 2009 and 2010.
The incident reports not only document harmful killer whale behavior, but also show that, time after time, SeaWorld had no explanation for why an incident occurred and was ineffective at preventing similar or even identical behavior from happening again. Although not all incidents resulted in serious injury, a number were chillingly similar to incidents with less happy endings. One example of whale aggression that continued regularly over twenty years is whales grabbing trainers’ feet or legs, often pulling them into or under the water.
Operant Conditioning Does Not Keep Trainers Safe
SeaWorld depends almost exclusively on operant conditioning to ensure safe
interactions with its captive killer whales. As the implementers of operant conditioning, trainers are thus responsible for their own safety and are “the primary source of management’s knowledge.
Through operant conditioning, trainers are expected to recognize precursors to aggressive or
unwanted behavior and to respond appropriately, whatever that may mean (when presented with unexpected behavior by a killer whale, “the trainer has to make a judgment there on the spot as to how to deal with a new scenario.”).
Unfortunately, operant conditioning is an imperfect system. It has 2 major flaws:
1. Killer whales do unpredictable things and
2. Trainers make mistakes
These facts are amply documented in the incident reports. Examples include:
A: “I think this is one of those situations where we will never quite understand the intent of [the whale’s] movement. I cannot rule out that it was done either on purpose or by accident. I think we have learned our lesson of our unpredictability of our animals even in the best of situations.”
B. “Knootka is a very unpredictable whale.”
C. “Because [the whale’s] never done anything like this, no one was expectingit.”
D. “As evident by this episode, our whales should never been [sic] viewed as routine, nor predictable.”
E. “I am still very confused that mistakes like this can be made by our senior trainers.”
F. “[The trainer] put himself in a very compromising situation.”
G. “[The trainer] should never have attempted to get out of the water while the whale still had a hold of her sock.”
SeaWorld argues that it learns from each incident and attempts to prevent similar
recurrences. However, as the incident reports demonstrate, SeaWorld is not as successful as it purports to be in preventing recurrences of dangerous killer whale behavior. In addition, SeaWorld’s “learn as you go” approach means that trainers are continually at risk from novel or unanticipated whale behaviors.
SeaWorld appears proud of the fact that its employees “controlled their own exposure to the alleged hazards,” calling this a “culture of empowerment.” However, placing the responsibility for employee safety on the employees themselves is, as the ALJ correctly pointed out, in contravention of the OSH Act. “An employer cannot shift [OSH Act] responsibility to its employees by relying on them to, in effect, determine whether the conditions under
which they are working are unsafe.” ("Final responsibility for compliance with the requirements of this act remains with the employer.”).
Likewise, the fact that some employees “testified that they felt safe” does not mean that they were safe. “The particular views of work[ers] are not necessarily, and often times are not the best determination as to what is safe and what is unsafe.”
Finally, SeaWorld contends that it cannot be held responsible for mitigation of the hazard posed by close contact with killer whales because all of the potential harm to its employees comes from “exceptional and unpredictable whale behavior.”
SeaWorld even suggests that Ms. Brancheau’s death was beyond the company’s control. (“SeaWorld could not have predicted this terrible incident”).
But the whole point is that SeaWorld knows killer whale behavior is unpredictable. Given the known unpredictability of killer whale behavior and the record of past incidents, it was entirely foreseeable that an event like Dawn Brancheau’s death could occur. That Tilikum attacked Ms. Brancheau without providing the “precursors” SeaWorld relies on shows the failure of operant conditioning to keep trainers safe; it does not render Ms. Brancheau’s death – or the hazard of close contact with killer whales – unpreventable.
SeaWorld’s Emergency Procedures Do Not Keep Trainers Safe
The second major failing in SeaWorld’s operant conditioning program is
that the emergency rescue procedures meant to “recall” or distract a whale from
dangerous behavior have proven to be grossly inadequate. The incident reports
document at least seventeen instances, dating to 1989, where killer whales ignored
attempts to “recall” them from unwanted behavior.
Most recently, recall attempts were useless in the deaths of both Alexis Martinez and
Dawn Brancheau. SeaWorld’s emergency procedures are contained in its “Animal Training
SOP.” When an emergency occurs, trainers are to sound a siren, after which “the senior ranking trainer should attempt to establish control of animal(s) in the environment with recall stimuli.” These “recall stimuli,” which SeaWorld has been using at least since the late 1980s, include trainers slapping the water and the sound of “recall tones.” The killer whales are supposed to respond to these signals by “calmly swimming to [the] stage.”
As the incident reports recount, however, and as SeaWorld employees admitted at the hearing, these procedures are ineffective in most serious emergencies. Kelly Flaherty Clark admitted that both a recall tone and recall slaps were attempted when Dawn Brancheau was in the water with Tilikum; she also admitted that there was no expectation that they would work.
Brian Rokeach agreed:
Q: So, Sea World knows from experience that emergency callback procedures performed while the whale is in a heightened state, if you will, will rarely succeed in getting the whale tocome back?
A: I guess – I’m sorry, there hasn’t been a lot of success in that specific scenario.
In an incident in 2004, killer whale Kyuquot repeatedly swam over trainer Steve Aibel. Mr. Aibel wrote: “[The whale] blocked my exit from the pool and sat in front of me. I asked for a recall tone and paired it with a point to control. There [sic] were both ignored. There were two more recall tones and three or four more hand slaps. All were ignored.” Ex. C-6 at 749. Indeed, SeaWorld staff apparently decided later that these recall tones and hand slaps did nothing but agitate the
One (unidentified) SeaWorld commenter observed, “Let’s face it, in these types of incidents, I don’t recall any whale responding to any hand slap, food bucket, or any other distraction we tried to implement.”
SeaWorld continued to use recall tones and hand slaps for more than twenty years even though such techniques were demonstrably unable to keep trainers safe. As the ALJ found, “[d]espite the repeated failures of the recall signals, SeaWorld continued to rely on them to protect its employees.”In short: Two killer whales trained under SeaWorld’s operant conditioning program killed two trainers two months apart. Under these circumstances it cannot be said that SeaWorld’s training program has reduced the recognized hazard to a significant degree. It clearly did
not eliminate the recognized hazard. The Secretary has established SeaWorld’s safety training program, both for killer whales and for its trainers, is inadequate as a means of feasible abatement.
This factual finding is supported by substantial evidence and should not be disturbed by the Court.