Science, policy, and the public discourse of shark “attack”: a proposal for reclassifying human–shark interactions

The scientific system of species classification and description originated by Carl Linneaus (1758)—the same system that identifies humans as Homo sapiens—was the genesis for the label of “man-eater” for the white shark, Carcharodon carcharias. Linneaus’ description of this species included noting that “it strikes” (dorfo plano), has teeth of armor (dentibus ferrates), and was likely responsible for swallowing Jonah (Linneaus 1758, p. 235), whose story had been widely published in the 1679 Lectiones Morales in Prophetam Jonam by Angelo Paciuchelli and Charles de Marimont. Linnaeus’ historic volumes redefined the scientific and social world, and white sharks were singled out for their motivation as a man-eater.

From there, the story of the danger posed by sharks grew. In Thomas Pennant’s 1812 volume British Zoology, a stated characteristic of white sharks was their “greediness after human flesh” (p. 140). In 1845, Samuel Goodrich wrote that the shark is the “dread of mankind in the seas where it is found” (Goodrich 1845, p. 317). “Man-eater” came to virtually define white sharks (Jordan and Gilbert 1880), but the label also was used for other species in other areas. For example, in Hawai’i, the term niuhi meaning “man-eater” was used in native songs, most likely referring to the tiger shark (Titcomb and Pukui 1951, p. 4).

In Europe, colonial experiences shaped understandings of sharks. British big-game hunter Sir Samuel Baker’s expeditions in Asia and Africa led him to conclude that individual tigers, Panthera tigris, can become “man eaters” within a local area (Baker 1890; Blanford 1891). This concept of a predator that has acquired a “taste for human flesh” was projected onto sharks. In 1899, William Bryce wrote in the British Medical Journal about shark bites on three people on the same day in Port Said, Egypt. He noted that “many people have expressed the opinion that it must have been one shark which bit all three boys, and I think this very likely” (Bryce 1899, p. 1534).

Historical accounts at this time in the USA often differed, with scientists arguing that dangerous sharks could only be found in warmer southern waters. A report in the New York Times in 1865 recounted the story of Peter Johnson, a fisherman aboard a schooner in Lubec, Maine, who was bitten by a shark “that must have been of the species known as ‘man-eater’” (NY Times 1865). Yet the report notes that “man eaters” are “common in low latitudes” and “seldom, if ever attack mankind” (NY Times 1865). In 1916, interest in shark behavior rose dramatically following a cluster of fatal shark bites in New Jersey. Initially, Frederic Augustus Lucas of the American Museum of Natural History stated that it must not have been a shark because “sharks have no such powerful jaws” (Webster 1962, p. 87). He later concluded that he was wrong, and the shark in question must have been demented or “mad.”

In Australia, ocean swimming during the day was illegal from the late 1830s until 1902, due to concerns about propriety (Neff 2012), and the first reported shark bite was not documented until 1915, at an ocean beach in Sydney (Maxwell 1949). Fatal incidents followed at Sydney beaches, yet responses were limited. New South Wales Fisheries expert David Stead offered a statement on shark behavior in 1929, noting that “sharks do not patrol beaches on the off-chance of occasionally devouring human prey” (NSW 1929, p. 2). As a result, an Australian government report at the time referred to most cases of fatal shark bites as shark “accidents” (Neff 2012).

However, more shark bites ensued in Australia, and a 1933 study by Sydney surgeon Sir Victor Coppleson attempted to reconcile the competing international theories. Correspondence from the USA urged him to address this disagreement and to warn the public of possible “shark rabies” (Coppleson archives 1964). Coppleson concluded that the “evidence that sharks will attack man is complete” (Coppleson 1933, p. 466). Following publication of his article, the terminology in Australia changed to favor the more dominant shark “attack” language, which portrayed certain sharks as “man killers” (Coppleson 1933). Yet, the question of how to label different types of shark bites would persist, allowing the perception that all shark bites were of the “man-eating attack” variety.

In 1949, Australian author C. Bede Maxwell concluded that a “shark consciousness” was beginning to emerge, and she wrote that “‘accidents’ is the correct word to use in connexion [sic] with shark tragedies” (Maxwell 1949, p. 182). A global awareness of sharks was beginning to take shape with the deployment of hundreds of thousands of troops across the Pacific in World War II, bringing more sharks and people into contact than any previous time in human history. The US Navy produced a “Shark Sense” brochure to dissuade the concerns of pilots; however, the sinking of the U.S.S. Indianapolis in 1945, resulting in 60–80 fatalities by sharks (Sisneros and Nelson 2001), brought the subject of sharks and shark repellents to the forefront of government attention (Caldicott et al. 2001, p. 447).

Prior to the Indianapolis disaster, a chemical shark repellent called “Shark Chaser” had been developed and distributed to US military personnel at sea (Sisneros and Nelson 2001). The realization that this repellent was only partially effective led the Office of Naval Research (ONR) and other groups to focus further on the problem. Decades of ONR-sponsored research followed, resulting in great advances in our knowledge of shark physiology, behavior, and ecology (Gilbert 1963; Hodgson and Mathewson 1978). In 1958, a “Shark Research Panel” comprising Perry Gilbert, Sidney Galler, John Olive, Leonard Schultz, and Stewart Springer was established by the American Institute of Biological Sciences (AIBS), spurring a new discourse on categorizing human–shark encounters (Gilbert 1963; Caldicott et al. 2001).

That year, a meeting entitled “Conference on the Basic Research Approaches to the Development of Shark Repellents” was held in New Orleans, sponsored by AIBS, Tulane University, ONR, and the Navy Bureau of Aeronautics. The conference included 34 scientists from around the world. Shark “attack” classifications were suggested, with four categories outlined in a report by Leonard Schultz (Schultz 1963). These included: “unprovoked shark attacks,” in which sharks “make contact with the victim or gear;” “provoked shark attacks” that involve injuring, catching, or annoying sharks; “boat attacks” that involve contact with boating equipment; and “air and sea disasters” (Schultz 1963). To track trends in human–shark interactions around the world, the Shark Research Panel established the Shark Attack File, later known as the International Shark Attack File (ISAF), currently curated at the University of Florida.

The Schultz (1963) report advanced our thinking about shark attack in several ways. Bites or ramming on boats were separated out from bites on people, reducing the number of perceived “attacks” that occurred each year. The special cases of air and sea disasters were identified as unusual circumstances possibly evoking shark behavior not typical of that seen off ocean beaches. Most importantly, “provoked” responses by sharks that had been antagonized by swimmers, divers, or fishermen were separated from “unprovoked” attack behavior on what was perceived to be an innocent human victim. This provoked vs. unprovoked criterion is still used in today’s ISAF.

During the same post-WWII period, people began using ocean beaches for recreation at an increasing rate, as leisure time and personal and public transportation all rose after the mid-1940s. In the USA, “vacation travel boomed…and beaches on the East, West, and Gulf coasts were particularly popular destinations” (Harper 2007, p. 37). This expansion occurred in Australia and South Africa as well. Davies (1964, p. 141) stated that in South Africa, the “increased usage of the sea is related to such factors as increasing population, improved methods of transport, shorter working hours, and the increase of leisure time.” As a result, the growing popularity of ocean swimming, surfing, snorkeling, and scuba diving brought more people into potential contact with sharks every year.

This analysis gave sharks human agency and moved them from unseen “monsters of the deep” to a potentially more terrifying image as resident serial killers lurking in wait for human prey. Coppleson thus concluded that “such a shark must be hunted until it is destroyed” (Coppleson 1950, p. 8). It is this perceptual change in the treatment of shark behavior that constitutes the criminalization of this human–animal interaction (Michalowski 1985).

While sharks are not traditional criminals and shark bites are unconventional crimes, this characterization is consistent with the way actions and groups receive benefits or burdens. Jenness (2004, p. 150) notes that criminalization “targets a set of activities perceived to be attached to a social group deemed ‘in need of control’ by those in a position to stimulate, define and institutionalize criminal law.” In this case, sharks’ swimming is no longer considered innocent behavior; instead, this patrolling or cruising constitutes a threat. In David Webster’s book Myth and Maneater (1962, pp. 50, 67), this change can be seen as shark bites on people are labeled as a “molestation” and “assaults.” In response to these concepts of shark behavior, “shark control” programs are enacted by nations around the world. Shark bites have been treated as if they were crimes in the USA, South Africa, Australia, and elsewhere. During the spate of incidents in New Jersey in 1916, a request for action was discussed during a War Cabinet meeting (Webster 1962). Following a series of four fatal shark bites in 13 days in 1957 along the Natal South Coast of South Africa, rogue sharks were identified as the culprits (Davies 1964, p. 71). In response, the South African Navy was enlisted, and a frigate, the S.A.S. Vrystaat, was sent from Cape Town to drop more than sixty 100-lb depth charges near the shoreline off the South Coast (Davies 1964). In the USA in 1961, fear of shark attacks led to directed longline fishing for sharks off the coast of New Jersey (Carlson et al. 2008). In Australia, culling nets were added to Queensland beaches in 1962 (Gribble et al. 1998). These measures were consistent with current scientific understanding of sharks. Indeed, Davies (1964, p. 65) wrote that the “Blue Pointer [a previous name for white sharks] is a voracious and aggressive species which attacks humans and even small boats with little hesitation.” The prevailing attitude at the time had become “the only good shark is a dead shark” (Gruber and Manire 1991).

By the time the best-selling novel Jaws and the subsequent blockbuster movie appeared in 1974 and 1975, respectively, the criminalization of shark bites was nearly complete: the “man-eater” label implied an intent-driven monster that was seeking human prey. The rare discoveries of human remains inside sharks reinforced this image and validated scientists’ worst fears (Maxwell 1949). The rogue shark theory appeared to provide a scientific basis for malicious shark behavior, from which there could be only one outcome in human–shark interactions: the fatal outcome resulting from being hunted, killed, and eaten alive by a shark. When Peter Benchley’s book appeared, the story of a rogue shark terrorizing a seaside town brought the graphic fear of criminal sharks home to millions. The enactment of anti-shark policies including shark hunts, shark derbies, and beach nets became punitive measures for the perceived public good. Recreational “monster fishing” for sharks skyrocketed after 1975, and shark “kill” tournaments in the USA became more popular than ever before (Hueter 1991). More recently, real shark hunts mirroring the fictional response in Jaws have been used following clusters of shark bite incidents in Egypt, Russia, the Seychelles, Mexico, Réunion Island, and Western Australia (Neff and Yang 2012).

Jaws was a worldwide phenomenon and framed a clear story: humans were on the shark’s menu. Risk theorist Paul Slovic (2004) noted that for hazards like shark bites, images, words, and symbols can be triggers to paint a picture of scary outcomes. He further stated that “[w]e have found that every hazard has a unique profile of qualities (much like a personality profile) that influences perception and acceptance of its risk” (Slovic 2004, p. 985). After Jaws, the shark “attack” profile was linked to the unforgettable images of the film and reinforced one, and only one, vivid and dreaded outcome. All shark “attacks” were perceived as equal, and wherever sharks roamed—which is off most swimming beaches of the world—going into the ocean meant you were risking your life to the bloodthirsty jaws of a shark.

As a label for a broad array of human–shark interactions, “shark attack” is an erroneous characterization for a number of reasons. Shark “attack” language unreasonably amplifies social perceptions of risk. It provides a heuristic model that facilitates mental shortcuts to connect words with images and feelings (Slovic 2004). A loaded phrase such as “shark attack” compounds the perceptions of other crimes and outcomes, similarly to the term “home invasion.” Such “perceptually contemporaneous (PC) offenses” can cause potential victims’ images to cascade into perceptions about other events (Warr 1984, 1987). For example, the fear of a robbery is amplified by the perceived connection it has with being assaulted, raped, or murdered. Likewise, the term “shark attack” does not elicit thoughts of minor scrapes or bites; instead, it conjures horrific, bloody scenes of being consumed alive by an evil predator.

As a result, the generic term “shark attack,” which is used to label a multiplicity of human–shark interactions, is misleading to the public. Reports of shark “attack” make little distinction between minor events and fatal incidents. Bites from non-threatening sharks like the wobbegong, which account for 5.5 % of all shark “attacks” in Australia since 1900 (NSW 2009), are not distinguished from more serious bites by other species of sharks when all events are labeled shark “attacks.” The term “shark attack” can even include events where there is no physical contact with a person. For example, sharks simply making contact with kayaks may be counted and reported as “attacks” (NSW 2009). Clearly, when the phrase “shark attack” is used, the public is led to conclude that this must involve direct contact resulting in major injuries to the “victim.”

In addition, shark attack language has been politicized, and the shark “attack” label is sometimes used for political purposes when shark bites occur. The words “shark attack” can create a perception of a premeditated crime, lowering the public’s threshold for accepting shark bite incidents as random acts of nature. The narrative establishes villains and victims, cause and effect, perceptions of public risk, and a problem to be solved. A shark fatality in Western Australia in 2011 led the local shire president to state that “[a] lot of people say the water is the shark’s territory, but I think if they can find the shark (responsible) they should get rid of it.” And he added, “[i]f they have attacked [our italics] a human in one of those areas they may want to do it again” (Hickey 2011). Policies that respond to shark “attacks,” therefore, may be more understandable and acceptable to the public when they rely on familiar stereotypes, even if they have no relationship to what actually takes place. In short, a pattern exists in which the designation of a shark “attack” raises media attention that provokes a government response, even when the event may not be serious or governable (Neff and Yang 2012).

It is not surprising, therefore, that the selection of policy responses can include “hunting the shark down,” drawing from the movie legend. This reaction raises an important final point, that the “rogue” shark theory is unsupported by scientific data. Although sharks may have ample opportunity to feed on humans, the incredible rarity of incidents suggests that shark bites in most cases represent what amounts to a tactical and biological dead end for the sharks. Raising public awareness of this involves conceptualizing the scientific data in a way that makes this clear, even as dramatic and sometimes fatal incidents occur and are publicized. Fatal shark bites in 2010–2012, in Sharm El Sheikh, Egypt, the Seychelle Islands, Réunion Island, and off Western Australia, brought back stories of “rogue” sharks, yet the post-Jaws era science stands firm on this issue: sharks have not put humans on their menu.

In 1974, David Baldridge’s groundbreaking analysis of shark bite data, conducted at Mote Marine Laboratory in Florida, introduced the first outcome-based approach to categorize human–shark interactions. A number of Baldridge’s conclusions ran counter to Coppleson’s. For example, Baldridge did not support Coppleson’s argument that water temperature was the primary dependent variable linking shark feeding behavior with shark attack predictability. Baldridge (1974) stated that “the correlation between warm water and shark bites is a reflection of when people are more likely to go into the water, not when sharks are excited to feed” (p. 17). In addition, Baldridge confirmed Springer’s (1963) hypothesis that, beyond a certain water temperature, sharks show an inverse relationship between feeding and temperature. Baldridge (1974) stated that “the real shark attack season” would be that time when local waters are warm enough for human swimmers (above 70 °F) and not too hot for sharks (less than 85 °F) (Baldridge 1974, p. 18).

Over the past 30 years, many shark biologists have dedicated themselves to understanding shark behavior, including biting and feeding patterns. In 1984, Timothy Tricas and John McCosker suggested that white sharks might mistake humans for seals in certain circumstances, resulting in a “bite and spit” behavior that could explain why human fatalities do not always occur in white shark attacks (Tricas and McCosker 1984). Samuel Gruber reviewed the analyses by Baldridge (1974) and Nelson et al. (1986) and decided it is unreasonable to draw conclusions about why sharks bite without looking at the totality of shark motivation and behavior and ecosystem conditions (Gruber 1988, p. 10). He stated that the contributing factors for shark bites include random “opportunity,” “interference” with reproductive activity, defensiveness against a threat or competition, and trespassing into a shark’s space. Gruber concluded “the majority of cases where sharks act aggressively against humans are probably motivated by social factors such as fear, aggression or sex and are entirely unrelated to feeding” (Gruber 1988, p. 12).

This discussion reveals a primary shortcoming of the Schultz (1963) categorization of “provoked” vs. “unprovoked” shark attacks. It is easy to classify cases of divers spearing sharks, grabbing their tails, or attempting to ride them, or fishermen hooking sharks and pulling them into their boats, as “provoked” incidents. What is not so easy to judge is if a swimmer who passes very close to a shark “minding its own business” is, or is not, provoking that shark to respond to a violation of its personal space by biting the intruder. This and other examples underscore the failings of motivation/intent-laden categories of human–shark interactions, such as the Schultz (1963) provoked/unprovoked distinction, rather than an outcome-based system. We simply do not, and probably cannot, understand enough about the specifics of shark behavior in all cases to classify the motives and intentions of every biting shark in a clear, objective way.

More recently, several factors have driven an alternative narrative to the human–shark interaction story. These include: (a) the emergence of shark conservation biology demonstrating the worldwide vulnerability of sharks, including white sharks; (b) a decline in shark bite fatality rates due to better first responder medical care; and (c) an increased awareness of the diversity of human–shark outcomes, even when shark bites occur. Heneman and Glazer’s (1996) article documenting new legal protection for white sharks, successfully implemented in California in 1994, was entitled “More rare than dangerous” and noted that a shark bite victim wrote in favor of protection despite having been “mauled” by a white shark (p. 485). Analysis by Caldicott et al. (2001) highlighted the increasing survival rates from shark bites, up to 87 % (p. 449). The authors of that study identified a number of shark bite situations including “hit and run,” “sneak attack,” and “bump and bite” (Caldicott et al. 2001, p. 449).

Efforts to focus on the diversity of outcomes from human–shark interactions, even when shark bites occur, have increased over time. Immediately following the US east coast’s so-called “Summer of the Shark” in 2001, one of the authors (RH) recommended that most “shark attacks” should more correctly be referred to as “shark bites,” in the same way that we distinguish an aggressive but nonfatal “dog bite” from a serious, sometimes fatal “dog attack.” This author particularly focused this message on the news media and its overuse of the term “shark attack” that resulted in a misled and sometimes panic-stricken public. Other scientists also have begun to advocate for a change in “shark attack” language. Cliff (1991) used a modified definition to calculate shark bites in South Africa. Recent research by Lentz et al. (2010), using data from the ISAF, led to the development of a “shark bite severity scoring system” that identifies five levels of shark bite injuries, from minor to fatal wounds, providing a leap forward in clinical assessments of shark injuries. Extending into the popular culture, in her 2011 book Demon Fish, journalist Juliet Eilperin sought to find a middle ground by referring to shark bites as “strikes” (Eilperin 2011).

Observations of sharks in proximity to human swimmers in the ocean demonstrate that the animals do not usually take an interest in people. Sightings by New South Wales Fisheries staff, including Vic Peddemors and Amy Smoothey, have revealed that bull sharks regularly swim close to hundreds of human swimmers in Sydney Harbour and ignore them all (ABC 2011). In Cape Town, South Africa, the Shark Spotters program has reviewed more than 1,100 sightings of white sharks swimming around surfers and near bathers. Bathers were alerted and got out of the water, and the visiting sharks swam away (Shark Spotters 2012). This story repeats itself in Port Stephens, Australia, where shark biologist Barry Bruce has studied juvenile white sharks that consistently ignore people in the nearby surf (Gilligan 2012). These observations point to the need for a more sophisticated public education effort on the subject of shark aggression (and nonaggression) towards humans. One example of such an effort can be seen in beach signage used by the city of Port Lincoln in South Australia. The signs identify varying levels of concern when white sharks are in local waters, using sighting categories of: (a) sharks away from shore, (b) sharks inshore, and (c) emergency situations where humans and sharks are very close together.

Clearly, classifying virtually all contact between sharks and humans as some form of shark “attack” misrepresents the facts and misinforms the public. It is invalid on both scientific and public policy grounds. Different language and categories of human–shark interactions are needed.

Australia and the USA provide valuable case studies for this analysis of shark attack terminology, as these nations lead the world in the number of reported shark “attacks” (ISAF 2012a). The US state of Florida was ground zero for the “Summer of the Shark” in 2001, while the Australian state of New South Wales (NSW) and its capital Sydney saw its own “Summer of the Shark” in 2009. The case studies below examine legacies from these episodes. In NSW, the government’s 2009 report on shark “attacks” is evaluated using our new categories; in Florida, the state’s global reputation as a hot spot for shark “attacks” is reconsidered.