Content uploaded by Randall Lockwood
Author content
All content in this area was uploaded by Randall Lockwood on Oct 26, 2016
Content may be subject to copyright.
bulls is significantly higher than for other breeds
(Pickney and Kennedy 1982; Multani and Clifford
1985; Wright 1985). However, many factors, in-
cluding the following, can bias breed-specific
bite rates:
COMMENTARY
ARE"PIT BUllS" DIFFERENT?AN
ANALYSISOF THE
PIT BUll TERRIER
CONTROVERSY
Randall Lockwood and KateRindy
One of the most controversial aspects of animal
control legislation is characterizing a dog as vi-
cious, or in some way restricting ownership of a
breed, on the basis of breed description alone.
Thus far, breed-specific regulations have affected
only "pit bulls," but breed clubs and other or-
ganizations of dog owners fear that such restric-
tions might extend to other breeds in the future
(Lockwood 1986). This article reviews some of
the historical, ethological, and epidemiological
evidence relevant to the question of whether pit
bull terriers present special animal control
problems justifying unusual legislative action.
From an epidemiological perspective, it is dif-
ficult to draw scientifically sound conclusions
about the dangers posed by a specific breed.
Many lists of the breeds most likely to bite have
appeared in the popular press, but accurate
breed-specific bite rates are very difficult to com-
pute. Such statistics require good data for both
the numerator (number of bites attributed to a
particular breed) and the denominator (number of
animals in that breed). To get good data, one
must have detailed and accurate reports of all
bites, including reliable information about the
breed(s) and registration of all animals in ques-
tion, and detailed demographics of the whole dog
population of the community.
Several studies suggest that the bite rate for pit
Randall Lockwood, Ph.D., and Kate Rindy work in the Higher
Education Programs of the Humane Society of the United States,
2100 L 51. NW, Washington, DC 20037.
This article is reprinted from The Pit Bull Terrier Controversy
(1967) by permission 'of the Tufts Center for Animals, Boston,
Mass.
2 ANTHROZOOS, Volume I, Number 1
1. Overreporting of bites attributed to a particular
breed
2. Difficulty in identifying a particular breed
3. Underreporting of the population of a par-
ticular breed, including aberrant registration or
licensing rates
4. A tendency to find specific breeds within
populations of dog owners more likely to
maintain their animals irresponsibly
All of these factors may apply to analyses of pit
bull bite rates.
First, dog fighting and bites attributed to fight-
ing breeds attract considerable attention in the
print and electronic media. If a community is
having a problem with dangerous dogs, any bite
or attempted bite involving pit bulls is likely to
find itsway into newspapers and local records.
Second, people commonly use the term pit
bull to describe a variety of registered and un-
registered dogs, including the American pit bull
terrier (registered by United Kennel Club and
American Dog Breeders Association), the
American Staffordshire terrier, the Staffordshire
bull terrier, the bull terrier, and the bulldog (all
registered by the American Kennel Club), and
many mixtures of these breeds with one another
and with other breeds.
There is considerable controversy over the
ability of animal control officers, law enforce-
ment officials, and veterinarians to positively
identify individual dogs as pit bulls. One survey
of over 2,000 bite reports (Beck, Loring, and
Lockwood 1975) four)d that any medium-sized
black and tan animal was likely to be recorded as
a German shepherd. Similarly, any stocky short-
haired animal involved in an attack is likely to be
recorded as a pit bull. It is not unusual to find
newspaper accounts of "pit bull attacks" accom-
panied by a picture of a boxer, pug, or some
,]
Randall Lockwood and Kate Rindy
other breed.
Third, low estimates of the total population of
specific breeds will inflate bite rates. Several of
the preceding studies used AKC registrations to
estimate the frequency of bites for various breeds
in the total population of dogs. This approach is
likely t~ produce erroneous results for pit bulls,
since many pit bull owners register their dogs
with organizations other than the AKC (including
UKC and ADBA), and very few dogs have dual
registration. Also, pit bull owners are probably
less likely to register or license their animals than
owners of other breeds, given past attempts to im-
pose restrictions on the breed.
Finally, although there are many well-bred pit
bull terriers with responsible owners, the tradi-
tional association of pit bulls with illegal dog
fighting means that a disproportionate number of
these dogs belong to that class of dog owners
likely to exercise less responsibility for the care
and supervision of any type of dog. Bite rates,
therefore, may say more about the irresponsibility
of owners who tend to prefer pit bulls than about
the viciousness of pit bulls themselves.
The few communities that claim to have docu-
mented higher bite rates for pit bulls have not
taken into consideration the confounding factors
just mentioned. Unfortunately, there are no
statewide or nationwide reporting systems that
would support epidemiological generalizations.
When addressing problems in the real world,
however, it is important to separate issues of
epidemiology and ethology from issues of public
safety and legislation. The public may demand
protective legislation if it perceives any increased
bite risk, no matter how small, to be associated
with a breed. Recent court actions suggest that
law enforcement and animal control agencies
have a broad mandate to give the protection of
the public priority over the right to own property
that may cause harm. In the absence of con-
clusive data, legislators tend to err in the direction
of safety.
With these problems in mind, we can address
several questions. First, are there biological or
ethological reasons why pit bull. terriers might
present special dangers? Second, do all pit bulls
present these dangers? Ifnot, are there other pre-
dictable factors associated with dogs or owners
that are likely to cause harm?
Are "Pit Bulls" Different?
A review of the origins of pit bulls provides
some insight. We do not intend to provide a
detailed history of the various pit-bu II-type breeds
in this paper. For in-depth information see Pit Bull
Report (Lockwood and Miller 1986) or other
standard references (for example, Matz 1984;
Semenic 1984). We can, however, briefly point
to some illuminating facts. All dogs ofthe pit bull
type trace their ancestry to the bulldogs of the
nineteenth century. These animals were originally
used for bullbaiting in England. When England
passed laws against bullbaiting in 1835, or-
ganized dogfighting became popular, resulting in
a proliferation of smaller dogs bred for combat.
These smaller dogs became popular in America
ataboutthetime of the Civil War.
The United Kennel Club was founded in 1898
to register pit bull terriers and to standardize the
rules of dog fighting. In 1935 the American Ken-
nel Club, which had registered pit bull terriers
since itsfounding, began registering them as Staf-
fordshire terriers; it renamed them American Staf-
fordshire terriers in 1972. Today, even though
both AKC and UKC have taken a stand against
dog fighting, owners of registered dogs still pride
themselves on their animals' "gameness."
A long history of breeding for bullbaiting and
fighting has had a profound effect on the genetic
predisposition of many of the fighting breeds. In
many cases, a shorter history of selection for
qualities that might make these animals suitable
as household companions has counteracted this
effect. The extent to which breeding has altered
the original temperaments of fighting breeds is
often difficult to predict in a given animal. The
following characteristics of fighting dogs con-
tribute to the problems these animals may
present:
1. Aggression against dogs and other animals.
The primary quality for which pit bulls have been
selected is "gameness." A game animal is ready
and willing for combat and unyielding in battle
with another creature. Certain genetically based.
characteristics reflect gameness. One important
characteristic isa low level of fighting inhibition.
Most wild and domestic dogs fight one
another only to drive a rival away from some dis-
puted object~food, mate, or territory. The attack
ends when the rival withdraws or displays signals
ANTHROIOOS, Volume I, Number 1 3
of surrender. Bluff, such as growling or staring, is
usually the preferred tactic. Actual attacks are
usually last-resort confrontations. In fighting
breeds this inhibition against attacking has been
selected against. These animals will fight with no
provocation, and a game animal will fight until
complete exhaustion or death. Inthis sense, these
animals ~are not "doing what comes naturally."
Their behavior is totally abnormal in an evolu-
tionary or ecological sense and is strictly the
result of human intervention. This lowered inhibi-
tion against aggression may also apply to other
species, particularly smaller animals such as cats.
The flight of a potential prey animal usually
triggers predatory attacks in wild and domestic
dogs. This is why many breeds may pursue or at-
tack moving people or objects like joggers,
bicyclists, and cars. But animals selected for
bullbaiting and fighting had to show gameness
against animals that were either restrained or
confined, so these animals and their descendants
are more likely to attack targets that do not flee or
show other "provocative" behaviors.
Gameness also seems to include a genetically
based lowering of sensitivity to pain. Many fight-
ing breeds show no outward sign of disturbance
by severe injuries.
2. Decreasedcommunication. Dogs, like wol-
ves, are highly social and possess a rich reper-
toire of signals to communicate their moods and
intentions to others. Animals selected for fighting
do not reveal their intentions or weaknesses and
are not inhibited by opponents' displays of sub-
mission or surrender. Fighting dogs offer little or
no indication that they are about to charge or at-
tack. They often fail to give warning with a growl,
an aggressive facial expression, or other sign.
They are often insensitive to behaviors that usual-
ly stop aggression. For example, dogs not bred for
fighting usually display defeat in combat by roIl-
ing over and exposing a light underside. On
several occasions, pit bulls have been reported to
disembowel other dogs offering this signal of sub-
mission.
3. Attack behaviors. Dogs use many different
styles of attack against members of their own and
other species. Many breeds have styles of biting
that reflect the purposes for which they were
bred. Guard dogs such as German shepherds, for
example, tend to restrain their enemies by grab-
4 ANTHROZOCS, Volume I, Number 1
bing and holding. The fighting breeds, on the
other hand, have been selected to inflict maxi-
mum damage on their opponents by sustained
grabbing, holding, shaking, and tearing. To our
knowledge, there is no direct evidence of un-
usually great biting force in fighting dogs. These
animals do not possess any unusual adaptations
for "locking" their jaws. The increased destruc-
tiveness of pit bull bites is attributable to the be-
havioralfactors of persistence and stamina rather
than to any biomechanical factors.
4. Aggression toward people. The fighting
dogs of the nineteenth century generally posed
little or no threat to people. These animals were
disqualified in the pit if they exhibited aggression
to their handlers or other people. Early in this
century several former fighting breeds such as the
bull terrier and the English bulldog were specifi-
cally selected for their good dispositions around
people.
As mentioned earlier, AKC, UKC, and ADBA
animals are all descended from fighting stocks.
Breed standards for the American Staffordshire
terrier and the American pit bull terrier make little
or no reference to temperament, although an
animal that attacks a person or a dog in the show
ring may be disqualified. Many individual
breeders have attempted to produce animals with
stable dispositions toward people. There are
many examples of well-behaved dogs of these
breeds, but there have been no uniform standards
in this direction. Nonregistered and pit-bull-mix
dogs, which are becoming increasingly popular,
have been subjected to even less selection for
stable temperament than their registered counter-
parts.
The widespread practice of hybridizing
American Staffordshire terriers and American pit
bull terriers with other breeds can produce par-
ticularly dangerous animals. American Staf-
fordshire and pit bull terriers were bred to show
little aggression to people. Other breeds with
which they are commonly hybridized, such as
the German shepherd, Bullmastiff, Rottweiler,
and Rhodesian ridgeback, have been selected for'
use as guard dogs against human intruders. The
result can easily be an animal with the fighting
potential of the classic pit dogs and the potential
aggressiveness to people of guard dogs.
Randall Lockwood and Kate Rindy
Table 1. Fatal Dog Attacks (1986)
Date Place Dog(s) involved
Victim's
age/sex
12/29
11/21
10/26
9/- ~_.~.
9/18
9/2
Apison, Tenn.
Decatur, Ga.
Denver, Colo.
Elizabethtown, Pa.
Forest City, N.C.
Dallas, Tex.
F, 3 yrs
M, 4 yrs
M, 3 yrs
M, 7 yrs
M, 4 yrs
F, 14 mos
1 malamute
3 pit bulls
1 pit bull
1 coonhound
1 wolf-shepherd
1 pit-bull-mix
1 mixed-breed ,.
1 "husky-type"
1 pit bull
1 wolf-husky
1 pit-bull-boxer
1 mixed-breed
1 pit bull
At least 4 pit bulls
The lack of uniform standards of temperament,
the lack of inhibition of aggression, the strength
f
and tenacity of attacks, and the failure to show
appropriate warning signs of aggression all repre-
sent potential risks associated with fighting breeds
and their hybrids. As previously mentioned, there
is little reliable evidence about breed-specific bite
rates. We have conducted research to gain some
additional insight from two other sources--
reports of fatal attacks and a survey of press
reports of dog bites.
Although many dog bites go unreported to
either the press or to a board of health, we are
certain that virtually all dog-related fatalities are
reported. When we learn of a dog-related fatality
through local humane groups, veterinarians,
health departments, or the press, we contact the
appropriate authorities to get a complete record
of the incident and subsequent investigations. In
several cases, we have been able to conduct on-
site investigations.
In 1986 we received reports of 12 fatalities
fromdog attack (seeTable 1). Seven of these at-
tacks involved at least one pit bull. Eleven of the
12 fatalities involved children aged 7 or under. A
less comprehensive survey of fatal attacks be-
tween October 1983 and December 1984
yielded reports of 9 additional fatalities, 7 of
Are "Pit Bulls" Different?
which involved at least one pit bull. Thus, two
thirds of the fatalities we have learned of during
the last three years have involved pit bulls. Past
and current AKC and UKC registrations and AKC
estimates of the ratio of unregistered to registered
dogs show that there are roughly 500,000 to 1
million pit-bull-type dogs in the United States, or
an estimated 1-2% of the entire dog population.
It seems clear, then, that pit bulls are over-
represented in the small population of dogs in-
volved in human fatalities.
The injuries inflicted by pit bulls in the cases
we have studied are noticeably different from the
injuries inflicted in fatal attacks by other breeds.
Pit bull victims typically had large portions of tis-
sue torn away, whereas victims of other breeds
typically died from a smaller number of exsan-
guinating injuries orfrom a single crushing injury
to the brain or spinal cord. We are preparing a
more detailed review of these incidents.
In order to gain insight into serious but nonfa-
tal dog attack injuries, we reviewed press clip-
pings of 278 dog attacks compiled by two clip-
ping services from approximately 1,100
newspapers forthe period from January"1986, .
to October " 1986. We abstracted as much infor-
mation as possible from each report, following
the format used by Beck, Loring, and Lockwood
ANTHROIOOS, Volume I, Number 1 5
7/1 Kobuk, Alaska F, 2 yrs
6/10 Ramsey, Mich. M, 20 mos
5/5 Anchorage, Alaska F, 2 yrs
4/24 Osteen, Fla. M, 79 yrs
4/10 Gresham, are. M, 5 yrs
1/26 Longview, Tex. M, 6 yrs
(1975) intheir survey of police reports.
We realize that we cannot use this analysis to
draw breed-specific conclusions about bite rates,
since there may be a tendency to report pit bull
attacks more often than others. In fact, 143 of the
reports, or 51.4%, dealt with pit bull incidents. Of
the remainder, 11.5% dealt with German
shepherds or German shepherd mixed breeds,
7.2% 'with Dobermans, 4.7% with Labradors,
2.9% with Chows, and 22.3% with other uniden-
tified breeds. Thus, we are not asking the ques-
tion "Are pit bulls different?" but instead are as-
king "Are pit bull attacks different?" Our analysis
of press clippings indicates several relevant dif-
ferences.
Beck, Loring, and Lockwood (1975) reported
that most serious dog bite cases involve children,
and our analysis agrees (see Table 2). There is,
however, a higher proportion of adolescents and
adults among pit bull victims (54.1 %) than
among victims of other breeds (38.1 %). This sug-
gests that greater size and maturity are less of a
defense against pit bulls than they are against
other attacking breeds. Familiarity with the
animal also appears to provide less protection in
the case of pit bulls. Out of 143 pit bull attacks,
19 (13.3%) involved attacks on the owner; out of
135 attacks by other breeds, only 3 (2.2%) in-
volved the owner.
We characterized an injury as serious if the
report indicated a need for suturing, hospitaliza-
tion, or other medical intervention. Of the 143
reports of pit bull attacks, 55 (38.5%) were
serious. Of the 135 reports of attacks by other
6 ANTHROZOO5, Volume I, Number 1
breeds, only 36 (26.7%) were serious. These
figures suggest that the press is not more likely to
report nonserious pit bull bites just because they
involve pit bulls; if they did, we would expect a J
higher proportion of reports of nonserious pit bull'i
bites. Out of the 91 serious bites reported,
however, over half (60.4%) involved pit bulls.
Thus, these reports indicate that pit bulls are
more likely to be involved in serious bites, and
serious bites tend to involve pit bulls more often
than other breeds.
Two other measures ofthe severity of bites are ~
the incidence of bites to the face and the number j
of bite"s involving multiple injuries to several .~
body areas. Victims under 9 years of age tended
to receive a high proportion (around 60%) of fa-
cial bites from al/ breeds. Pit bulls do not inflict
more facial injuries than other breeds to any
group. However, pit bulls are more likely to in-
flict multiple injuries on older victims: 35% of
older pit bull victims received multiple injuries,
compared with 18.5% of older victims of otherj
breeds. !
Previous studies of dog bite epidemiology (for
example, Beck, Loring, and Lockwood 1975) sug-
gest that the majority of incidents involve free-
roaming, owned animals. Virtually all the dogs in
the cases we studied were owned. A surprising
number, however, were restrained at the time of
the attack. In the case of pit bu II bites, 61 of 143
(42.7%) involved animals that were fenced,
chained, or inside prior to the incident. Twenty
cases (14%) involved pit bulls that escaped by
jumping fences or breaking chains immediately
before the attack. Of the 135 cases involving
other breeds, 36 (26.7%) involved restrained
animals, but only 1 (0.7%) broke restraint to in-
itiate the attack.
The press accounts support the fact that most
dog bites are unprovoked. Table 3 describes the
victims' interactions with dogs in the 163 reports
in which details were provided. The most
noteworthy distinction between pit bull attacks
and attacks involving other breeds is that 24.8%
of the former involved the victim coming to the
aid of an animal or person already injured by ttie
attacking animal. Thi~ occurred in only 11.3% of
the attacks by other breeds.
Our overview suggests that some pit bulls
present special problems. They account for a dis-
Randalilockwood and Kate Rindy
Table 2. Age of Victim in Nonfatal Dog Attacks
No. of victims No. of victims
Age of victim (dog=pit bull) (dog=other breed)
<5 yrs 22 (18.0%) 25 (21.2%)
5-9 yrs 24 (19.7%) 31 (26.3%)
10-14yrs 10 (8.2%) 17 (14.4%)
1"5"':1t) yrs 5 (4.1%) 3 (2.5%)
20-29 yrs 8(6.6%) 3 (2.5%)
30-39 yrs 7 (5.7%) 5 (4.2%)
40-49 yrs 4(3.3%) 4 (3.4%)
Adult, age
unspec. 42 (34.4%) 30 (25.4%)
Total 122 (100%) 118 (100%)
I
proportionate number of fatal attacks, although
these are few; and they are more likely than other
breeds to inflict serious injuries, to attack while
restrained or after breaking out of restraint, and to
attack adults, including their owners.
These generalizations seem to be supportable,
but we feel that we cannot use them to make
predictions about the behavior of an individual
animal. A dog's tendency to bite is a product of at
least five factors:
.The dog's genetic predisposition to be aggres-
sive
.The early socialization ofthe animal to people
.Its training for obedience or mistraining for
fighting
.The quality of care and supervision provided
by the owner
.The behavior of the victim
All of these factors interact. Genetic pre-
disposition is the on Iy factor directly relevant to
the issue of breed-specific restrictions. .Are pit
bulls as a group genetically uniform and predict-
ably aggressive enough to warrant special restric-
tions? Responsible breeders argue that they are
not. None of the 1986 fatal ities involved AKC- or
Are "Pit Bulls. Different?
UKC-registered animals, nor did press accounts of
nonfatal bites ever mention registration. Although
the nature and severity of pit bull attacks reflect
the effects of the dogs' selection for fighting, we
must recognize the variability in the animals that
we call pit bulls and in their owners.
The genetics of canine aggression are still
poorly understood, although the existence of
many breeds intentionally selected for aggression
under different circumstances clearly demon-
strates a strong genetic component to some
aspects of aggressive behavior. It is quite possible
that the t~rm pit bull encompasses a variety of
genetically diverse animals. The long history of
selection for gameness has produced a charac-
teristic fighting dog. The shorter history of breed-
ing for pet qualities has clearly overcome many
negative characteristics in responsibly bred
animals.
The remaining factors affecting dog attack are
all human variables related to the level of owner
responsibility and supervision. Many owners are
responsible people, well aware of the history of
pit bulls, and they attempt to correct problems of
aggression inherited from the past. Other owners
are ignorant of the breed. Most troublesome are
owners specifically seeking a "mean" dog. In their
hands, any dog is likely to become a menace, a
pit bull particularly so. The interest among less
responsible owners and breeders in overall
"meanness" has affected at least the last 10 to 20
generations of dogs; this fact may partly account
for the recent increase in the number of problem
animals. Finally, there continues to be an interest
in dog fighting. The dogs that prove to be too ag-
gressive to people to be acceptable for dog fight-
ing often wind up in the hands of owners seeking
a "mean" dog.
The common theme in virtually all of the fatal
and nonfatal attacks we reviewed was that the
owner had not taken appropriate steps to prevent
his or her animal from becoming a problem.
Simply placing an animal behind a fence or on a
chain is not sufficiently responsible behavior, par-
ticularly in the case of a breed or individual
animal inclined to attack others.
Problems of irresponsible ownership are not
unique to pit bulls, nor will they be in the future.
For this reason, we feel that effective animal con-
trol legislation must emphasize responsible and
ANTHROZOOS, Volume I, Number 1 7
Table 3. Victim Interaction with Dog (Where Known)
No. of cases No. of cases
Interaction (dog=pit bull) (dog=other breecl)
No direct
interactio. .59/101 (58.4%) 30/62 (48.4%)
Inactive,
walking 38 (37.6%) 15 (24.2%)
Run, bike, play 9 (8.9%) 7 (11.3%)
Other 12 (11.9%) 8 (12.9%)
Interacting with
dog 42/101 (41.6%) 32/62 (51.6%)
Feed, pet, play,
misc. friendly 8(7.9%) '7(27.4%)
Helping injured
animal 5 (5.0%) 3 (4.8%)
Helping injured
person 20 (19.8%) 4(6.5%)
Deliberate
provocation 5 (5.0%) 1 (1.6%)
Other 4(4.0%) 7 (11.3%)
humane ownership of sound animals as well as
responsible supervision of children and animals
when they interact. We believe that this can be
accomplished in a number ofways:
.Str€Agthen and enforce laws against dog fight-
ing to eliminate the "macho" image of this ac-
tivity.
.Introduce and enforce strong animal control
laws to identify problem animals and owners
before tragedy strikes. (Guidelines for such or-
dinances are available from the Humane
Society of the United States, 2100 L St. NW,
Washington, DC 20037.)
.Introduce programs to educate the public
about responsible ownership and the
problems of dog bite.
;;~
We feel that it is possible to protect the health
and safety of the public and at the same time
preserve the rights of pet owners. By placing
greater emphasis on responsible and humane
animal care, communities can go a long way
toward solving their current animal problems and
preventing new ones.
REFERENCES
Beck, A. M., H. loring, and R. lockwood. 1975. The
Ecology of Dog Bite Injury in St. louis, Missouri.
Public Health Reports 90(3):262-267.
lockwood, R. 1986. Vicious Dogs. Humane Society
News (Winter).
lockwood, R., and P. Miller. 1986. Pit Bull Report.
Humane Society of the United States, 2100 l St.
NW, Washington, DC 20037.
Matz, K. S. 1984. The Pit Bull: Fact and Fable.
Sacramento: De Mortmain Books.
Multani, P., and D. H. Clifford. 1985. Are Pit Bulls
Giving Good Dogs. a Bad Name? Community
Animal Control 4(3):16-17.
Pickney, l. E., and l. A. Kennedy. 1982. Traumatic
Deaths from Dog Attacks in the United States.
Pediatrics 39:193-196.
Semenic, C. 1984. The World of Fighting Dogs. Nep-
tune City, N.j.: T.F.H. Publications.
Wright, j. C. 1985. Severe Attacks by Dogs: Charac-
teristics of the Dogs, the Victims, and the Attack Set-
tings. Public Health Reports 700(7):55-61.
8 ANTHROlOOS, Volume I, Number 1 Randalilockwood and Kate Rindy
