PICTURE a lab rat imprisoned for life in a cage the size of a shoe box. Or a
fox racing to escape from a pack of hounds in full cry. Are these animals
suffering? To our eyes, it may seem obvious that they are having a pretty rough
time. But two British zoologists claim it is not that simple. To discover
whether an animal is likely to be suffering, they say, you need to ask if
evolution has designed it to deal with such conditions. If it has, then the risk
of suffering may not be nearly as great as we might imagine. And if there is
suffering, at least where farms and labs are concerned, it may be possible to
alter the animal’s living conditions accordingly.
To unravel the puzzle of when an animal might be suffering, we need to
understand how it has evolved to “value” its own survival and wellbeing, argue
Chris Barnard, professor of animal behaviour at the University of Nottingham,
and Jane Hurst, a behavioural ecologist also at Nottingham, who studies wild
rodents. In a controversial paper published last November in the academic
journal Animal Welfare, Barnard and Hurst argue that animals cannot be
assumed to suffer—no matter how nasty the situation may seem to
us—if in fact natural selection has “designed” them to function in that
particular situation. We cannot automatically assume that a rabbit fleeing a fox
is suffering, they claim, because it is operating in familiar territory,
evolutionarily speaking. Foxes, by contrast, are predators without natural
enemies, so they may not have developed the rabbit’s capacity to cope with the
experience of being hunted.
“We say animals may not suffer when running away from predators,” says Hurst,
“because it is a contingency that natural selection has adapted them to deal
with.” An animal may experience the subjective state of “fear”, but that just
tells it to run, just as “hunger” tells it to eat. Fear and hunger only lead to
suffering when the animal’s adaptive responses fail to assuage them, Barnard and
Hurst claim. Suffering is triggered when the world frustrates an animal’s
adaptive drives.
Advertisement
This may sound like a subtle distinction, but the Nottingham researchers see
it as a crucial one that has been neglected by animal welfare campaigners weak
on evolutionary theory.
Researchers have a range of techniques for assessing whether an animal is
suffering. In addition to behavioural assessments, they include physiological
measurements such as blood pressure, stress hormone levels and the strength of
the immune system. But even if you use behaviours which seem to reveal what the
animal is feeling, can you be sure that you will be able to interpret these
subjective feelings correctly? Surely, you may well ask, anything that the
animal experiences as “bad” or aversive leads to suffering?
Questions of pain and suffering have been debated across the centuries and
across all manner of disciplines. Stephen Clark, professor of philosophy at the
University of Liverpool, points out that in classical Greece, Epicureans thought
that any pain was “a bad thing” that caused suffering in all animals. Stoics, on
the other hand, “thought that pain only mattered as a sign of some impediment to
natural functioning”. They reasoned that animals were concerned only with
maintaining their existence and their kind. “The Stoics pointed out that animals
didn’t shrink from doing what might seem to be painful in pursuit of their
natural ends.”
But it does not follow, Clark concludes, that animals never feel or object to
pain, nor that their pain is always appropriate. And distinguished scientists,
such as Oxford zoologist Marian Stamp Dawkins, have suggested that “bad” or
painful subjective experiences we call suffering evolved to alert animals to a
state of emergency and trigger evasive action. Whether or not these actions are
frustrated, they argue, those responsible for animal welfare cannot ignore them
(“The emotional chicken”, Â鶹´«Ă˝, 22 January 1994, p 28).
But Hurst and Barnard downplay the role of subjective feelings. “Many `bad’
feelings will simply reflect an internal gauge which animals use to regulate
damage within adaptive bounds,” says Barnard. The conventional approach to
animal welfare which gives a high priority to individual
self-preservation—staying alive—simply reflects our own priorities
as a long-lived species, he says.
“People’s conception of what is good for animals tends to be the cat on a mat
sort of thing, rather than seeing things from the animal’s point of view,”
agrees Robin Dunbar, professor of psychology at the University of Liverpool, who
sympathises with Barnard and Hurst’s approach. People have an anthropomorphic
concept of suffering and fail to “see things from the animal’s point of view”,
he says.
Design for life
Barnard uses an analogy to explain his criticisms of conventional approaches
to animal welfare. It is like having a Ferrari, polishing it and maintaining it
meticulously for years, yet never taking it out of the garage. What we should be
trying to achieve, he says, is a well-functioning Ferrari on the road. The
animal performing its natural behaviours is like the finely tuned motor car in
action. You will run into problems only if you treat the car (or animal) in a
way that its designers (evolution) didn’t intend. “Animals suffer when they are
being forced to perform outside their design criteria,” he says.
It goes without saying, explains Barnard, that many experimental procedures
performed on animals infringe their design criteria and prevent them from acting
in line with inbuilt “decision rules”—evolutionary rules of thumb guiding
their behaviour. “An animal riding a hot plate to test analgesia, an animal
restrained and injected—in these conditions we are self-evidently
inflicting an imposition on the animal that it is not designed to cope with,” he
says. “We know that already.” What welfare campaigners and experimenters need to
consider, argues Barnard, is whether less clear-cut issues such as day-to-day
husbandry—feeding regimes, laboratory housing and the
like—could be addressed to better suit the design criteria of individual
species.
Life behind bars
Hurst points to a telling example. In the laboratory, she says, female rats
and mice make many more attempts than do males to escape from their cages when
housed singly or in same-sex groups. The males seem much more at ease on their
own or with other males. Hurst explains these observations by looking to their
evolutionary design criteria. In the wild, she says, male rats and mice compete
with other males as they try to establish dominance and to set up and defend
territories. In a sense, males expect either to interact with other males, or
exclude them. But in nature, a mature female mouse or rat would never live alone
or only with other females. She would move into a male’s territory to mate, and
would then aggressively fend off other females who might threaten her offspring.
Males expect direct competition with each other, but females compete for
space.
The dullness and predictability of life in captivity can also frustrate an
animal’s design rules, say the zoologists. Carefully planned “environmental
enrichment” can improve matters by offering animals suitable decision-making
challenges. Dispersing food around cages, for example, simulates the patchiness
of natural resources far better than a single feeding station.
But there again, anyone wishing to promote animal welfare must first have
their evolutionary spectacles firmly in place. Increasing the apparent
naturalness of an animal’s surroundings may not always be beneficial. “You have
to understand what changing the environment means to the animals, and how to
interpret their response to the change,” Barnard argues. For instance, if you
put shelves and nest boxes in a bare cage to give mice exercise and sleeping
places, the frequency of aggressive encounters goes up and their immune
competence falls. The boxes and shelves give dominant mice something to defend,
and trigger their aggression. In a cage, their hapless subordinates have nowhere
to run.
Immune competence is commonly regarded as a sound, objective physiological
measure of stress in studies of animal and human welfare: the greater the fall
in immune competence, the greater the stress. But the situation is more
complicated than it may seem. Taken out of their social and evolutionary
context, isolated measures of physical cost, such as stress responses and immune
depression, do not necessarily indicate that something unwelcome is happening to
the animal.
A male mouse may even “choose” to do something that damages his long-term
survival prospects, if by so doing he can boost his reproductive chances,
Barnard and Hurst argue. But according to conventional measurements, his actions
would make it appear that he was suffering. In recent experiments, Barnard and
Hurst found that a high-ranking mouse in a group of captive males boosts his
testosterone levels to maintain a mean, competitive edge. But this rise in
testosterone and other hormones conventionally linked to “stress” actually
lowers the efficiency of his immune system and increases the risk of succumbing
to infection. Barnard claims that the animal has, in an evolutionary sense,
“chosen” to pay the physiological cost in this case.
“We humans think the maintenance of the individual is the priority, because
we live a long time and have a long period of parental care,” says Hurst. Mice,
however, may make very different adaptive trade-offs, exchanging a brief and
uncertain future for the greater certainty of here-and-now reproduction.
In Barnard’s view, the dominant male’s stress and increased susceptibility to
disease is an example of adaptive decision making. Even so, dominant individuals
may exercise a degree of damage limitation. The Nottingham research has found
that a male mouse produces less immunosuppressive testosterone if its immune
system is already at a low ebb. In other words, mice will sacrifice some
competitive oomph in the interests of health. “Looking for these kinds of
damage-limitation strategies may give us a better idea of how animals assess the
impact on them than simple measures of cost,” says Barnard.
The plight of subordinates is altogether different. Caged rats can cope with
a certain amount of stylised aggression, signalled by postural cues for
dominance and submission. This helps them to avoid out-and-out fights and
promotes peaceful coexistence. But they are not designed to cope with sustained
attacks, and eventually a nastier sort of aggression between caged rats can
emerge, which may cause suffering. This happens when one rat is not willing for
another to remain in the same cage, and takes to chasing it. The
get-the-hell-out-of-here rule that the victim would apply in the wild is
frustrated in the confines of a laboratory cage, so the aggressor attacks, often
by repeatedly pulling at the victim’s fur with its teeth. In this situation, the
victim can quickly become ill from stress alone, let alone physical injury, and
would rightly be judged to be suffering, say Hurst and Barnard.
Victims of this type of aggression spend a lot of time climbing, twirling and
chewing at the cage bars. Such repetitive but useless behaviour is often written
off as a dysfunctional and neurotic “stereotypy” caused by some ill-defined
“frustration”. Hurst, however, sees this behaviour as a thwarted attempt to
escape and a sign of suffering.
Born to flee
So has Barnard and Hurst’s theory rattled the cages of fellow academics? So
far, their paper has met with a somewhat mixed response in the world of animal
welfare.
“It’s a thought-provoking view,” says Mike Mendl, who researches into the
welfare of pigs at Bristol University’s school of veterinary sciences. His
misgivings centre on the difficulty of defining a “decision rule” or “design
criterion” for a particular animal. He fears that Barnard and Hurst’s approach
risks degenerating into academic speculation about what the animal is or is not
designed to cope with. “It seems to me that you might be able to argue back and
forth about whether a particular situation is overtaxing a species in light of
its evolutionary history,” he says.
Veterinarians may find the argument even harder to swallow. Caroline Manser,
a vet in Cambridge who specialises in animal welfare issues, argues that the
Nottingham zoologists have no evidence that a mouse fleeing a cat, for example,
is not suffering. In the circumstances, says Manser, “I would rather give the
benefit of the doubt to the mouse”. She wonders whether animals wounded as they
flee from predators or in competitive disputes can rightly be said to be free
from suffering, simply because they have made what seems to be the “right”
evolutionary decision.
Barnard and Hurst reply that there is a good case for invoking suffering if
the animal’s decision rule—”get away from the predator”—has
manifestly failed. But they maintain that, in some circumstances, an animal
might accept physical damage and reduced survival prospects as the price to be
paid for acquiring mates, food, territories or whatever else is the object of
competition. “The paradoxical implication,” says Barnard, “is that cost itself
may not have welfare significance, but frustrating the animal’s ability to incur
a cost may have.”
In Manser’s view, Barnard and Hurst seem to imply that suffering is
unimportant in short-lived species. “But suffering is a subjective experience,
of concern to the individual sufferer,” she argues. “Small rodents can certainly
behave as though they are experiencing pain or distress, so how can we simply
dismiss their apparent suffering?” But Barnard and Hurst say that the argument
is not that short-lived species are less likely to suffer, but that they are
designed to make and cope with self-sacrificing trade-offs that are very
different from those of long-lived species.
Feeling hurt
A major problem remains. If welfare means what Barnard and Hurst suggest,
doesn’t it apply equally to plants and bacteria as well as what we think of as
“sentient” animals. After all, plants and “lower” organisms have also evolved
with “design criteria” that can be infringed. In fact, says Don Broom, professor
of animal welfare at the University of Cambridge, almost everyone would agree
that the word “welfare” applies only to animals. Animals are unlike other
organisms in the ways they try to cope with adversity—running away,
seeking out better conditions or modifying their environment—and in having
feelings. “Feelings may be difficult to measure,” says Broom, “but we should try
to obtain precise information about them.” He agrees with Dawkins’s stance, that
feelings are something scientists can study. They “have evolved because they are
useful to animals in their attempts to cope with their environment,” says
Broom.
“Agreed,” says Barnard. The trouble is, “this stance spawns arguments about
what does and what does not have `feelings'”. He believes that many researchers
waste time assessing suffering by drawing “subjective lines between organisms”
based on whether they have nervous systems and their complexity. A far more
objective way to assess an organism’s potential to suffer, he says, is to look
at its ability to control its circumstances and judge whether this ability is
being frustrated. “Understanding the animal’s adaptive decision rules gives us a
direct window onto its coping mechanisms.” This, he argues, is the best way we
can hope to find to measure suffering.
- “Animal welfare defined in terms of attempts to cope with
the environment” by D. M. Broom, Acta Agriculturae Scandinavica Section A,
Animal Science, vol 27, p 22 (1996) - “Welfare by design: the natural selection of welfare criteria” by C. J.
Barnard and J. L. Hurst, Animal Welfare, vol 5, p 405 (1996).
