I think it’s time to put a stop to this nonsense and shoot down a few flying
pigs. Their reputation for being tractable, easy to train, intelligent and cute
is far from deserved. Pigs are simply not all they are cracked up to be.

As is widely known, the Russian physiologist Ivan Pavlov used dogs in his
pioneering experiments on the conditioned reflex. He placed a dog in a harness
on a stand and presented it simultaneously with food and the ring of a bell. The
dog salivated as a reflex response to food. After a while, the dog came to
associate food with the bell and it would salivate in response to the bell
alone. Pavlov tried using pigs in his experiments, but they proved to be
difficult subjects.

When he tried to produce the same response from a pig, Pavlov got nowhere. As
soon as the pig was lifted onto a stand, it squealed at the top of its voice and
all work in the laboratory was impossible. Nearby colleagues complained, and
whatever Pavlov did to try and soothe the animal, it would not cooperate. He
spent a month working with it and, gaining no results, he finally gave up. The
pig was the most nervous of animals, so he concluded that all pigs are
hysterical. So much for the pig’s reputation for tractability.

What about pigs being easy to train, then? Keller and Marian Breland
attempted to do this in the 1950s and found the truth to be otherwise. The
Brelands were students of the great American psychologist B. F. Skinner. They
gave up promising research careers to train animals for shopping centre displays
and television commercials. Eventually they conditioned a veritable Noah’s Ark
of some 6000 individual animals from 38 species, including reindeer, cockatoos,
racoons and whales.

They succeeded in conditioning a pig to pick up large wooden coins, carry
them several feet and deposit them, of course, in a piggy bank. Every time the
pig deposited four coins it received a food reward. At first it eagerly picked
up the coins, ran to the bank and deposited them, performing the actions rapidly
and neatly. Over a period of weeks, however, the process became slower and
slower. The pig might run over for the coin, but on the way back drop it, root
it, drop it again, root it along the floor, pick it up, toss it in the air, drop
it, root it some more, and so on. Increasingly its behaviour deteriorated until
it was taking about 10 minutes to transport four coins about six feet, and the
pig was earning so few rewards that it was not getting enough to eat in a
day.

Ardent porcophiles will doubtless maintain that this was an isolated
occurrence, but this behaviour developed repeatedly in successive pigs. The
Brelands interpreted their observations as a complete and utter failure of
conditioning theory. The conditioned response had been displaced by instinctive
foraging behaviour. However, I would interpret the Brelands’ observations as the
utter failure of the pig to accept training.

Which brings us to intelligence. Surely pigs are smart? Psychologists Dudley
Klopfer and Frank Wesley of Washington State University touched on pig
intelligence when they tried to condition pigs to discriminate between different
visual stimuli. They trained four sows to respond to stimulus cards that
differed in form, colour and size, and presented them randomly either to the
left or the right. In 500 trials of 24 problems, the pigs’ scores never improved
over what would be expected if the cards were chosen by chance—not exactly
genius material.

So, that leaves us with the epithet “cute”. Cuteness is perhaps something
that resides in the eye of the beholder and so may be impossible to measure in
the laboratory. However, in my work I condition pigs to lift a lever repeatedly
for a food reward. I gradually increase the number of lifts required until the
pig stops responding, and use this point as a measure of their hunger. My
current record-holder is a sow which lifted the lever 3282 times an hour before
giving up. That is some persistence, some hunger.

From such a close working relationship with pigs I can only conclude that
pigs are, well, ravenous. Pigs are pigs. How could anyone love an animal which
is that hungry? And they smell, to boot. Not much cuteness there.

I suppose most people think that a sheep is, well, a sheep. An animal that blindly follows other animals, whatever the outcome, with no capacity for independent thought. It is true that sheep have a strongly developed flocking and following instinct that often overrides other, perhaps more rational, behaviour.

For example, a 19th-century textbook on sheep management describes how a flock of sheep was being driven through the streets of Liverpool “when a man with a broomstick tried to stop them running down a street. He stooped, brandishing the broomstick, and the entire flock, following their leaders, jumped over him in quick succession.” Another passage describes one sheep in a flock of 800 jumping over a fence in fear of a dog and “falling down a great descent”. The other sheep in the mob followed and 123 were killed. How smart is that?

Early scientific studies of learning appeared to confirm that the sheep was rather dim. In the 1920s Howard Liddell, working at Cornell University’s Behaviour Farm in Ithaca, New York, compared the learning abilities of various farm animals using a classical conditioning technique. When he applied a brief, mild electric shock to the sheep’s foreleg the animal responded with an unconditioned response – it raised its leg. After 5 to 10 pairings of the sound of a metronome with the shock the sheep reacted to the metronome alone by lifting its leg. Similar studies of the number of combinations of signal and shock necessary to establish a conditioned motor reflex showed that one trial was sufficient for the pig, 3 to 5 for the dog, 7 to 9 for goats, and 14 to 24 for the rabbit.

Pearl Gardner, also working at Cornell University but in the 1930s, used a different technique to determine the sheep’s ability to learn. Her experiment involved getting sheep to open one of three wooden boxes by knocking off its lid. Gardner draped a black cotton cloth over the box that contained food. She found that sheep learnt rather slowly, partly because of their fear of the cloth signal. Sheep urinated, stamped their feet, bunted the cloth, or avoided it altogether. But eventually all sheep learnt to go to the box with the cloth signal. In a fascinating extension of this study Gardner compared the learning abilities of other species when confronted with the same three-box problem. The average error per trial for horses was 0.37, compared with 0.44 for cattle and 0.55 for sheep. However, sheep did rather better than people with IQs of less than 50, who averaged 0.61 errors per trial for a sweet reward.

So, the general conclusion from these early studies was that the sheep are relatively slow learners – in about fourth place on the farm animal intelligence scale. However, recent discoveries have turned this conclusion on its head.

Keith Kendrick of the Agricultural and Food Research Council’s Institute of Physiology and Genetics Research, Babraham, Cambridge, recorded the electrical output from single cells in the sheep’s brain (“Through a sheep’s eye”, New ĐÓ°ÉÔ­´´, 12 May 1990). Cells in the hypothalamus appear to specialise in the visual recognition of foods and respond to the sight of known palatable foods. The more an animal likes a particular food, the more these cells respond to the sight of it. If a sheep encounters a novel food, the cells do not respond. But if a sheep eats the food just once and likes it, the cells will respond the next time it sees the food, even if the sheep has not seen it for a month or more. This is most remarkable evidence of single-trial learning at the cellular level in the brain of the sheep.

And Kendrick is not alone in demonstrating single-trial learning in the sheep. Fred Provenza of the range science department at Utah State University has spent many years studying conditioned food aversions. Because food is so crucial for survival and sheep spend up to 12 hours a day grazing, they must be able to learn about and remember the diverse array of foods they find in their environment. Provenza found that lambs could learn to discriminate between a familiar safe food (alfalfa pellets) and a novel harmful food in a single trial. The harmful food was a palatable feed, corn that had been treated with lithium chloride (LiCl), a gastrointestinal poison.

Provenza went further and suggested that sheep must be able to learn to avoid novel foods given delays of several hours between ingesting the food and the “feedback” they get afterwards, and they must also be able to do this while consuming a meal of a variety of foods. Together with E. A. Burritt, he tested this proposition by giving the lambs LiCl at 0, 2 and 4 hours after they had consumed a novel food. All lambs reduced their intake of the novel feed after the single dose of LiCl, regardless of when it was administered. When mixed meals were tested by offering the animals a novel food for 5 minutes followed by four familiar foods for 20 minutes and then LiCl, the lambs refused to eat the novel food for the next three days.

Sheep are also capable of learning by observation. Justin Lynch of the CSIRO, Australia’s national research organisation, has demonstrated the crucial importance of the social context of learning. Working at the Division of Animal Production in Armidale, New South Wales, Lynch found that animals accepted a new feed more quickly if they saw other sheep feeding during their test. Sheep nominally took more than 13 days of 15 minutes’ exposure to wheat grain before eating even 10 grams of it. In contrast, lambs given wheat for 15 minutes a day with their wheat-eating mothers on five consecutive days subsequently ate an average of 120 grams a day when offered wheat after weaning. This knowledge was retained by the sheep for life, and was still used after years of encountering no wheat.

The ultimate accolade for animal intelligence is to be classified as a tool user, along with chimpanzees, gorillas and Californian sea otters. According to S. F. Bates writing from Cheshire (Letters, New ĐÓ°ÉÔ­´´, 15 March 1979) sheep can use tools. He reported that Gritstone sheep in Lyme Park, Cheshire, had scratched away snow to reveal hidden grass using pieces of wood that they held in their incisor teeth. It seems that they used the technique where the snow had thawed and refrozen, becoming hard and icy.

So the modern view of the sheep is that it is a very smart cookie indeed, capable of single-trial learning, delayed learning, observational learning and using tools. Keep this in mind next time you are in Liverpool and are confronted by an angry mob, armed only with a broomstick.