Is the dog's bowl half empty or half full?

Animals can exhibit pessimistic and optimistic traits. Michael North asks a psychologist exploring the consciousness of emotion what light this casts on the human condition.

The word "murderer" flashes up on the computer screen, followed shortly by the word "sunshine". I press a yellow or red key depending on whether I think the words unpleasant or pleasant. The only other requirement of the test is to press the buttons as quickly as I can.

I am in the office of Liz Paul, a psychologist at the Centre for Behavioural Biology at Bristol University's Veterinary School, wondering what the exercise is all about. Paul will explain shortly with contagious enthusiasm that she is testing the core aspects of my emotional functioning.

She lets me know after the test that the pleasant and unpleasant words were preceded by dark and light backgrounds respectively. I was not conscious of this, but apparently my brain could pick up the association.

"The idea is that even if you don't consciously recognise the association, you would have been speeding up slightly." Paul says a graph can be plotted to record this. She reassures me that even scientists will fail to spot the test's secret, whereas one of the school's secretaries might be quicker on the uptake.

The next stage of the test is to alter the background's dark or light shade slightly so that the unconscious mind gets confused about what is coming next and slows its responses. "What it is measuring is an unconscious interpretation of the colour of the screen. The next stage would be to find out whether unhappy or depressed people interpret intermediate grey screens as more likely to precede negative words than positive ones," she says.

Part of Paul's research has been aimed at devising reliable indicators of what she calls the "core affect system" - the response to pleasant and unpleasant stimuli that is at the heart of emotional behaviour.

But she is not just interested in humans. "Even a worm would withdraw from a punishing stimulus and approach a rewarding one," she says. Human beings just have a more complex range of emotions.

If Paul can devise reliable indicators of human emotional response, she hopes to be able to do the same for animals, which will aid monitoring of animal welfare.

In pursuit of this goal, Paul and colleagues have raised attention in the animal science community with their studies of "optimism" and "pessimism"

in rats. In humans, she explains, it has been observed that sugar is a mood enhancer - a trigger to the core affect system. "That's why when you are unhappy or get slightly depressed you will get this tendency to eat more sweet things. But if you become clinically depressed that liking/affect system breaks down and, for some time, you may not be able to feel pleasure at all."

In the experiment, Paul's research student Emma Harding induced negative mood states in rats by subtle disruption of their housing - say, by changing their bedding at unpredictable times. "The rats did not show signs of being depressed as such. But with subtle cognitive tests we found that they appeared to be less optimistic. The rats were trained that a sound of a particular pitch predicted a good event - the arrival of food - and that another sound, of a different pitch, predicted a bad event - no food and a short noise. They were then presented with sound of intermediate pitch to see whether they treated these ambiguous sounds as indicating a good or bad event. Rats from the unpredictable environment were less likely to treat these sounds as heralding the arrival of the good event than were those housed in the stable, predictable surroundings.

"The disrupted rats tended to expect something less good, so what we were suggesting was that they had a tendency to use that sort of marker to see things in a less positive light - an indicator of a negative affective state."

This is not conclusive proof that rats feel in the same way that we do.

Asked about the criticism of anthropomorphism, Paul replies that the behaviourist movement in science, which treated animals like emotionless black boxes, is thankfully over. "I would shy away from jumping to conclusions about what animals feel, such as whether they can feel complex emotions like guilt or regret. But humans have conscious experience, and we need to discuss whether and in what form animals have conscious experience as well. The question is not solved by ignoring it."

Paul says that one of the difficulties in interpreting animals' behaviour is that they have such different reactions from humans to certain kinds of stimuli. Chinchillas, for example, love to dust-bathe. "They dive into a tray of sand and they are so highly motivated. As humans we have no motivation to do anything like that."

The behaviour of hunted animals and physiology showing a similarly heightened emotional state is also difficult to interpret, she says. "Are they terrified or are they enjoying doing what comes naturally?"

But pessimistic rats, dust-bathing chinchillas and dogs that look guilty seem like a diversion compared with the key issue that concerns Paul - consciousness of emotions in humans and animals. Paul says that the affect system can be observed in most animals, but consciousness is a different matter. Only some primates, such as macaque monkeys and dolphins have shown they possess metacognition: they betray awareness of their knowledge of something in a cognitive task.

"But whether that means they are having a phenomenal experience of what it's like to know, or feeling what it's like to know, we can't be sure because it's the problem of other minds, that age-old philosophical chestnut: how can we be sure that anyone or anything except ourselves is conscious?"

Paul says it is a fair guess, however, that most primates have conscious knowledge. She now wants to find out whether consciousness of an emotional state, as opposed to an unconscious emotion, changes the way humans, and animals, function. "Hypothesise that you have a zombie who could not feel anything. But you know its heart thumped and its palms sweated and it ran away from scary things. But it never had a conscious experience. Would there be some loss of function in that zombie?"

The problem is how to measure a conscious emotion. Paul explains an experiment in which a person whose brain is monitored by an MRI scanner is shown images of neutral faces. But milliseconds before each face is shown an angry face is flashed at such speed that he or she will not initially be conscious of it. Yet a part of the brain that deals with emotion, the amygdala, "lights up".

She says some people will be more conscious of the subtle emotion being induced than others. "Can we compare those people who have emotions they notice consciously to those who are having the emotions done to them but for whom the emotions are not actually experienced consciously?"

The next step, says Paul, is to impose some form of information-processing on the participants to see if those conscious of the emotion process information differently from those unconscious of the emotion. "The ultimate goal is to find an information-processing task that works only where the affective state changes that information-processing if it is consciously felt by the person.

"If we could then show the same thing for animals, we would perhaps be one step further along the argument to say that they are using two levels of processing: they are doing the emotion and they are doing the extra emoting (that is to say, thinking about that emotion)."

Paul says that such a result would cause great debate in science and philosophy and prompt at least another ten years of research. She hopes that she may be able to develop quick non-linguistic tests of emotional states for humans who do not use language - such as premature babies - and for animals.

But she admits that she is a long way off the day when she can say whether a laboratory rat, say, is conscious of being depressed. She says science is a "slow, painful process" and she does not expect "Eureka moments" any time soon. "I would say it's an ongoing process, and people studying consciousness cannot expect to get to the end of the problem by the time they die, I'm afraid."

But she is determined to continue this fundamental research. "It's at the core of all our beings. It's the crux of everything."