7.1 Studies on animals’ understanding of ‘seeing’

While chimpanzees have been the animal most frequently focused on by researchers interested in Theory of Mind (ToM) in animals, the abilities of non-primates have also been studied. Donna Nissani (2004) discusses some studies she has carried out with elephants. In one set of studies, Nissani used adapted versions of studies on chimpanzees’ understanding of seeing (carried out by Povinelli and Eddy, 1996). Figure 14 shows some scenes from Nissani’s studies.

Figure 14: Nissani’s study: (a) The elephant’s natural begging pose; (b) The elephant making the ‘incorrect’ choice between food and a rock; (c) The elephant making an equivocal choice between a facing and an away-facing human (that is, it wasn't clear whether the elephant was choosing correctly or incorrectly); (d) The elephant making the ‘incorrect’ choice between the facing and away-facing human

Show description|Hide description

Figure 14 comprises four black and white photographs of an elephant interacting with one or more humans; each photograph depicts a scene from the studies described by Nissani (2004). In (a) the elephant faces a human, standing on the other side of a fence, with its trunk pointing towards the human in a ‘begging‘ pose. In (b) the elephant points its trunk towards a human, standing on the other side of a fence, who is facing it with both hands outstretched. In (c) the elephant points its trunk between two humans, on the other side of a fence, one of whom is facing toward the elephant and the other facing away. In (d) the elephant points its trunk towards a human, on the other side of a fence, who is lying on the ground facing away from the elephant; another human is lying nearby facing towards the elephant.

Figure 14: Nissani’s study: (a) The elephant’s natural begging pose; (b) The elephant making the ‘incorrect’ choice between food and a ...

As you can see from Figure 14, the elephant is required to make a choice, which will either be ‘correct’ (will gain them food) or ‘incorrect’ (will not gain them food). In order to get them used to the general procedure, they are first presented with a number of trials where they need to learn to choose between a piece of food and a rock, using their natural begging gesture (image (a)). They are then presented with the experimental trials in which they are required to choose between a human who can see them, and a human who cannot see them (images (c) and (d)).

Now watch this video: it shows the procedure used by Nissani in this series of studies, which looked for evidence of an understanding of ‘seeing’ in elephants.

Download this video clip.Video player: The procedure used by Nissani in her series of ‘seeing’ studies with elephants

Show transcript|Hide transcript

Transcript: The procedure used by Nissani in her series of ‘seeing’ studies with elephants

MAN:

[INAUDIBLE] It’s February 12, and it just, in the last portion of his pre-training, he’s learning how to beg. He had a hard time to still swing his trunk first before he does it, but now we are waiting, and it’s at least one second before we let him go. That’s the usual set-up, and the pre-training is usually done by either Donna [Nissani] or myself.

Again, he’s swinging it. That’s the way he always did it, but at least, now he’s able to settle.

DONNA NISSANI:

Next, an example of our session number nine.

This is session nine. This is Ade-Oui-Maang. It’s February 21, and we’re ready to go. This is Scarecrow Person. All righty.

OK. Now we have fans. They’re black fans. They’re not the old-fashioned ones that we had before, and he is going to the one that can’t see him.

OK. Third round here, we’re ready to go.

Much better this time. Here we go with the bamboo wall, and they’re ready to go. Go!

And that was an easy one to choose from.

Here we go. We’re going to do black fans again. And the wrong one again.

Ready. Here we go. Go!

Oh, that was much better. That was much better.

OK. Now this one is the guys each have on one of my two rings. One is an elephant ring. It’s got the elephant hair from Momo A, from Cheshire. The other one is my amethyst ring. And they’ve just switched rings.

And whoever has on the elephant, Momo A ring, the elephant tail ring, is the one that is the correct one. And we’re doing this as a way to tell that everything is kosher. Because there’s no way, no matter how smart the elephant is, it’s going to figure this out.

Here’s the rings, just so we can have an idea of what we’re talking about. The elephant ring obviously is the one on the top.

OK. Here comes the old scarecrow again. Not looking his best. He got attacked the other night, brutally, by one of the elephants trying to get the food out of his pocket. This one is, depending on which ear your hand is covering, is the one that the elephant is supposed to choose.

Now we’re going to do another one of the ear ones. These are controls to make sure that we’re not giving any unconscious signals to the elephants to help them decide which one to choose from. That’s not the end of session nine. There’s still some more to go, but they’re all repeats of what I’ve already shown, so we’re going to conclude the tape. All righty.

We may note in passing that the performance of our elephants in this task, 69 per cent correct, was roughly equal to that of our chimpanzees, 67 per cent. This means that both species, in about 30 per cent of the trials, begged from a person who could not see them. Here is just one brief example from Beauty, our star performer at the Detroit Zoo. Note that she is facing a person who can see her begging gesture, and goes out of her way to beg from a person who can’t see her at all.

This is Beauty again. This is the last one, and it’s buckets, and we’re ready to go whenever you are.

[Are elephants conscious?]

We went to the Detroit Zoo and the jungles of Burma, because we love elephants, and we’re curious about them. We were fortunate enough to gather some new information about their vision, olfaction, short-term memory, learning abilities, and behaviour. But we failed to come up with an unequivocal answer to the question, which was uppermost in our minds, and which constitutes, in our view, the fundamental question of both animal behaviour and comparative psychology. That is, are animals conscious? Is our Lila conscious?

Are elephants, for instance, to paraphrase Colin Allen, big zombies walking about without any awareness? Or are they, as Joyce Poole believes after decades of closely observing them in the wild, nearly as aware as human beings are? Can they think or solve problems in their head? Do they understand anything? Are they aware of their own selves and of the selves of others? Are they capable of empathy, compassion, or deliberate cruelty?

Once resolved, the answer to these questions will forever alter our view of elephants and, perhaps, also our conception of all other animals, ourselves included. When Moti and I started our adventures with elephants, we took their consciousness for granted, and hoped to come up with an unequivocal experimental evidence in its favour.

Instead, although the meagre evidence we have gathered so far is only suggestive and circumstantial, and although it does not rule out consciousness, to Moti this evidence appears more consistent with the odd view that elephants do not think. I, on the other hand, feel that there is enough of an anecdotal and field evidence to support the view that elephants are conscious.

In a few years, we hope we shall have a more definite answer. For now, we just want to end this film by saying that we are profoundly touched by our brief sojourn with elephants, and that we shall be forever grateful to the people and the institutions who made this sojourn possible.

Thank you, too, for joining us.

End transcript: The procedure used by Nissani in her series of ‘seeing’ studies with elephants

Download

The procedure used by Nissani in her series of ‘seeing’ studies with elephants

Nissani (2004) reported that elephants initially performed poorly, failing even to distinguish the presence of food − as determined by the food versus rock task (Figure 14 (b): the elephant gestures towards a human holding a rock, rather than one holding food). However, the elephants soon learned to beg correctly (e.g. approaching the human who was facing them, and thus could see them, rather than the human who was facing away) in all the conditions, on most of the trials (around 70%) in which they were tested. They performed in the facing and away-facing condition as well as the chimpanzees in a previous study by Povinelli and Eddy (1996), and outperformed the chimpanzees in other tests (such as, when a bucket was placed over the head of one experimenter, obscuring their vision, while the other had no bucket over their head). But despite the elephants performing better than the chimpanzees in Povinelli and Eddy’s study, they nevertheless begged incorrectly around one-third of the time. So, as Nissani points out in the video, findings from her studies remain equivocal − or inconclusive − on the issue of whether elephants do understand seeing.

Considering the issue of how methods and study designs used might influence the results obtained, Nissani identified a number of methodological issues with the original Povinelli and Eddy chimpanzee studies. She pointed out that these factors may have had an impact on the chimpanzees’ ability to demonstrate an understanding of perception in these studies. In particular, Nissani notes various aspects that made the procedure artificial and unnatural (such unnatural procedures are said to lack ecological validity). Among the key points that Nissani (2004) raises, and some of the adaptations she made to address these in her studies with elephants, are

• The required ‘begging’ (for food) gesture used was ‘natural’ for the elephants, but (most likely) not for the chimpanzees in the original Povinelli and Eddy studies.
• The elephants in Nissani’s studies had been raised in more natural conditions, whereas the chimpanzees in the original studies had mostly been born in captivity and subjected to experimentation for most of their lives.
• Povinelli and Eddy’s studies used young chimpanzees, between the ages of 5 and 6 years.
• The way in which the experimental and control trials were set up in the original procedure, and the nature of the rewards offered, made it quite plausible that the chimpanzees may have misunderstood what was required of them, and/or they may have lacked motivation to carry out the intended task (e.g. owing to a lack of sufficient rewards being offered in the test trials).

In her adapted studies, Nissani addressed many of the methodological issues identified. For example, she used adult elephants, and tried to create a more natural context by using tasks and behavioural responses that occur more spontaneously in a naturalistic setting. As mentioned earlier, the performance of Nissani’s elephants was better than that of the chimpanzees in the Povinelli and Eddy studies (although the elephants still made errors around one-third of the time).

Nissani also reported replicating the original Povinelli and Eddy study with chimpanzees. She made procedural variations in order to address many of the issues she identified, including using adult chimpanzees (in a zoo setting) who had not previously taken part in psychological experiments. She also made some more subtle adjustments: for example, altering the way the practice and test trials were organised, and the way in which the rewards were offered. Nissani found that the performance of the chimpanzees was improved (they performed above chance levels, which means they were not just choosing randomly), with findings comparable to those obtained with the elephants (Nissani, 2004).

These various procedural adaptions demonstrate the importance of paying careful attention to the methodological details of these types of laboratory studies, since relatively minor variations may influence the outcomes.

The issue of laboratory studies being artificial, and the extent to which they can be said to have ecological validity (in other words, to represent behaviours and abilities that occur in a naturalistic setting) is important to keep in mind when evaluating the results of comparative research. As you have seen, researchers such as Nissani have attempted to maximise the ecological validity of laboratory studies by using contexts and tasks that are more akin to what occurs in an animal’s natural environment. Some of these studies have found that animals do appear to show evidence of abilities, which previous studies using more artificial contexts suggested they did not possess, such as the understanding of seeing. Another approach, however, is to observe animals’ naturally occurring behaviours in fully naturalistic settings. You will consider this approach in the next section.