In my future career I want to work with dolphins and find out whether their metacognition and consciousness is similar to that of humans. In their study in 2003, Smith, Shields and Washburn tested how well dolphins and humans were able to differentiate between pitch heights. In order to do this, they played a 2100 Hz reference tone, and then several others that were lower or higher. Both humans and animals then had to decide whether the heard tone was lower, higher, or the same as the reference tone.
In a first experiment, the dolphin showed very good answers if the tone was far away from the reference, but close to 2100 Hz his answers were around chance. Therefore, in a second experiment, the “uncertain” response-paddle was introduced: Does the dolphin know that his answers are close to chance? The dolphins (and humans) again had to say whether the tones were higher or lower than the reference, but they could also choose a third answer: “Uncertainty”. Dolphins chose the “uncertain” response very frequently if the played tone was close to the reference tone (e.g. a 2085 Hz tone). We could interpret that they were able to realize when the tones were very similar (2085-2100 Hz), and, more importantly, that they noticed that they had difficulties with these decisions. This would show that dolphins have at least some kind of metacognition, as they know when their answers aren’t too sure. This is supported by the fact that they showed behavioral signs of uncertainty, as they hesitated to give an answer if the asked tone was close to the reference tone.
Other possible explanations would be that the dolphin does not show metacognition, but that e.g. the dolphin trainer sends unconscious cues, therefore enabling an “uncertain” reponse when the trials are difficult, or that the animal responds with “uncertain” to avoid aversive, error producing stimuli.
Whatever might be the actual answer to this question, it remains a fact that in order to conduct experiments like these, the use of applied behavioral analysis is necessary. The dolphin first has to learn to use the response paddles, and this can be done by either reinforcing or punishing him. An example for a good way of training a dolphin with the help of reinforcement is clicker training (Kaplan, Oudeyer, Kubinyi, & Mikosi, 2002). If you press the clicker, it emits a sharp sound, and this sound has to be associated with a primary reinforcer (e.g. fish, a toy, …). Thanks to this association, the clicker is then seen as a secondary reinforcer. Therefore, if you click it, the animal will know that the reward will come soon. After charging the clicker (associating its noise with the primary reward), the behavior is introduced: The movements the dolphin is supposed to learn are shaped, by rewarding a rather vague behavior at the beginning, and then only clicking for more special behaviors over time. Finally, the dolphin will have learned how to respond with the response paddles, and the actual experiment can be started.
Kaplan, F., Oudeyer, P., Kubinyi, E., & Miklosi, A. (2002). Robotic clicker training. Robotics and Autonomous Systems, 38, 197-206.
Smith, J. D., Shields, W. E., & Washburn, D. A. (2003). The comparative psychology of uncertainty monitoring and metacognition. Behavioral and Brain Sciences, 26, 317-373.