Conclusions about the nature and magnitude of dolphin intelligence have not yet been reached. There are many different species of dolphin (see the cetacea article for a full list) and generalizations can be easily misapplied. There are only a select few species that live up to the ideal of dolphin intelligence.
For example, dolphins' cerebral cortex is 40% larger than human beings', with sulci and gyri ("wrinkles") of near equivalent complexity with a similarly developed frontal lobe (ibid) however, for example, "no patterns of cellular distribution, nuclear subdivision, or cellular morphology indicate specialization of the LC (coeruleus complex)" despite the large absolute brain size and unihemispheric sleep phenomenology of cetaceans. Moreover, it is generally agreed that the growth of the neocortex, both absolutely and relative to the rest of the brain, during human evolution, has been responsible for the evolution of intelligence, however defined. In most mammals the neocortex has six layers, and its different functional areas (vision, hearing, etc) are sharply differentiated. The cetacean neocortex, on the other hand, has only five layers, and there is little differentiation of outer layers according to function. The neocortex of the cetacean brain has a highly developed layer I and VI, which is a pattern that has been labeled "archaic" or phylogenetically primitive and superficially similar to that of hedgehogs. Therefore the evolutionary development of the cetacean brain has taken a different route than that of advanced terrestrial ones.
All sleeping mammals, including dolphins, go through a stage known as REM sleep. Unlike terrestrial mammals, dolphin brains contain a paralimbic lobe, which may possibly be used for sensory processing. The dolphin is a voluntary breather, even in sleep, with the result that veterinary anaesthesia of dolphins is impossible, as it would result in asphyxiation. Ridgway reports that EEGs show alternating hemispheric asymmetry in slow waves during sleep, with occasional sleep-like waves from both hemispheres. This result has been interpreted to mean that dolphins sleep only one hemisphere of their brain at a time, possibly to control their voluntary respiration system or to be vigilant for predators. This is also given as explanation for the large size of their brains.
Dolphin brain stem transmission time is faster than that normally found in humans, and is roughly equivalent to the speed found in rats. As echo-location is the dolphin's primary means of sensing its environment -- analogous to eyes in primates -- and since sound travels four and a half times faster in water than in air, scientists speculate that the faster brain stem transmission time, and perhaps the paralimbic lobe as well, support speedy processing of sound. The dolphin's dependence on speedy sound processing is evident in the structure of its brain: its neural area devoted to visual imaging is only about one-tenth that of the human brain, while the area devoted to acoustical imaging is about 10 times that of the human brain. (This is unsurprising: primate brains devote far more volume to visual processing than those of almost any other animal, and human brains more than other primates.) Sensory experiments suggest a high degree of cross-modal integration in the processing of shapes between echolocative and visual areas of the brain. Unlike the case of the human brain, the cetacean optic chiasm is completely crossed, and there is behavioral evidence for hemispheric dominance for vision.
Many scientists now tend to rank dolphins about the level of elephants in "intelligence" tests and say that dolphins haven't shown any unusual talent with problem solving compared with the other animals classed with very high intelligence.
Researching the behaviour of dolphins in the wild is a difficult task. However, several researchers have examined the social behaviour of dolphins and tried to extract an understanding of the level of communication between individuals, which in turn is interpreted as a measure of intelligence.
Resident orcas living in British Columbia, Canada, and Washington, United States live in extremely stable family groups. The basis of this social structure is the matriline, consisting of a mother and her offspring, who travel with her for life. Male orcas never leave their mother's pod, while female offspring may branch off to form their own matriline if they have many offspring of their own. Males have a particularly strong bond with their mother, and travel with them their entire lives, which can exceed 50 years. It is interesting behaviour, as it may seem that there would be no benefit from this except perhaps in hunting techniques, although they could join other groups to hunt. There are two interesting examples of this familial bond in males. Two male sons, identified as A38 and A39, constantly accompany their mother A30, despite that she needs no protection and they can all hunt by themselves, and rarely leave her side. Researchers have noted that if one son does wander off, one always remains with the mother. Another example are the brothers A32, A37 and A46, whose mother (A36) died. Instead of the family disbanding, the three brothers remain constantly together.
Relationships in the orca population can be discovered through their vocalizations. Matrilines who share a common ancestor from only a few generations back share mostly the same dialect, making up a pod. Pods who share some calls indicate a common ancestor from many generations back, and make up a clan. Interestingly, the orcas use these dialects to avoid in-breeding. They mate outside the clan, which is determined by the different vocalizations. On one occasion, an orca's mother and father were determined to be in the same clan, although in different pods.
In bottlenose dolphin studies by Wells in Sarasota, Florida, and Smolker in Shark Bay, Australia, females in a community are all linked either directly or through a mutual association in an overall social structure known as fission-fusion. Groups of the strongest association are known as "bands", and their composition can remain stable over years. There is some genetic evidence that band members may be related, but these bands are not necessarily limited to a single matrilineal line. There is no evidence that bands compete with each other. In the same research areas, as well as in Moray Firth, Scotland, males form strong associations of two to three individuals, with a coefficient of association between 70 and 100. These groups of males are known as "alliances", and members often display synchronous behaviours such as respiration, jumping, and breaching. Alliance composition is stable on the order of tens of years, and may provide a benefit for the acquisition of females for mating.
Starting with the dolphin named Malia, the methodology of the experiment was to choose a particular behavior exhibited by her each day and reward each display of that behavior throughout the day's session. At the start of each new day Malia would present the prior day's behavior, but only when a new behavior was exhibited was a reward given. All behaviors exhibited were, at least for a time, known behaviors of dolphins. At approximately the two week mark Malia apparently exhausted "normal" behaviors and began to repeat performances. This was met without reward.
According to Pryor the dolphin became almost despondent. However, at the sixteenth session following no novel behavior, the researchers were presented with a flip they had never seen before. This was reinforced. As related by Pryor, following the new display: "instead of offering that again she offered a tail swipe we'd never seen; we reinforced that. She began offering us all kinds of behavior that we hadn't seen in such a mad flurry that finally we could hardly choose what to throw fish at..."
The second test subject, Hou, took thirty-three sessions to reach the same stage. On each occasion the experiment was stopped when the variability of dolphin behaviour became too complex to make further positive reinforcement meaningful.
The same experiment was repeated with humans, and it took the volunteers about the same length of time to figure out what was being asked of them. After an initial period of frustration or anger, the humans realised they were being rewarded for novel behaviour. In dolphins this realisation produced excitement and more and more novel behaviours - in humans it mostly just produced relief.
Captive orcas have often displayed interesting responses when they get 'bored' with activities. For instance, when Dr. Paul Spong worked with the orca Skana, he researched her visual skills. However, after performing favourably in the 72 trials per day, Skana suddenly began consistently getting every answer wrong. Dr Spong concluded that a few fish were not enough motivation. He began playing music, which seemed to provide Skana with much more motivation.
At the Institute for Marine Mammal Studies in Mississippi, it has also been observed that the resident dolphins seem to show an awareness of the future. The dolphins are trained to keep their own tank clean by retrieving rubbish and bringing it to a keeper, to be rewarded with a fish. However, one dolphin, named Kelly, has apparently learned a way to get more fish, by hoarding the trash under a rock at the bottom of the pool and bringing it up one small piece at a time.
There is strong evidence that some specific whistles, called signature whistles, are used by dolphins to identify and/or call each other; dolphins have been observed emitting both other specimens' signature whistles, and their own. A unique signature whistle develops quite early in a dolphin's life, and it appears to be created in an imitation of the signature whistle of the dolphin's mother.
Xitco reported the ability of dolphins to passively eavesdrop on the active echolocative inspection of an object by another dolphin. Herman calls this effect the "acoustic flashlight" hypothesis, and may be related to findings by both Herman and Xitco on the comprehension of variations on the pointing gesture, including human pointing, dolphin postural pointing, and human gaze, in the sense of a redirection of another individual's attention, an ability which may require theory of mind.
The environment where dolphins live makes experiments much more expensive and complicated than for other species; additionally, the fact that cetaceans can emit and hear sounds (which are believed to be their main means of communication) in a range of frequencies much wider than humans' means that sophisticated equipment, which was scarcely available in the past, is needed to record and analyse them. For example, clicks can contain significant energy in frequencies greater than 110 kHz (for comparison, it is unusual for a human to be able to hear sounds above 20 kHz), requiring that equipment have a sampling rates of at least 220 kHz; MHz-capable hardware is often used.
In addition to the acoustic communication channel, the visual modality is also significant. The contrasting pigmentation of the body may be used, for example with "flashes" of the hypopigmented ventral area of some species, as can the production of bubble streams during signature whistling. Also, much of the synchronous and cooperative behaviors, as described in the Behavior section of this entry, as well as cooperative foraging methods, likely are managed at least in part through visual means.
While there is little evidence for dolphin language, experiments have shown that they can learn human sign language. Akeakamai, a bottlenose dolphin, was able to understand both individual words and basic sentences like "touch the frisbee with your tail and then jump over it" (Herman, Richards, & Wolz 1984). Dolphins have also exhibited the ability to understand the significance of the ordering of each set of tasks in one sentence.
The most widely used test for self-awareness in animals is the mirror test, developed by Gordon Gallup in the 1970s, in which a temporary dye is placed on an animal's body, and the animal is then presented with a mirror. Some scientists still disagree with these findings, arguing that the results of these tests are open to human interpretation and suceptible to Clever Hans effect. This test is far less definitive than when used for primates, because primates can touch the mark or the mirror, while dolphins cannot, making their alleged self-recognition behaviour less clear. Critics argue that behaviours that are said to identify self-awareness resemble existing social behaviours, and so researchers could be mislabelling social responses to another dolphin. The researchers counter-argue that the behaviours shown to evidence self-awareness are very different from normal responses to another dolphin, including paying significantly more attention to another dolphin than towards their mirror image. Dr. Gallup called the results "the most suggestive evidence to date" of mirror self-recognition in dolphins, but "not definitive" because he was not entirely clear that the dolphins were not interpreting the image in the mirror as another animal. Whereas apes can merely touch the mark on themselves with their fingers, dolphins show less definitive behavior of self-awareness, twisting and turning themselves to observe the mark.
As a further response to these criticisms, in 1995, Marten and Psarakos used television to test dolphin self-awareness . They showed dolphins real-time footage of themselves, recorded footage, and another dolphin. They concluded that their evidence suggested self-awareness rather than social behaviour. However, this study has not been repeated since then, the results remain thus uncorroborated.
Examples of cognitive abilities investigated in the dolphin include concept formation, sensory skills, and the use of mental representation of dolphins. Such research has been ongoing since the late 1970s, and includes the specific areas of: acoustic mimicry, behavioural mimicry (inter- and intra-specific), comprehension of novel sequences in an artificial language (including non-finite state grammars as well as novel anomalous sequences), memory, monitoring of self-behaviours (including reporting on these, as well as avoiding or repeating them), reporting on the presence and absence of objects, object categorization, discrimination and matching (identity matching to sample, delayed matching to sample, arbitrary matching to sample, matching across echolocation and vision, reporting that no identity match exists, etc.), synchronous creative behaviours between two animals, comprehension of symbols for various body parts, comprehension of the pointing gesture and gaze (as made by dolphins or humans), problem solving, echolocative eavesdropping, and more. Some researchers include Louis Herman, Mark Xitco, John Gory, Stan Kuczaj, Adam Pack, and many others.
While these are largely laboratory studies, field studies relating to dolphin and cetacean cognition are also relevant to the issue of intelligence, including those proposing tool use, culture, fission-fusion social structure (including tracking alliances and other cooperative behaviour), acoustic behaviour (bottlenose dolphin signature whistles, sperm whale clicks, orca pod vocalizations), foraging methods (partial beaching, cooperation with human fishermen, herding fish into a ball, etc.). See: Richard Connor, Hal Whitehead, Peter Tyack, Janet Mann, Randall Wells, Kenneth Norris, B. Wursig, John Ford, Louis Herman, Diana Reiss, Lori Marino, Sam Ridgway, Paul Nachtigall, Eduardo Mercado, Denise Herzing, Whitlow Au.
In contrast to the primates, cetaceans are particularly far-removed from humans in evolutionary time. Therefore, cognitive abilities generally cannot be claimed to derive from a common ancestor, whereas such claims are sometimes made by researchers studying primate cognition. Though cetaceans and humans (in common with all mammals) certainly had a common ancestor in the distant past, it was almost certainly of distinctly inferior cognitive abilities compared to its modern descendants.