Evidence of self-awareness in the bottlenose dolphin
(Tursiops truncatus)

by Kenneth Marten and Suchi Psarakos

Dr. Ken Marten e-mail: porpoise@lava.net

Reprinted with permission from Cambridge University Press. All rights reserved.
Self-awareness in Animals and Humans: Developmental Perspectives.
Edited by Sue Taylor Parker, Robert W. Mitchell, and Maria L. Boccia.
Chapter 24, pp. 361-379. New York: Cambridge University Press, 1995.

Dolphins are big-brained, socially sophisticated mammals, socially and cognitively comparable to monkeys and apes in memory capacity, language comprehension, and other cognitive abilities (Herman, 1980). Their remarkable capacities suggest that dolphins, like apes, may come to recognize themselves in mirrors (e.g., Anderson, 1984; Gallup, 1970, 1982; Lethmate & Dücker, 1973; Suarez and Gallup, 1981). This chapter reports on the results of a series of studies utilizing a mirror to assess whether dolphins recognize contingent representations of themselves or use the mirror to examine an area of the body not otherwise visible that has been marked with a highly salient substance. We adapted the mirror mark test (Gallup, 1970) for use with the bottlenose dolphin. We employed several control conditions, including mirror without mark, no mirror and no mark, and first encounter between unfamiliar dolphins through a barrier. We also devised and conducted several new tests for self-recognition, tailored for dolphins rather than primates. These tests, which utilize self-view television and video playback, are summarized here. This chapter focuses on interpreting mirror-directed behavior (both marked and unmarked) by comparing it to the control data. We address the central question of whether the dolphins' mirror-directed behavior is social or self-examination. We also discuss the role of environmental, social, and individual influences on the test results.


Subjects and setting

Five dolphins, 6-14 years old, living at Sea Life Park, served as subjects of this study (Table 24.1). The groups of dolphins changed throughout the course of the research. The dolphins dwell in a two-tank complex (Figure 24.1). The laboratory has five underwater windows looking into the large tank. One window is a 1.2-m-diameter circle, which can be made into a large one-way mirror. Another window is 0.6 m x 0.6 m and houses a 20-in. Sony Trinitron color television that the dolphins can watch. Data were collected by videotaping the animals through the window or a one-way mirror with a Minolta S-VHS Series V-200 video camera. Notes on behavior were also recorded on an ongoing basis during testing. No food rewards were provided.

Table 24.1. Mark test subjects and methods

                                      Mirror  (b)
                                      expos. Mark
                   (a)                before test
                  Origin       Mirror first  mirror Body   (c)
         Age      and          mark   test   size   part  Marked
Subject (yr) Sex  subspecies   test   (hr)   (m)    markd with

Keola    14  male CB,truncatus  1     <1     1.2    side    ZO
                                2    225     1.2    fluke   ZO
                                3    260     1.2    rost./ Ich/GV
Hot Rod   6  male CB,gilli      1     <1     1.2    side    ZO
                                2     225    1.2    pec fin ZO
                                3     260    1.2    melon/  ZO/GV
Okoa     13 female WC,gilli     1     270    1.2    side    ZO
Laukani  14 female WC,gilli     1      80    0.8x0.7 side   ZO
                                2      84    1.2    melon   ZO
Itsi     13 female CB,truncatus 1      75    0.8x0.7 behind GV
Bitsi                                                pec fin
                                2      78    1.2    side    ZO
                                3      86    1.2    side    ZO

(a) CB, captive born; WC, wild caught.
(b) Single dimension is diameter.
(c) ZO, zinc oxite; GV, gentian violet; Ich, ichthammol.


The mirror was a reflective aluminum-coated polyester Mylar film made by DuPont, which we applied to the window with soap and water. Viewed from the tank, it makes a perfect mirror, yet because of its one-way properties we were able to videotape the dolphins' mark tests through the same mirror from inside the lab. In some tests we covered the entire 1.2-m-diameter circular window with the reflective Mylar, while in others, we used a smaller 0.8 m x 0.7 m rectangle of Mylar in the same window (Table 24.1).


Mirror mark tests. All dolphins had some mirror exposure prior to their first mark test. The dolphins were not isolated during their mark tests, and sometimes more than one animal was marked at a time. All but one of the dolphins had multiple mark tests. Table 24.1 presents information regarding the hours of mirror exposure prior to each mark test, the body part marked, and other relevant information.

[Figure 24.1:50K GIF Schematic]

Figure 24.1. Two-tank complex (top view) and underwater viewing laboratory, Sea Life Park, Hawaii. Videotaping of the body mark tests was done through a one-way mirror on the large, flat window. A camera next to a 20-in. television in the small square window provided the frontal self-view on the television (mirror mode) for the television tests summarized in the text.

Unless noted otherwise, all animals were marked on their sides with approximately 3/4 oz of zinc oxide (Figure 24.2). Zinc oxide is a tactile as well as visual stimulus. (We could not find an appropriate stain without tactile stimulus.) Other stains used were gentian violet, a purple topical antiseptic, and ichthammol, a black antiseptic ointment. One subject, Itsi Bitsi, was sham marked (with the mirror present) with Vaseline six times prior to her first zinc oxide mark test. Mark locations were selected for their likelihood of being visible only with the aid of a mirror. Bottlenose dolphin eyes point laterally and cant forward and down. This creates a binocular field in front of and below the animal, with a blind area behind and above, where the marks were located.

[Figure 24.2: 83K GIF (11 photos)]

Figure 24.2. Ten-second sequence of adult male Keola engaged in maneuvering and turning postures in front of the 1.2 m diameter one-way mirror. In contrast to this brief visit, interactions with real dolphins (controls) were characterized by a straighter, more perpendicular orientation (to the underwater gate) and much greater intensity, duration, and speed. (Small insets are a side view. Note that sequence proceeds from left to right for first four photos, then down to second row, etc.)

Each dolphin was observed and filmed after being marked, and recorded for latency to approach the mirror, time in front of the mirror, frequency of approaches to the mirror, and behavior while in front of the mirror. Behaviors at other locations in the tank were also noted. In some cases if a marked dolphin did not approach the mirror for approximately half an hour, we tapped on the window to attract it to the mirror area. Although enticing is undesirable, the important data was not whether the dolphin came. to the mirror, but what it did in front of it once it saw its mark.

Was the dolphin examining its mark? Guidelines for "passing" the mirror mark test. The purpose of the tests presented here was to evaluate objectively if the dolphin treated its mirror image as itself or as another dolphin. These guidelines were used to ascertain whether a given dolphin used the mirror to examine its mark:

1. The dolphin engaged in maneuvering, turning, or twisting postures that brought the mark into view. (The dolphins also occasionally performed similar postures when not marked. These were most likely instances of the animal simply using the mirror to examine the unmarked lateral area.)
2. The dolphin visited the mirror and made its mark visible to itself within 10 min of being marked.
3. The dolphin made at least one visit of 10 see or more to the mirror with the mark visible to itself.
4. The dolphin's physical distance to the mirror during a visit (including the one or more that lasted at least 10 sec) was no more than a foot.
5. The dolphin's eye was distinctly oriented toward the mirror reflection of its mark (optional: this was not noted in all cases, could only be seen in person, and was too subtle to show up on videotape data).
6. Sea Life Park staff familiar with these dolphins and their behavior, as well as colleagues who have seen videotape of the tests, had the subjective impression that the dolphin was using the mirror to examine its mark.

Control conditions. Three control conditions were used with some subjects for comparison with the above-described mark tests:

1. The mirror was available, but the animals were not marked.
2. The animals were not marked and there was no mirror available.
3. The animals were exposed to unfamiliar dolphin(s) through a metal grate, under varying conditions.

Results were videotaped, and samples of the data were analyzed. For control Conditions 1 and 2, we videotaped the unmarked dolphins' behavior both with and without the one-way mirror in place. We analyzed 24- and 35.5-hour samples from two different groups of dolphins (Table 24.2). The samples were picked to be close in time to the mark test. We looked especially for the maneuvering postures (Figure 24.2) we later saw in mark tests. We counted these postures and generally observed and characterized the dolphins' behavior. For control Condition 3 we videotaped behavior from the surface whenever new individuals met each other at Sea Life Park.


Mirror mark test results

The dolphins varied considerably in their interest in the mirror during the mark test and at other times. There was a brief period of heightened tank activity shortly after the marks were applied to Keola and Hot Rod, when all four dolphins in the tank engaged in fast, circular swimming for 2 min. All of the animals tested engaged in behaviors suggesting that the dolphins were examining the zinc oxide mark during the mirror mark test. In addition, all of the subjects played with objects in front of the mirror, suggestive of self-examination. Finally, each animal's behavior was unique, but gave evidence of utilizing the mirror in various ways. Results described here are not exhaustive, and are meant only to convey test highlights.

Table 24.2. Mark test results

Subject      A      B      C      D      E      F      G      H

Keola        1      5     13     13      5     24    0.08      yes
             2      0      0      0     NA
             3     17   35.5     15     11
Hot Rod      1     16   18.5      4      4     24    0.12 probably
             2      3    7.5      3     70
             3     23   54.5      9      7
Okoa         1      4   25       11      5    24&    0.25/hr&0 yes
Laukani      1     12   14       12     10    34.5   0.06      yes
             2      3   11.5      8     48
Itsi Bitsi   1     15   57       12     20    34.5   0.03 possibly
             2      0    0        0     NA
             3      4   14       12     46

A = Mark test
B = No. of deliberate visits to mirror
C = Total time in front of mirror with mark visible (sec)
D = Maximum single-visit time with mark visible (sec)
E = Latency to approach mirror after mark (min)
F = Mirror-no mark control: Length of sample analyzed (hr)
G = Mirror-no mark control sample: Maneuvering posture rate during      sample (times/hr)
H = Did dolphin examine its mark in first mark test?

Keola, 14-year-old adult male Atlantic bottlenose. In Keola's first mark test, he swam to the mirror while we were still in the process of putting it up. As he approached, we were producing the noises the mirror-application process makes. As he arrived we stopped the application process and made no further sounds. He positioned himself so that his mark was probably visible to him and engaged in some extreme "maneuvering" postures and contortions for 13 sees (Figure 24.2). In his second mark test, Keola started removing his mark right away by rubbing against the tank. Keola passed by the mirror many times during this and his third mark tests, but did not give the mark or the mirror as much special attention as he did in his first test. The gentian violet that was applied during his third mark test washed away in the tank water almost immediately.
In a later mark test done with the 0.8 m x 0.7 m rectangular mirror, on at least one pass by the mirror Keola swam so that his mark moved in front of the mirror, then stopped just as the mark reached the edge of the mirror. He remained relatively motionless for 15 see, with the mark still visible near the edge of the mirror, and then calmly swam on.

Okoa, 13-year-old adult female Pacific bottlenose. Within 5 min of being marked Okoa swam by the mirror in such a way that the mark was visible to her, but she did not stop and engage in any "maneuvering" postures. Thirteen minutes later she returned and stayed for 11 see, less than a foot from the mirror and roughly parallel to it, in a position where the mark was most likely visible to her. Her behavior included an unusual lifting of the midsection, a behavior that we scored as a maneuvering posture. Nineteen minutes later, upon being enticed to the mirror area from another part of the tank by our tapping on the mirror, she engaged in another 11-sec visit close to the mirror with the mark visible to her. This time she exhibited very deliberate maneuvering and turning postures, positions similar to those of Keola (Figure 24.2), and similar as well to six occurrences of maneuvering postures we counted during the 24-hour sample of unmarked control data for her.

Laukani, 14-year-old adult female Pacific bottlenose. In her first mark test, Laukani did not spontaneously go to the mirror like Keola and Okoa, but needed to be enticed to the mirror by tapping on the underwater window. Three times she was positioned within a foot of the mirror in such a way that her mark could have been visible to her. The first instance, 10 min after the mark was applied, lasted only 2 see, but included a brief maneuvering posture. The second instance lasted 20 see: She stayed horizontal and parallel to the mirror, only a few inches in front of it, with her eyes straining backward (toward her mark). The third was similar to the second and lasted 3 see, until movement inside the laboratory inadvertently caused her to leave.

Itsi Bitsi, 13-year-old adult female Atlantic bottlenose. Although she visited the mirror several times during her gentian violet mark test, and was positioned so that the mark was probably visible to her, she did not engage in any obvious behavior or postures that seemed to favor viewing the marked area.

Itsi Bitsi did not go to the mirror at all during her first zinc oxide mark test, but instead aggressively chased the other dolphins in the tank. In her second zinc oxide mark test she visited the mirror four times, but we had to entice her there by tapping on it. Once, for 3 see, she was positioned within a foot of the mirror so that her mark was probably visible to her. At this time she engaged in a brief, subtle maneuvering posture.

Hot Rod, six-year-old juvenile male Pacific bottlenose. During his first mark test, while the mirror was being applied to the window, Hot Rod swam by the mirror with his mark facing it, but did not stop. Four min later he engaged in maneuvering postures virtually identical to those of Keola in Figure 24.2, but with his unmarked side facing the mirror, then turned around immediately so that the mark was visible to him for 2 sec. Five min after that, he stopped in front of the mirror with the mark visible for 3 sec. In his third mark test, 7 min after receiving the mark he stopped in front of the mirror, where he stayed relatively motionless, with the marked side facing the mirror, for 10 sec. Two min later he did this for another 4 sec.

Control condition results

Mirror, no mark.
During the mirror-no mark control condition, the dolphins approached the mirror regularly. In one control sample, for the first 5 days the dolphins moved slowly and acted passive in front of the mirror, with long, stationary watches. After 5 days (50 hours of mirror exposure) their behavior shifted to more active body movements, mouth movements, bubble blowing, and open-mouth behaviors. During mirror visits, the dolphins engaged in the same type of posturing seen during mark tests: In a 24-hour sample, twice for Keola, three times for Hot Rod, six times for Okoa; in a different 34.5-hour sample, twice for Laukani and once for Itsi Bitsi (Table 24.2). (These counts are minimums.)

Interactions while facing the mirror were characterized by mouth movements, bubbles, vocalizations, and melon butts against the window; they appeared more social, aggressive, and interactive than sideways interactions. During sideways interactions, in which the dolphin's body is parallel to the mirror, the dolphins were more passive and relatively motionless, engaging in occasional body twisting, posturing, and slight humping of the midsection. In these interactions the dolphins seemed self-absorbed rather than interactive, and appeared to be maneuvering either to bring certain parts of their bodies into view or to stay in place in front of the mirror. Once, after approximately 200 hours of mirror exposure, Okoa remained motionless in front of the mirror and released a continuous, slow, fine stream of bubbles for approximately I' min. She repeated this behavior a short time later for a similar period of time, although the bubble release was not continuous. Her eye movement in both episodes indicated she was watching the bubbles rise to the surface.

During mirror exposure, the dolphins occasionally approached the mirror with an object and manipulated it. This play was sometimes characterized by the dolphin relocating the object in front of the mirror if the object moved out of view. After being fed, the dolphins often spit fish out while in front of the mirror.

Adult male Kamalii, who was never marked, was exposed to a mirror on one occasion for 2 hours. He approached the mirror several times during this test and engaged in a variety of behaviors. Once, he blew a bubble, and moved quickly through a series of maneuvering postures while whistling. Two min later he came close to the mirror, and while oriented perpendicular to it, engaged in rapid, rhythmic, repetitive shallow pumping of his mid and lower section, punctuated by an abrupt, audible sideswipe of his flukes at the end. The movement was accompanied by a continuous click train. The movement sequence lasted approximately 10 see and contained two bouts of the pump and sideswipe. Later, he returned and repeated the behavior, accompanied by the same rapid click train.

No mirror, no mark. When no mirror was present, posturing in front of the window occurred only occasionally, and was subtle and brief. In addition, the dolphins did not approach the window with the same frequency or regularity as when the mirror was up. Playing with toys occurred with no mirror, but infrequently. Spitting out fish also occurred with no mirror.

Encountering an unknown dolphin. If dolphins do not recognize their mirror reflection as representing themselves, then one might expect them to respond to the mirror in the same way they respond to encountering an unknown dolphin. (We found no literature on dolphin encounters with unfamiliar dolphins, although Ostman, 1985, and Overstrom, 1983, present some ethological and acoustic information.) In the third control condition we tested how dolphins respond to encountering other dolphins for the first time. The adult females (Okoa, Laukani, and Itsi Bitsi) were introduced to an unfamiliar adult mate, Kamalii, separated by an underwater barred gate with bars 15 cm apart. These females spent time inside the channel at the underwater gate during most of the 40 min of the experiment. There were often two or three females at the gate together. The dolphins appeared excited, but were not physically hyperactive (i.e., did not create a lot of turbulent water). All three females, but most frequently Okoa, blew bubbles. All four of the dolphins were usually oriented perpendicular to the gate. Laukani and Kamalii engaged in three bouts of open-mouth interaction with each other. During two of these, they touched rostra. During one, Kamalii engaged in a brief vertical undulating motion, much like simultaneously swimming and pushing his rostrum through the gate. In the third open-mouth episode, both Laukani and Kamalii looked as if they were taking small, rapid bites at one another. At one point, Kamalii swam slowly and closely by the gate and turned sideways, presenting his ventral aspect to Okoa, who released a large bubble. He immediately circled back and faced her.

Interactions with strangers of a closely related species through an underwater barrier. Four bottlenose dolphins, including juvenile male Hot Rod, encountered two false killer whales through an underwater barred gate. Although some of the dolphins had previously lived with a false killer whale (and one Sea Life Park dolphin has successfully mated with one, producing a fertile hybrid), they had never met these two. The dolphins gave the strangers their undivided attention for the entire 1-hour observation period. There was a brief period of prolonged, rapid, and loud airborne vocalizing during the initial confrontation across the gate. Unlike in mirror mark tests, the dolphins oriented almost exclusively perpendicular to the gate (there is much parallel orientation to the mirror in mark tests.) There appeared to be little or no body posturing. Their occasional open-mouth behavior was not accompanied by rhythmic head movements as it often is in front of a mirror.

No underwater barrier. Watching stranger dolphins interact without the restraint of an underwater barrier between them shows how the dolphins would behave if they were swimming freely and not separated by glass (as a subject and its reflection are). Keola, the adult male with the most convincing mark test, was put in a tank with Maka, an adult male Keola had never seen or heard before. The result was very unlike Keola's mirror behavior: high-speed synchronous swimming around the edge of the tank and quick, intense
movements. Keola repeatedly put his rostrum near Maka's genitals. Keola's intensity and undivided attention with respect to Maka lasted for most of a day (i.e., much longer than a mark test)-

Summary of new tests: Self-view television (See Note 1 below)

Using television to distinguish between social behavior and self-examination

When a dolphin looks in a mirror it often opens its mouth and moves its head around in some rhythmic fashion. In a mark test, as it does not have a hand to touch its mark, the dolphin maneuvers its body in various postures to see it. In our opinion, the mark tests and controls presented in this chapter suggest that these behaviors are contingency-checking and self-examination, respectively. However, another possible interpretation is that the dolphin is using postures to interact "socially" with what he perceives to be another dolphin in the mirror. We therefore conducted additional tests to distinguish self-examination from social behavior in the context of dolphin-mirror interaction. Space limitations do not allow them to be included in full in this chapter, but they are summarized here.

Television tests: Real-time self-view versus playback of same

In previous experiments we determined that dolphins respond to a televised image as if real (Marten & Psarakos, 1992). To distinguish self-examination from social behavior, we put a video camera aimed into the tank next to the dolphins' 20-in. television monitor. This setup is referred to as "mirror mode," or real-time, self-view television. A real-time image is displayed on the dolphin's television, allowing the dolphin to look at itself as in a mirror; this material is videotaped. When we later play back the videotape of a mirror mode session to the dolphins on their television - "playback mode" - the movements of the taped dolphin on the television no longer mirror those of the dolphin watching. The dolphins were exposed to alternating sessions of 10 min of mirror mode and 10 min of playback mode. One would expect the self-aware dolphin to behave differently in the two television modes: perhaps to utilize the television image for self-examination in mirror mode, and merely to observe in "playback mode." If the dolphin perceives its mirror-mode television image as another dolphin, then whether it is watching mirror mode or playback mode, the dolphin should act the same way - possibly interacting socially with the "television dolphin," or perhaps just observing it, but behaving the same in both modes. The results for the adult male Keola were clear cut: In mirror mode he spent quite a bit of time opening his mouth wide and moving his head in various rhythmic ways, whereas he never did this during playback mode. The results for juvenile Hot Rod and babies Maui and Tinkerbell, however, did not show such clear-cut differences between the viewing modes, although the babies seemed to spend more time in front of the television during mirror mode.

Television mark tests

Some of the dolphins who participated in the mirror mark tests were also exposed to the mirror mode-playback test described above. The video camera was located next to the television, to the dolphins' left, and thus favored showing the left side when they looked at the television. We marked them on their right sides to see if they would counter this bias by maneuvering the marked side into view. Indeed, presentation of the marked (right) side during mirror mode was three times what it was in playback mode, and 2.5 times what it was in the unmarked control.

Mouth mark tests

We wondered if the wide-open-mouth behavior that the adult male Keola exhibited in mirror mode, but not during playback, was Keola examining the inside of his mouth. To test this, we marked his teeth with colored denture adhesive and compared his open-mouth behavior in front of the television before and after he was marked. After his teeth were marked, he came to the self-view television twice and repeatedly opened his mouth extremely wide. However, in 20 min of control data immediately preceding his mark, he visited the television five times and engaged in open-mouth behavior during three of them. Nevertheless, he never opened his mouth as wide or as long as when he had the dental adhesive on his teeth.

Turn tests

We showed the dolphins a real-time frontal self-view on their television, then suddenly changed it to a side view to see if this would cause them to turn. If so, it would suggest that the dolphin might know that the image it is watching is itself. Two-year-old Maui's turn rate during the "turn trials" was five times what it was in control tests with no view-switching, and 3-year-old Tinkerbell, who never turned in the control situation (no view-switching), turned five times in the 39 turn trials of her last two sessions.


Mirror mark tests

We have concluded from the behavior of the dolphins during their mark tests that three of the adults, Keola, Okoa, and Laukani, used the mirror to visually inspect their bodies where the mark was located. The locations the dolphins chose to take in front of the mirror, and the body orientation they assumed during these inspections, suggested they were looking toward the mark. Their behavior while in front of the mirror also suggested this. For example, during one close approach to the mirror, Laukani moved her eyes far back, seeming to strain to see the reflection of the mark behind her head.

For some subjects, the data were harder to interpret in terms of self-directed behaviors. Hot Rod spent as much time as the adults in front of the mirror with his mark visible to him, but his movements were not as suggestive of examination as those of Keola, Okoa, and Laukani. The presence of the mirror and the mark did not appear to alter Itsi Bitsi's behavior. Interestingly, she was the animal who had received the vaseline sham marks prior to her zinc oxide mark tests. She had the shortest visit to the mirror of all the dolphins who were mark tested, and she was the only adult who did not spend more than 10 see directly in front of the mirror with the zinc oxide mark visible to her. Perhaps the novel tactile stimulus of the zinc oxide without sham marking is needed to elicit longer visual investigation in the mirror.

Interpretation of combined results from mirror mark and television tests

One of the most important results from this research comes from a combination of the mirror mark tests and the television tests. Keola, the only adult to have both tests, behaved significantly differently when the television was in mirror mode as compared to playback mode. His behavior strongly suggests that he did not perceive the television image as another dolphin in either mode and was not interacting with it "socially." When he was marked with zinc oxide, he almost immediately positioned himself within a foot of the mirror so that the mark could have been visible to him and then engaged in rapid, extreme, postures. It is very likely that he was examining his mark. The only reasonable alternative is that he might have been posturing socially to what he believed to be another dolphin in the mirror. Since this "social" explanation is rendered not very likely by the television tests, the data suggest that this dolphin examined his mark in the mirror during his mark test, and took his mirror image to be himself.


The comparison of the three control conditions with the mark test suggest that the dolphins' behavior during the mark tests was unique. Although behaviors suggesting postural adjustments to enable viewing otherwise invisible parts of the body occurred during the mirror-no mark condition, the dolphins, particularly Keola and Okoa, engaged in this behavior at a much higher rate during their mark tests. The virtual absence of these postural adjustments during the no-mirror condition further supports a self-examination interpretation of these postures.

Although there were similarities, Kamalii and the females' behavior at the underwater gate differed substantially from their behavior at the mirror. The biggest difference was their attention span: the females crowded at the gate, and one or more was almost always at the gate throughout the 40-min experiment. In contrast, when the mirror was present in their tank, single individuals made only occasional visits of a few seconds. Although open-mouth behavior occurred in front of the mirror, we never saw the type of rapid, open-mouth biting that Laukani and Kamalii engaged in across the gate. Bubble blowing occurred both at the gate and at the mirror.

As for play behavior, which occurred both with and without the mirror, it is our impression that mirror play behaviors were qualitatively different from play behaviors elsewhere in the tank, and that the dolphins appeared to utilize the reflective properties of the mirror while playing.

Conclusion from encounters with unknown dolphins

Overall, when we compare dolphins' mirror behavior to their behavior with unfamiliar dolphins, there are more differences than similarities. Mirror-directed behavior is different from social behavior in these experiments. The high rate of vocalization when the four dolphins encountered the false killer whale through the underwater gate was never duplicated by a dolphin in front of the mirror. Dolphins pay close, rapt attention to a real stranger for virtually 100% of a test period, even when it lasts hours. They pay close attention to their mirror image for less than 1% of test periods. This critical and important difference between the two types of conditions supports the interpretation that dolphins' mirror behavior is not social, and that they recognize their mirror images as themselves.

Rate of "maneuvering postures" in mark test compared to no-mark controls

An examination of the column labeled "Mirror-no mark control sample: maneuvering posture rate" in Table 24.2 shows rates of 0-0.25 per hour (average = 0.09/hour). Thus, the control data would predict these postures to be an unlikely event in a 1-3-hour mark test. However, during their mark tests, Keola and Okoa spent the majority of their time in front of the mirror engaging in such maneuvering postures; the postures occurred as well in Hot Rod's, Laukani's, and Itsi Bitsi's mark tests, although at a lower rate and more subtly. If this twisting behavior is considered to be social behavior, we see no reason to expect it to increase during a mark test; if it were self-examination, such an increase would be expected. The virtual absence of the maneuvering postures in the no-mirror condition further supports a self-examination interpretation of these postures.

Comparison with primates: Amount of time spent in front of mirror

Dolphins spend less time in front of mirrors than do primates. Children attend to mirrors and contingent video images on the order of 50% of their test time (Lewis & Brooks-Gunn, 1979); but a dolphin swimming in a tank is very different from a baby sitting in front of a mirror. The baby is more or less physically bound to remain sitting in front of the television, and is thus more likely to watch it. Dolphins, being primarily mobile, are less likely to remain in one spot for an extended period of time. This difference between dolphins and primates forced us to resort to enticing marked dolphins to the mirror; however, we view this as equivalent to putting a mirror in front of a marked primate. (As mentioned, dolphin attention changes virtually to 100% when faced with a real stranger, not a mirror, as in the unfamiliar dolphin controls.)
Furthermore, most studies of mirror behavior are conducted with the subject isolated from conspecifics. Our subjects were tested in social groups, and this may account for differences in behavior. Animals that were tested in larger social groupings tended to spend less time with the mirror. In this context, the dolphins were more interested in social interactions than in mirror interactions.

Dolphins appear to respond to the mirror quite rapidly, in comparison to data reported for primates. The explanation may lie in the presence of windows in their tanks. The quality of the reflections vary, due to time of day, but appear to be adequate to provide experience to the dolphins with reflective surfaces. The lack of novelty of reflections may also explain the relatively brief attention span of the dolphins during mirror exposure: Responses to mirror mark tests might diminish as the novelty of the new "good" mirror wears off.

Although dolphins have good vision, the relative importance of visual stimuli and visual images to them is open to question. In his first mark test Keola rushed over to the mirror as soon as it was being put up and appeared to took at his mark, but then did not return to the mirror for the rest of the 2-hour test. He did, however, eventually rub the zinc oxide off, on the side of the tank. This suggests he may have been more concerned with the tactile stimulus than the visual stimulus. We have the impression from our research that visual images do not have as high a priority for dolphins as they do for primates.

Contingency checking compared with primates

Primate researchers deal with their subjects as if the animals can recognize their own static images. We must guard against this kind of approach for dolphins, as the role of movement could outweigh visual form. It appears that in primates, both self-recognition and contingency checking occur in the great apes, but are absent in monkeys (Gordon Gallup, personal communication, 1991; contra Boccia, SAAH23). In humans, contingency play and self-recognition (as measured by a rouge mark test) appear together developmentally, whether the contingency play is in a mirror or on self-view television. Lewis and Brooks-Gunn (1979, p. 109) feel that this supports their belief "that contingent play is not a precursor but rather an indicator of self-knowledge." We observed contingency play in the mirror in all of the dolphins tested including the youngest (1.5 years). Anyone who watches tapes of dolphins from our laboratory looking at themselves in mirrors cannot help but be impressed by the rhythmic head and mouth movements and sounds involved. In dolphins, contingency checking may be necessary for self-recognition.

Environmental, social, and individual influences on test results

. Our tests were conducted in a situation in which the dolphins are accustomed to seeing people on the other side of the window used for the one-way mirror. They know it is people habitat, not dolphin habitat. Also, we were present during tests, so although the dolphins could not see us, they could hear us if we made any sounds. On the other hand, the test in Marino, Reiss, and Gallup (SAAH25) had the one-way mirror located in a channel leading to another tank. The dolphins were used to encountering strange dolphins and interacting with them there, and during these tests they could hear other dolphins behind the mirror. These differences in the test situation would make it much more likely for dolphins in that study to be deceived into thinking they were seeing other dolphins in the mirror than those tested here. Furthermore, when we resorted to tapping on the window to attract the dolphins to the mirror area, we reminded the dolphins that there were people on the other side of the mirror, not dolphins. Most of the Mirror-directed behavior we saw did not appear to be the kind of social behavior that generally accompanies encounters with other dolphins, such as hyperactivity and pronounced social cohesion.

Social. Dolphins are intensely social creatures, and the influence of social factors on these tests should not be overlooked. When there were not many dolphins in the tank (two or three), they interacted with the mirror and television quite a bit. When more dolphins were present (four or five), the dolphins were so busy with each other that they paid less attention to their mirror or television image. Social interactions could have influenced the animals' ability to focus attention on the test. A given animal's position within the social hierarchy may have contributed as well. For example, Itsi Bitsi ("top" dolphin at the time) failed to go to the mirror on her first mark test, and instead chased other dolphins, as if immediate hierarchical imperatives prevailed. In addition, our inability to isolate the animals may have affected the unknown-dolphin control test. The ratio of "unknown" to known animals is 1 : 1 in a mirror setup, whereas Kamalii was faced with three new females. This imbalance could have altered Kamalii's behavior. The gender of the individuals involved in a first-time encounter could possibly influence their behavior. The females' being of opposite sex to Kamalii (as well as of different race, in Laukani and Okoa's case) was a confounding variable that could have contributed to differences in behavior through the gate, compared to the mirror.

Individual variation. It is worth mentioning that positive and negative responses to tests like these may reflect individual personality differences rather than intellectual characteristics of the species. There may be a correlation between a natural proclivity for looking in the mirror and the use of it for self-examination. While unmarked, Keola was very interested in the mirror, and he gave intense results when he was marked. Throughout their mirror exposure when unmarked, Itsi Bitsi and Laukani showed no special interest or preference for the mirror, and had to be enticed over to the mirror when marked. The seemingly negative result of an adult animal being mark tested for self-awareness may indicate more about that animal's likes and dislikes than it does about self-awareness. Dolphins who show no natural interest in mirrors may not be good candidates for mirror mark tests.


No single test presented here proves self-recognition in bottlenose dolphins. The tests were developed mainly from primate research paradigms, and their limitations for interpretations of dolphin behavior are apparent. Nevertheless, the data taken together make a compelling case for self-recognition in this species. Four of five dolphins apparently examined their marks in a mirror; most brought objects to the mirror and played with them in front of it, even moving the object back when it drifted out of view; and most of the mirror-mode television tests designed to distinguish self-examination from social behavior suggested self-examination. Not only did dolphins attend to their mirror (or television) images less than 1% of the time as compared to 100% for real dolphins, but they engaged in different behavior with mirrors than they did with other dolphins. The results obtained in the experiments presented here are consistent with the hypothesis that these animals are using the mirror to examine themselves. More definitive results, however, will have to come from methodologies developed specifically for dolphins.


We express special appreciation to Don White and Dexter Cate for their dream of an underwater laboratory that would probe the dolphin mind by allowing dolphins to interact with modern human technology, and for their founding work on the lab. In addition, we recognize Don White for creating the project, conceiving the dolphin-video interaction system, and working to secure funding for the project. His thoughtful and valuable ideas, discussions, and suggestions have made this work possible. Sea Life Park, Hawaii, also made this research possible. The project is a cooperative effort between the conservation organization Earthtrust and Sea Life Park. Earthtrust supplies the research staff; Sea Life Park supplies the dolphins, their care, feeding, and maintenance, as well as the laboratory space. We greatly appreciate Sea Life Park's participation in general, as well as help with the mark tests from Sea Life Park curator Marlee Breese and trainers Stephanie Vlachos, Susan Rodgers, Roberta Horne, Keana Pugh, and Rich Nunes. Trainer Carol Chang provided valuable information on the dolphin Kamalii. We would also like to express appreciation and admiration for the tremendous help received from Lori Marino. Shortly after we began exposing dolphins to self-views and delayed playback, Lori and Diana Reiss independently exposed two juvenile males to a one-way mirror and conducted a zinc oxide control-mark and mark-test experiment at Marine World in California. Later, once we knew about each other's work through Dr. Reiss, our research, especially the methodology for systematic tests, benefited greatly from Lori's input. Gordon Gallup gave valuable advice for the mirror mark tests. Maria Boccia, Marlee Breese, Gordon Gallup, Emily Gardner, Don Griffin, Chris Johnson, Lori Marino, Bob Mitchell, Kati Nagy, Sue Parker, Diana Reiss, Don White, and Jana Wolff reviewed the manuscript. Dave Hack provided the schematic rendering of the tank. Soltech in Honolulu, Hawaii, and Security Glass in Pleasanton, California, donated the one-way mirror polyester film. Oceanic Institute provided facility support. The research was supported by Mollie Malone, Wendy Grace, the Boudjakdji Family, the Sarah Stewart Foundation, the Barbara Gauntlett Foundation, Outrigger Hotels, and the members and supporters of Earthtrust.


1. The television tests are described in detail in a paper entitled, "Using self-view televi-
sion to distinguish between self-examination and social behavior in the bottlenose dolphin (Tursiops truncatus)," submitted to the journal Consciousness and Cognition.


Anderson, J. (1984). Monkeys with mirrors: Some questions for primate psychology. International Journal of Primatology, 5(l), 81-98.

Gallup, G. G., Jr. (1970). Chimpanzees: Self-recognition. Science, 167, 86-87.
(1982). Self-awareness and the emergence of mind in primates. American Journal of Primatology, 2, 237-248.

Herman, L. M. (1980). Cognitive characteristics of dolphins. In L. M. Herman (Ed.), Cetacean
behavior.- Mechanisms and functions (pp. 363- 429). New York: Wiley.

Lethmate, J., & Dücker, G. (1973). Untersuchungen zum Selbsterkennen in Spiegel bei Orang-
Utans und einigen anderen Affenarten. Zeitschrift für Tierpsychologie, 33, 248-269.

Lewis, M., & Brooks-Gunn, J. (1979). Social cognition and the acquisition of self. New York: Plenum.

Marten, K., & Psarakos, S. (1992). Do dolphins perceive television as reality? Earthtrust
Chronicles (November), 8.

Ostman, J. (1985). An ethogram for dolphin social behavior, and observations on changes in aggressive and homosexual behavior among two subadult male bottlenose dolphins (Tursiops truncatus) in a captive colony. Unpublished master's thesis, San Francisco State University.

Overstrom, N. (1983). Association between burst-pulse sounds and aggressive behavior in captive Atlantic bottlenose dolphins (Tursiops truncatus). Zoo Biology, 2, 93-103.

Suarez, S. D., & Gallup, G. G., Jr. (1981). Self-recognition in chimpanzees and orangutans, but not gorillas. Journal of Human Evolution, 10, 173-188.

25 Kaneohe Bay Drive, Suite 205
Kailua, HI 96734 USA
(808) 254-2866
FAX: (815) 333-1158
E-mail: EarthTrust

Return to Project Delphis page