Study and analysis of animal cognition sets the findings and theories on human cognition in a broader comparative perspective. The study of the Atlantic bottlenose dolphin (Tursiops truncatus) offers a unique challenge to psychologists in that it possesses on average a larger brain than humans, demonstrates a very complex repertoire of vocal and manual behaviors, and is a social predator which hunts in cooperative groups, not unlike humans (and earlier hominids, presumably). The present form of dolphin and human cognitive abilities developed out of the same processing capacity of the early, rudimentary mammalian brain; more than 60 million years ago after phylogenetic separation, humans and dolphins continue to share a number of analogous cognitive, behavioral, and neuroanatomical traits. Similar behavioral and cultural pressures, such as prominent vocal behavior, protection through social unity and cohesion, hunting in groups, and mother-offspring transference of social norms, may have produced a communally-shared symbolic system in dolphins, as language has developed in humans. Thinking is not a unitary phenomenon or process. Examination of linguistic and cognitive abilities in other mammals, such as chimpanzees (Pan troglodytes and paniscus) and the sea lion (Zalophus californianus), a smaller-brained marine mammal, illuminate cognitive structures and processes which may be common, in different degrees, to all or most mammalian species.
Table 1. Taxonomy of Animals Discussed: Common name (Genus species subspecies)
Many scientists and laypersons alike ascribe a unique status to the human mind in evolutionary theory. While human anatomy and physiology is accepted as the product of selective pressures which acted in the past, some find it difficult to accept nonhuman analogs to human psychological phenomena. Philosophical issues aside, proponents of discontinuity theory argue that some elements of behavior and mental experiences are qualitatively unique to a given species, sharing no evolutionary heritage with other living or archaic species (e.g., language in Homo sapiens sapiens; Chomsky, 1980). Discontinuity theory removes psychology from biological and evolutionary paradigms; human abilities are studied as novel specializations without preadaptation in other primates. Proponents of continuity theory, in contrast, argue that mental traits of extant species, like other biological traits, follow the rules of natural selection, have evolutionary precursors in more primitive forms, and appear within related species which have descended from a common ancestor (Roitblat & Herman,1988). Other theories, such as homology and convergent evolution, may be useful in the isolation, description, and explanation of related phenomena such as mental representation, social behaviors, and even human language. The concept of homology posits that a high degree of genetic relationship may imply a high degree of physical and psychological similarities. Proponents of evolutionary convergence maintain that similar ecological pressures is likely to result in similar traits, physical or mental, being expressed in otherwise evolutionarily divergent species. The study of linguistic or language-like behavior in nonhuman animals allows the construction of an intellectual framework in which to conceptualize and interpret human behavior and human evolution.
The human mind is the result of millions of years of evolution, the last four or five million having occurred separately from any living species. Although hominid brains have greatly increased in size over the last two to four million years, it is not clear that there has been any qualitative reorganizations of neural machinery (cf. additive model, Jerison, 1973; cf. Holloway, 1968, for opposite view). Heuristic neural models based on parallel distributive processing theories of brain function (e.g., McClelland & Rumelhart, 1986) suggest that brain size and neural interconnectivity are the critical variables which determine whether a behavior appears as intelligent or instinctive. Increasing numbers of neurons and increasing numbers of connections per neuron automatically result in increased differentiation of sensory and motor units, without large-scale neural reorganization. Large, highly-interconnected brains produce, by nature of their architecture, increased numbers of discrete sensory and motor behaviors. A consequence of brain size expansion in hominid ancestors, solely on the basis of architectural constraints, therefore would have been the development of more discrete motor and sensory units (of both representational and communicative kinds) than possessed by other smaller-brained animals. Combination and recombination of these discrete units may have enabled varied and more hierarchically constructed behavioral patterns -- the basis of language and higher cognitive functions (Gibson,1988;1990). In this sense, human intellectual development saw primarily the expansion of general information processing techniques in which some elements could have been accentuated, others diminished; but overall there may have been few if any structural or functional alterations in cognitive processing strategies. Relative brain size, as indexed by a structural encephalization quotient (SEQ), for example, which involves the brain enlargement beyond requirements for controlling the body, is related to information processing power (Jerison, 1986). Increased relative brain size correlates with other neurological changes as well: an increased size of neocortex, an increased size of association areas, and an increased neuronal interconnectivity (Jerison, 1973). The best index of intelligence and cognitive abilities may be cortical surface area (Jerison, in prep.). The human neocortex is 3.2 times larger in volume than it would be in a nonhuman primate with our body size and proportionally larger still in cortical surface area (Jerison, 1973). However, many species of odontocetes including the Atlantic bottlenose dolphin (Tursiops truncatus) has a greater average cortical surface area than humans, with a comparable body size to that of a human. Using this criteria, the brain of the Atlantic bottlenose dolphins, which also has an average EQ above all other animals except humans, should be capable of supporting highly complex representations (Jerison, 1986). However, it is risky to equate intelligence with allometric brain size; other variables need to be evaluated to determine the relative intelligence of a species. (Some have argued that cross-species comparisons of intelligence are meaningless; e.g., Bullock, 1986.) Human intelligence may have developed from prolonged maturation and other neoteneous traits (Gould, 1975). Compared with most mammals, however, higher primates are not neurologically altricial at birth (as might be expected of a highly neotenous species). Neonatal brain weight (as proportional to full adult brain weight) in higher primates fall in the precocial range: human neonates possess 25% of their full adult brain weight, whereas bear cubs possess only 2%, rats, cats and dogs approximately 10%; Tursiops possess 42.5%, killer whales (Orca orcinus) 50%, and ungulates as much as 75% (Gibson 1990b). Human maturation is longer than any other species, which confers at least two intellectual advantages: (1) the youthful period of play and learning is extended into at least the second decade. In contrast, apes must be able to forage for themselves after only 5 or 6 years of life; Atlantic bottlenose take 9 to 10 years to reach adult levels (Ridgway, 1986b); and (2) the brain is an extremely expensive organ in terms of metabolism and a long period of brain growth serves as a developmental buffer against serious malnutrition (and consequent brain damage) in larger-brained species (Morgan 1990).
Using a variety of molecular "clocks" biochemical studies place the Pan-Homo divergence between 3.5 and 5.5 million years, a relatively recent event (Sarich & Cronin, 1976, cited in Wolpoff, 1985). A minimum divergence date from paleontological evidence for the Pan-Homo divergence (Pliocene fossil) is 4 million years (Johanson and White, 1979, cited in Wolpoff.) Bipedal locomotion had already evolved close to 4 millions years ago, as shown by footprints found at Laetoli (Leakey & Hay, 1979). Stone tools are found as well as some evidence of cooperative hunting appear as early as two million years ago (Isaac 1979). DNA-DNA hybridization experiments indicate that the gorilla ancestor split from the ape-human stem approximately 8.0 -9.9 million years ago (mya) (Wrangham,1987). According to this method, the split of hominids from the common ancestor of chimpanzees, bonobos, and humans is determined to have occurred between 6.3-7.7 mya. Although there is criticism of exact dates provided by these and similar techniques, the general sequence of events provided by such methods is generally agreed upon: ancestral members of gorillas split prior to the Pan-Homo divergence, the two chimpanzee species split subsequent to the Pan-Homo divergence, as recently as 2.5 million years ago.
Behavioral similarities also suggest a closer relationship within the chimpanzee-bonobo-human clade (Pan paniscus, Pan troglodytes, and Homo). In this group we find female exogamy and the strict retention of male offspring in the natal social group, which is uncharacteristic of gorillas and orang-utans. According to Ghiglieri (1988), this sets the stage for the evolution of cooperative behaviors among male kin, from which communal reproductive strategies and more complex social groups evolved. The more moderate sexual dimorphism in body size in this clade (in contrast to the immense sexual dimorphism in body size in gorillas) suggests that members of these species have relied for a long time on cooperation among male kin. Fusion-fission sociality, which describes the social organization of humans, is found in both species of chimpanzees (Ghiglieri,1988).
Some have gone so far as to conclude that the pygmy chimpanzee may be a possible prototype for the common ancestor of humans, chimpanzees, and gorillas. (Zihlman, Cronin, Cramer, & Sarich, 1985). Pygmy chimpanzees (paniscus) are generally more neotenous than common chimpanzees (troglodytes): they possess a narrower trunk, a more gracile body, with smaller facial and canine dimensions. Pygmy chimpanzees have a hominid-like appearance: the sclera of the eye is white and the eyes themselves are round and large while the ears are small (Savage-Rumbaugh, Romski, Hopkins, & Sevcik,1988). Behaviors once thought unique to humans are also demonstrated by this species. For instance, ventro-ventral copulation is common; the female is swollen and copulates throughout her cycle. In captivity (and presumably in wild) the males and females other than the infant's mother of the species take an active role in child care. In captivity, pygmy chimpanzees are known for walking more bipedally than common chimpanzees (Savage-Rumbaugh et al..1988). Paniscus use a wider range of gestures and displays a greater diversity of social-sexual patterns. For instance, in captivity male pygmy chimpanzees have been observed using gestures to direct the position of the female into a variety of different postures for copulation (Savage-Rumbaugh et al., 1977). Socially, they possess more stable subgroups and male dominance is less pronounced than in common chimpanzee societies. Male-male cohesion is an important social factor in troglodytes; male-female bonds appears to be strongest in paniscus. There is no sexual dimorphism in cranial capacity, limb bone lengths, or robusticity in this species. They also possess a brain size comparable to Australopithecines (485 cm2) compared to Pan paniscus (350 cm2). Some investigators believes pygmy chimpanzees are a good living model of the early Australopithecine hominids and hence may be one of the most interesting ongoing investigations available to scientists. Of note, tool use has not been observed in this species, whereas tool use in common chimpanzees is well documented (e.g., Goodall, 1986). Common chimpanzees have also adapted to a wider range of environments. These were thought to be two qualities which signal evolutionary advance of cognitive skills. Examination of the differences between these two species may help us understand what was and was not retained or modified in hominid evolution and why humans achieved ecological dominance on the planet while other higher primates such as the pygmy chimpanzee did not.
The most prominent argument against continuity theory concerns human language. Development of linguistic capacities in hominids was the result of a recent radical adaptation in hominid evolution . For instance, Chomsky (1980) believes that the "evolution of language faculty was a development specific to human species long after it separated from other primates." Underneath such statements is also a misguided definition of language as an all-or-none phenomenon: either an individual possesses all aspects of language, or none at all. Language and other related cognitive capacities are not necessarily all-or-none phenomenon. It is not the case that an animal has complete veridical recall or no memory at all. Also, children under two, deaf children of hearing parents who create their own pidgin sign languages, and minor aphasics (e.g., a patient with slight anomia) would not possess language using this criterion, in spite of a being able to communicate to others on a variety of topics. Chomsky (1980) justifies his position by stating that vocal language results in obvious advantages: its development enabled the proliferation of the human species on the planet, hence it would be a "biological miracle" if a similar capacity in other primates existed but went unused. The communicative advantage of human language follows from the high rate at which information can be transmitted by sound signals (i.e., 5-7 syllables/second; 15-20 phonemes) (Lieberman, 1981). But adaptations do not develop in a "vacuum": they occur within an already complex network of social and ecological (pre)adaptations. If language evolved because it facilitated hunting in hominids, for instance, a social but less carnivorous humanoid species might have significantly less use for a vocal language. Other one-time candidates for establishing a unique status for humans in the animal kingdom (discontinuity) include tool use, tool fabrication, culture, cross-modal perception, religion, reason, and self-awareness, among others , which are unsupported by findings in cognitive ethology and related research (see Table 2 [note: from Humphreys, 1976, not reproduced here).
Language is claimed to facilitate free expression of thought and clarify one's ideas, as well as help establish social relations and to communicate information (Chomsky,1980). However, investigations which attempt to demonstrate how language facilitates cognitive abilities or development have met with controversy and rejection (e.g., Cheng, 1985). And other species establish social relations without human language that appear to be equally or more stable than human relationships, such as killer whales, and dolphins (Conner & Norris, 1982; J. Mann, personal communication). Nonverbal components of speech, such as paralinguistic and pragmatic cues, play significant roles in human communication (Brown, 1977). Human language is also unusual in that it is "a behavioral system that permits so many failures of communication" (Jerison, 1988). Linguistic processing can limit the free expression of thought as well (e.g., a dissociation between linguistic [preparation] and nonlinguistic [incubation] stages in creative problem-solving tasks may be necessary for illumination of a solution, Bogen, 1986). The loss of verbal language may even clarify thinking in artistic individuals (e.g., Gardner, 1979). Notwithstanding these claims, language is a symbol system shared by a community. Language is grounded in a social convention that attributes to certain substitutes (signifiers) the power to designate other substitutes (referents) (Vauclair, 1990). Like any symbol system, it is composed of a reasonable number of basic elements (e.g., 30-50 phonemes, or cheremes in sign languages; 26 letters in English, etc.) which in humans are used in both communicational and representational capacities. The dual roles of this symbol system is what makes human language powerful. The representational capacity of human language has promoted the phylo- and ontogenetic development of, among other concepts, the duality of patterns, which in turn has strengthened the communicational role of language, facilitating the acquisition and storage of knowledge.
It is widely recognized that many species from different phyla exhibit behaviors that express an internal or symbolic processing of external reality (Roitblat, Bever, & Terrace, 1984). There is also no social behavior without communication (Evans & Bastian, 1969); few species exist which do not possess a substantial repertoire of symbolic communicative devices such as vocalizations, displays, postures, and/or gestures. But these two symbolic systems may remain separate systems with little conceptual or featural overlap in most species. This may be where the location of human faculties and nonhuman faculties separate. If so, many questions need to be answered. For example, when did substantially integration of these symbol systems evolve in humans? Integration is in no way complete in present-day humans. Some elements of communication, such as phonemes or words, are used extensively in representational roles while others are hardly used at all, such as gestures or postures; and conversely, some aspects of our representations are cannot readily be communicated (e.g., the tip-of-the-tongue experience). To what degree are these two symbol systems integrated in humans? Do other mammals also use symbols which possess communicational as well as representational elements, and to what degree? What is the nature of communicative and representational integration? Which neural systems are responsible for this integration in humans? In other mammals? Does this integration affect either or both systems? The degree of integration is likely to vary in different domains of human experience, but behavioral and cognitive tests of human capacities may not show this for three reasons: (1) linguistic abilities of humans may overshadow or act to overcompensate weaknesses; (2) when nonhumans demonstrate behaviors comparable to that of humans (e.g., particularly young children), unwarranted capacities are generally ascribed to humans than to nonhumans (Greenfield & Savage-Rumbaugh, 1990); and related to the first reason; (3) the complexity of both communicative and representational processes in humans makes investigation of these questions difficult and the findings less reliable. The paradigm of language instruction in animals provides a structured environment in which variables can be isolated and explored.
Premack (1985) found evidence for integration of communicative and representational systems in the behavior of a common chimpanzee. According to Premack (1985) words have two functions in human language: (1) the external function of words is to retrieve or communicate information; and (2) the word serves as an intrinsic part of mental representation. He tested (1) the communicative function of lexigrams (plastic word) by determine how effective each lexigram was evoking a mental representation in a chimpanzee (troglodyte). He found that the most salient physical feature of a fruit for a chimpanzee was color: specific colors were effective in retrieving the animal's mental representation of the object. However, names were even more effective (e.g., fewest errors in tests). Consequently, lexigrams (plastic words) proved to be extremely effective communicative devices. To test (2) the representational function of a word, he applied a phenomena observed in the performance of matching to sample testing in apes. A blue triangle (lexigram), which refers to a red apple, can be matched to a red patch of color correctly; but a blue-painted apple cannot be matched to a red patch. Somehow a blue-painted word evokes a red apple whereas a blue-painted apple does not evoke a red apple. Either the blue apple is not recognized as an apple (which is very unlikely), or perception of a blue apple interferes with the representation of a red one. He found that perception of a distorted example of an object adversely influences the representation of a normal example of this object . Premack (1985) terms this the impairment effect. The adverse effect occurs within the same form of representation only (an object must be compared to another object). Evidence for an impairment effect of words would indicate that physical features of a word are being represented by a chimpanzee. It turns out that perception of a distorted word (orange triangle) does interfere with representation of a normal word (blue triangle). Chimpanzees find it difficult to match a distorted lexigram for apple (orange triangle) with a blue patch (normal color of word for apple). In contrast, chimpanzees could match blue-painted apples with a red patch of color . The impairment effect arises when perception and representation of the same form of information (lexigram and lexigram, or object and object) are incompatible. The impairment effect does not occur when perception and representation of different forms of information (lexigram and object) are incompatible. There is no impairment effect between word and referent; only when referent and referent or word and word are incompatible in same way does the effect arise. The impairment effect is offers a powerful tool, albeit complicated, for comparing mental representations of different species, and different levels of development within these species. Premack (1985) concluded that, like human words, plastic lexigrams possessed both representational and communicative functions. In spite of this demonstration of mental continuity, the degree of integration of representational and communicative systems may still quantitatively separate human and nonhuman experiences.
How is human language different than other communication systems in other animals? Humans have developed many communication systems that have little to do with language (e.g., stop lights). A definition of language would be useful here; but herein lies some of the difficulty of studying this problem: there is no explanatory nor conceptual definition which is widely agreed upon. Chomsky (1980), for example, states that the most elementary property of human language is that it involves a "denumerable infinity of functionally distinct expressions", which also describes the system of mathematics. And the productivity of human language is not infinite, but specific to those discernable elements in the social and physical problem-space of human experience (e.g., Parker, 1985). One fact is certain: language is a behavior that has certain consequences, and these consequences are in terms of representation and communication. An utterance has both an ideational meaning and interpersonal meaning or social purpose (Parker, 1985). Language conveys propositional and referential information and it provides alternative ways of expressing ideas as well as means for the speaker to communicate effectively, engagingly, appropriately (Parker, 1985). The system of language consists of hierarchically organized levels of processing (i.e., phonology, morphology, semantics, syntax, pragmatics) and consists of a number of design features (Hockett, 1960; see Table 3). It should be noted that many of these design features are physical characteristics of spoken language and every design features presented by Hockett (1960; Hockett & Altmann, 1968), except perhaps reflectiveness, is used in part by some animal communication system. Vauclair (1990) provides a useful conceptual definition of language: Language is a system that is both communicational and representational, grounded in a social convention that attributes to certain substitutes (signifiers) the power to designate other substitutes (referents). But was the process of integrating representational and communicational symbol systems abrupt or gradual? Did it occur before or after Homo split from Anthropoids?
TABLE 3: DESIGN FEATURES OF HUMAN LANGUAGE (from Hockett, 1960)
Added in 1968 in Hockett and Altmann
There are nearly as many questions about language evolution as there are theories, which probably number only slightly fewer than languages in the world. When did language evolve? Why did language evolve? How did language evolve? What effects did it have on behavior and cognition? When and how did primate vocalizations come to be supplemented, transformed, or replaced by the system of human language may be answerable. The transformation from ape to hominid entailed some hundred morphological, physiological, and behavioral evolutionary changes (Wind,1981). However, the changes may not been radical alterations. For instance, according to Wind (1981) the morphology of the apes' vocal apparatus cannot account for their inability to speak. He argues that the primate pharynx and airway were preadapted for speech-like vocalizations ever since the origins of the anthropoids. "(I)f a chimpanzee larynx could be grafted into an otherwise normal human being and if all the nerves could be connected such a human individual would be able to produce vocalizations and speech hardly or not discernible from the normal ones." Similarly, the human acoustic system shows its optimal performances at the pitch frequencies which predominate in speech consonants, around 2 kHz. Wind (1981) notes that the sounds emitted by breaking dry branches and leaves are also at 2 kHz. The hominid peripheral vocal organs for speech was likely preadapted as a result of selective pressures operant in arboreal species. In his opinion, the development of the association and motor areas of the brain was decisive for the origin of language as we know it. Falk (1980) also questions the supposed late appearance of a modern articulatory apparatus, saying that fossil reconstructions have been insufficient to determine if qualitative alterations have occurred. Wind (1982) believes that if the anthropoid vocal tracts were properly wired neurologically, apes would be capable of producing a sufficient variety of sounds to demonstrate at least some rudimentary speech. Fossils of Australopithecus robustus display an upper respiratory system closely akin to that of modern nonhuman primates, particularly apes. Subsequently, Laitman (1981) concludes that this species possessed a limited range of vocalizations. Fossils of Homo erectus, however, show a descent of the larynx, which would enable (with the pharynx) a greater range of vocalizations than possible to Australopithecines. An upper respiratory system similar to that of modern man, with a vocal apparatus enabling human speech, is found as early as 300,000 years ago in archaic Homo sapiens. Lieberman (1985) reports that major changes to the upper respiratory systems have taken place in the last 250,000 years; that even Neanderthal hominids retained most of features of the nonhuman supralaryngeal airway (e.g., they weren't able to produce the sound [i]) and would not have been able to encode speech as rapidly as even archaic Homo sapiens.
Language is more than rapid vocal behavior; neural mechanisms developed for encoding and decoding phonemic communication are also necessary. Tobias (1987) reports that Homo habilis possessed a prominent Broca's area in the posterior part of the left inferior convolution, which exceeds the prominence found in homologous areas in Australopithecus africanus . The pattern of sulci in this region of the brain of Homo habilis is comparable with that of modern humans and much different from apes. .. Tobias noted that the area around the parioeto-occipto-temporal junction (i.e., Wernicke's area) shows especially strong development in all four of the endocasts of H. habilis skulls. Of interest, he find no evidence of laterality (i.e., pronounced development more pronounced on one side than the other). However asymmetry of the Sylvian fissure is indicated by impressions in the endocranical walls of Homo sapiens neanderthalsis, Homo erectus, and even Australopithecus africanus (LeMay & Geschwind, 1975). However, the first true speech sounds may not have been uttered until very recently. Among the many neuroanatomical changes which occur during the hominidization process, there has been a progressive complication of the vascular system, particularly surrounding the sylvian region of the brain. As recently as 30,000 years ago, endocranial wall impressions of Cro Magnon humans were complicated, indicating better vascularization of the sylvian region than those humans who preceded them; but these impressions suggest that the vascular system was not developed enough to support speech. It was not until neolithic times (10-12,000 years ago) that first fossils which possess a "squaring of the parietal meningeal vascular system" are found, which indicates that humans now possessed the developed vascularization necessary for speech (Saban, 1981). Jaynes (1976a) argues that the evolution of speech cannot be detected by fossil remains (i.e., a skill may by physiologically possible but latent). However, the origins of spoken language, the acquisition of words, etc., may have produced behavioral changes in those hominids which possessed speech, and these cultural/behavioral changes may be reflected in the artefacts left behind. According to Jaynes (1976a), speech produced three changes which benefitted hominids: (1) spoken words enable one to train attention on specific salient features of objects and events (in contrast, feral children have more difficulty training attention, possibly due to however other factors besides lack of language; (2) verbal labelling also facilitates recalls (cf. Herman, 1986, analogous ability in dolphins); and (3) language allows one to code and compare attributes of objects verbally, thereby freeing us from the momentary perceptual impact of one attribute or another.
When did speech originate in hominid? First, the supposed radical change from a primate communication system to language must have occurred at a time where its benefits outweighed in disadvantages (e.g., greater likelihood of choking). When was there great enough ecological pressure to evoke such a change? Glaciation was likely the strongest ecological force at act upon hominid evolution. Each Ice age, lasting approximately 70,00 years, would have brought about a change in the habits and livelihood of hominids living at the time. There have been four major glacial periods since the transformation of H. erectus through archaic H. sapiens to H. sapiens sapiens: the coldest periods being approximately 600,000, 400,000, 150,000, and 35,000 years ago. Warm interglacial periods are probably without sufficient ecological challenge to provide language with any new survival value if it had developed during these periods. As mentioned above, the brain structures which mediate language -- Broca's area in frontal lobe, Wernicke's area around the Sylvian fissure, and the supplementary motor cortex -- are not strongly developed until H. habilis. The artefacts left behind from H. habilis and H. erectus were crude, suggesting little change from those of Australopithecines. The complete development of language was not likely. Although there was some progression in tool manufacture during the first three glaciation, the change was very gradual. Speech subsequently must have originated during the fourth glaciation. This age was characterized by large swings of temperature, beginning 70,000 years ago, achieving its coldest period about 35,000 BC. Normal temperatures returned around 8000 BC. An explosion of artefacts and new technologies coincides with the middle of the last ice age, approximately 40 - 35,000 BC (Jaynes, 1976a). The human brain had reached its present proportions (including its present-day increase in frontal lobes) between 250,000 to 100,000 years ago. Applying these time constraints, Jaynes (1976a) places the origin of speech at approximately 40,000 BC.
Attempts to reconstruct the earliest languages (e.g., Indo-European languages from 2000 BC) by linguists and anthropologists suggest that these languages were as complex as modern languages. If there has been little change in linguistic complexity in 4000 years, how much change could there have been in 40,000. The stability of language complexity suggests a prolonged phylogeny of language, possibly stretching back million of years. The rapid change in technology may not indicates a rapid change in cognition and linguistic capacities. Perhaps what occurred at this time of great cultural change was a cultural shift. Language may have already been present before this period, but the means and value of acquiring knowledge may have changed; knowledge itself gained a new value, beyond its value for helping to achieve an immediate goal. For the first time the distribution of labor may have included individuals who were not primarily hunters or plant gatherers but whose sole social function involved the transmission and storage of information, such as shamans, wise man, etc. The cognitive capabilities of humans probably have not change qualitatively since the evolution of language; however investigating the change in tools and artefacts (e.g., from stone tools to cities and nuclear power plants) are the work of the same intelligence. The rise of industrialization via mass automation marks a change in an individual's relationship to the process of knowledge acquisition. The individual contributes to his own knowledge as well as serve the cultural knowledge base.
Why did language evolve? What ecological or social pressures resulted in the development of speech? Gibson (1981) places the selection pressure within child-parent communication. The dietary niche of hominids -- omnivorous, extractive foraging with tools -- resulted in offspring depending heavy on parental guidance for long periods of time. Sophisticated recognition of requests for aid, for instance, for a specific tool or food item, would have required greater sensitivity between communicants. According to Jerison (1988) early hominids invaded an environmental niche (social predator) that was inappropriate for primates in critical ways. The niche of a social predator required behavioral and morphological adaptations which were not present at the time, nor likely to develop. Living species in this niche (e.g., wolves) must navigate, control, and defend an extensive territory and range. A scent marking system is ideal for differentiating territories (e.g., urine cues). Higher primates, with reduced olfactory system, would likely have had to rely on their most developed system, that is, the auditory-vocal system, to contribute to the same kind of cognitive mapping of the external world. Maintenance of these cognitive maps over time required active participation. This could be done by "linguistically-labelling" landmarks (e.g., a river labelled by a particular whistle; an old tree by a phoneme-like sound). Construction of these sensory map overlapped with the primate communication system (i.e., occurred in same modality). This may account for "the peculiar feature of human language." In language humans share a constructed reality; whereas animal communication (apparently) consists of primarily direct commands about behaviors. "Self-consciousness (arose) to distinguish the reality generated by one's own information (sensory, linguistic, etc.) from the reality generated by verbal information from another individual." (cf. Jaynes, 1976b). Livingstone (1981) also argues a similar origin of language. However, those mammals which do exhibit territorial vocalizations possess small territories, where all regions of the territory are within easy auditory proximity. Baboons and chimpanzees, who usually occupy large home ranges not unlike hominid hunters, possess no specialized territorial displays. Fischer (1981) postulates a vocal onomatopoetic theory. Speech originated in the "magical" imitation of other species' cries (food calls, mating calls). Such strange, less human vocalization would have allowed humans a closer approach to prey than ordinary cries. Unlike earlier hypotheses, there is some present-day support for this theory. For example, Amazon basin Indians imitate 35 different species, not to mention the wide use of decoys in hunting.The ability to imitate a wide variety of sounds of prey, and other environmental sounds as well, would have been selected for, as well as the imitation of postures, movements, and gestures (Fischer,1981). This theory is intuitively appealing in that, in essence, it means that Nature, the elements, taught human to speak. Many researchers (e.g., Falk, 1980) posit that deitic and iconic gestures were originally used to communicative references during hunts, etc., which were slowly supplanted by vocal referents. This may explain the ubiquitous supportive presence of gestures during speech. However, individual communications may have evolved not to describe the physical world as much as the social world (cf. Bateson, 1966; Humphreys, 1976).
How did language develop? Did it developed in parallel with cognitive abilities? Is it true that "language is less a gift of the gods than an exploitation of the primate potential" (Desmond, 1979)? Gibson (1981) compares the acquisition of object manipulation with the acquisition of language skills, In phylogeny and ontogeny both of these abilities mature through differentiation of existing behaviors into smaller component parts, whose parts are combined and recombined into new and varied behavioral patterns (see above). This occurs vocally, semantically, and syntactically. Gradually, through babbling, infantile coos differentiate into phonemes, which are combinable into a virtually infinite variety of words. Similarly, the semantic meaning conveyed by a child's first utterances is not clearly differentiated and parents must judge from the immediate context of the utterance. For example, the utterance "milk" made by a young child could indicate the simple recognition of milk, or the desire for a drink of milk. As grammatical and semantic skills developed, the child becomes capable of constructing sentences with various distinct meanings. In terms of syntactical constructions, the speech of children begins with single word utterances. Gradual increases on the mean length of utterances occurs during the second and third years, coinciding with the development of grammatical competence (Brown, 1980). The first grammatical constructions are those of simple position: e.g., agent-action-object. The ability to differentiate agent and object on the basis of more complex grammatical rules (as in "the girl is kissed by the boy" or "the girl that the boy kissed") does not emerge until about five to six years of age (Limber, 1980). Children at this age can construct hierarchical embedded structures and are able to judge the meaning of an utterance by the sum total of grammatical features, rather than by a single feature or position alone.
A constructional theory of the origin of language postulates a close correspondence between the communicative and tool-using abilities of primates (Reynolds,1981). (This theory is particularly interesting in light of the possible dissociation of these capacities in the two Pan species). According to this type of theory, the precursor of human linguistic structure may be found in anthropoid constructional ability in which objects are arranged into new functional configurations (such as a ladder). Constructional actions possess the following properties of (1) intentionality, (2) high-speed execution of both sequential and simultaneous constituents, (3) the creation of a new entity from constituent parts, (4) recursiveness, (5) generalization to new contexts, and (6) nesting of one operation within another. Ape ladder building has been observed in captivity. Elements of this skill which parallel the structure of human language are: (1) extensive practice of voluntary motor movements; (2) ladders can be built quickly; (3) new construction not reducible to constituent behaviors (no single element of ladder permits the ape to climb higher); (4) each pole in one ladder is often used in functionally different ways or positions in constructing other ladders (evidence of recursive application);(5) the ability is generalized to new contexts, with different poles, different locations; (6) behaviors are nested in that the output of one support operation is used as the input to another; and there is no implicitly correct order of actions in ladder construction, actions are executed within a hierarchical tree structure. Individual apes build each ladder, but it is use socially to the extent that others often hold or steady the ladder as others are climbing it. Hence, human language is the product of general evolutionary changes of primate constructional ability (Reynolds, 1981).
Gibson (1981) rejects constructional accounts of linguistic development which are closely tied to tool use. Gestures like tool use and object manipulation both depend upon differentiation and construction in visual and manual modes, but speech depends on vocal and auditory modes. The difference between the abilities in the two species of Pan may indicating some evolutionary divergence between vocal skills and the capacity for object manipulation. Neuropsychological evidence indicate that the highest constructional levels of tool use and language use are both mediated by the inferior parietal and anterior frontal association areas. For instance, lesions in these areas can result in any of the following: ideational apraxia (inability to use objects properly), ideomotor apraxia (inability to imitate gestures), inability to name objects, and inability to understand or construct complex grammatical relationships. But she notes that while aphasics need not be apraxic, in general apraxic patients are usually always aphasia (Gibson, 1981). She concludes that the synchrony between maturation of object manipulation and language skills, in light of neuropsychological evidence, reflects two closely tied but separate neurological processes.
This does not deny a possible fundamental parallel between language and skilled movement sequences. Complex motor skills are hierarchically organized: i.e., "the responses contained in a movement sequence exist as sets of sequential dependencies that effect the probability of subsequent responses" (Gibson,1981, pp.20-21). Calvin (1988) suggests the sequential processing involved in syntactical processing was first developed in complex motor sequences, such as throwing objects to hit targets. The sequence of muscle commands which are necessary for walking or breathing may be viewed as a preadaptation for rule-governed syntactical processing (Lieberman, 1981). Gregory (1970) suggests that the rule-governed nature of human language is an extension of neural rules that order retinal patterns into objects. The phylogeny of language involved a "take-over operation" in which humans exploited the development of the visual system in higher primates to structure vocal signalling: representational and perceptual processes were integrated with communicative processes.
The study of language in apes began with Furness (1916; see review, Ristau & Robbins,1982). Furness (1916) taught a female orangutan to produce vocally, papa, and cup and th over 11 months of instruction. Great apes learn to associate arbitrary signs with meaning, to generate new symbols with new meanings, and to use these signs to communicate simple statements, requests, and questions. They can refer to objects and events displaced in time and space; classify novel objects into appropriate semantic categories (Ristau & Robbins,1982). Like that of children, their sign use undergoes progressive decontextualization of meaning. They can even transmit their knowledge to peers and offspring (Fouts, Hirsch, & Fouts, 1982; cf. traditional transmission, Hockett, 1960). Symbol order does not appear to be salient to them (except in paniscus). Most great apes do not use symbols spontaneously, in casual conversation (except paniscus and possibly Koko). However, a few novel word combinations have been reported, primarily for Koko, a female gorilla taught American Sign Language (ASL). On separate occasions, Koko has called a stale cake cookie rock, a face mask eye hat, and a ring finger bracelet, which is the Chinese translation of their term for ring (Robbins & Ristau, 1982). In wild and captivity, many apes show tool use, invention of new means to solve problems, and purposeful deception of other group members. Overall, great apes demonstrate the intellectual achievements of 2- to 3-year-old children (i.e., symbolic subperiod of the preoperations period, Parker & Gibson, 1979). It appears that selection has favored advanced sensorimotor and even early symbolic intelligence in great apes (Ristau & Robbins,1982).
Terrace (1985) argues that common chimpanzees (Pan troglodytes) can use symbols as names or transmit information with them. This may likely be a result of training methods. If troglodytes first learn receptive capacity of symbols, as human children do, they might demonstrate greater productive capacities. Nim's tendency to imitate in producing strings of signs is probably a result of how he was trained. Terrace's criticism actually implicates his own particular methodology; in essence, he is criticizing the execution of his experimental design.
Some methodological problems of earlier studies were resolved with Kanzi's instruction, or lack thereof (Savage-Rumbaugh, Romski, Hopkins, & Sevcik,1988). Common chimpanzees (Pan troglodytes) had been initially taught to ask for an item by name (demonstrating productive skills), but they would not then retrieve these same items if someone else asked for it by name. They showed no evidence of comprehending the referential function of the lexigram symbols until they required specific instruction in reception skills. Kanzi was first introduced to lexigrams at six months when his mother Matata was being instructed in their use. He spontaneously began to use symbols for the purpose of communicating with no explicit training. Communicative value of symbols was apparently understood by him from the outset (Savage-Rumbaugh, Rumbaugh, & McDonald, 1985). When he use a symbol to request an apple, for example, if a banana was offered to him instead, he would reject it. He demonstrated behavioral concordance with his symbol usage. Kanzi can also use photographs of food as if they were lexigrams: he can name them, give the pictured food if asked, and use the photograph in requests. Naming, giving, and requesting required separate training in troglodytes studied by experimenters. Kanzi uses symbols mostly to direct attention to places and things not visible, or to activities in which he is currently not engaged in. In fact, twenty or more seconds may elapse between his request to travel to a particular food site and the time he arrives there. But he keeps his destination in mind at the time. Sherman and Austin required five to six years of training before they could represent absent objects or events. With comparable vocabularies of 90 symbols, Kanzi (paniscus) constructed as many utterances as Sherman (troglodyte), but included a wider range of communicative functions and wider range of topics (Savage Rumbaugh et al.,1988). Sherman's utterances were primarily task-elicited by the teacher, concerned with only food, actions, and locations. Kanzi, on the other hand, made comments, requests, and statements, about affilitiation, objects, agents. Savage Rumbaugh et al. (1988) conclude that symbol usage in paniscus was characterized by "spontaneity and freedom from contextual constraint not seen in troglodytes".
Comprehension of spoken English was surprisingly high in paniscus. Troglodytes understood a very limited amount of words, possibly none (tests may have involved paralinguistic cues.) Kanzi appeared to understand an entire sentence on occasion, including completely novel requests. Formal tests of matching uttered word to corresponding lexigram were given to Kanzi at 5 1/2 and Mulika at 2 1/2 (using 3 different experimenters with different accents). With cueing controlled, both paniscus demonstrated that they could associate a graphic symbol to the spoken word without being able to pronounce the words (which is not unlike young children). They possessed a double system of representation. At two-and-a-half years of age, Mulika was already able to link photographs, lexigrams, and spoken words for 77 items. Results such as these suggest that humans have underestimated the intelligence of infant ape in earlier studies. To reduce paralinguistic cues, a (Votrax) speech synthesizer was used in testing Kanzi ,who had had on incidental previous exposure to the synthesizer. With no advance training on this specific test, Kanzi was able to correctly identify 100 words, compared to 150 words spoken naturally. (Normal four-year-old child without exposure to the synthesizer were only able to identify 33 out of a possible 150 words.)
Vocal repertoires of two chimp species are also distinct. Also, in response to vocal comments or to questions directed to him by humans, Kanzi will occasionally produce sounds which are atypical of his species. Savage-Rumbaugh et al (1988) have recorded five vocalizations which are unlike any recorded from other paniscus. Kanzi's mother lived with him all the time (except for 4 months when she was away) so these unique vocalizations cannot be attributed to the lack of an appropriate model. Kanzi could select a photograph in response to a spoken word, select a photograph when shown a lexigram, and select a lexigram in response to a spoken English word. Out of a total of 204 stimulus and sample trials, Kanzi made only one error (Savage-Rumbaugh, Rumbaugh, & McDonald, 1985). Savage-Rumbaugh et al. (1988) concluded that, compared to troglodytes, paniscus possess a greater capacity for symbolization and for associating auditory stimuli to representational processes which may be used in the wild. Paniscus appears to have a considerably more complex social life: they establish elaborate male-female relationships, the unit group consists of larger, more stable subgroups, and the attack behavior in this species demonstrate a remarkable degree of coordination among a larger number of individuals of both sexes. Humphreys (1986) and Seyfarth & Cheney (1990) have separately advanced theories which propose that primate intelligence is driven by social interactions, as opposed to environmental variables (particularly in hominids). Kanzi's most complex symbol communication occurred in social interactions which involved three or more individuals, which provides some support for these theories. Of interest, retarded children instructed in a similar symbol system eventually demonstrate competencies beyond paniscus, most marked by increases in interpretable vocalizations. Also, it would be interesting to test if the sound of English words or image of lexigrams were used in Kanzi's mental representation, using tests which are sensitive to the impairment effect (Premack, 1985).
Imitation plays an important role in language use, in humans as in chimpanzees. Studies (e.g., Terrace, 1985) which have criticized language use in signing chimpanzees on account of signs of imitation have ignored the pragmatic functions served by imitation in a conversation. As a human child, imitation by Kanzi and Mulika were not random but selective and served a variety of conversational purposes in context. Between 30 - 47 months of age, only 11% of Kanzi utterances were immediate imitations, compared to 39.1% of Nim's utterances (Greenfield & Savage-Rumbaugh, 1990). Most of Kanzi's immediate imitations occurred while he was learning a particular lexigram and once it was learned, he rarely responded with immediate imitations (Savage-Rumbaugh, Rumbaugh, & McDonald, 1985). Kanzi's present rate of 6% immediate imitation is similar to children up to age three. Mulika at 21% is comparable to children between one and two. Troglodytes also use repetition to serve conversational/pragmatic functions (Greenfield & Savage-Rumbaugh,1984). Nim used partial imitations to construct longer utterances. Unlike Nim, longer utterances made by Kanzi provided additional information. Two or three-symbol combinations usually indicated individuals other than himself as the beneficiary of actions (36% of three-word combinations, at 30-47 months of age) (Savage-Rumbaugh, Rumbaugh, & McDonald, 1985). Nim always assumed he was the beneficiary of actions. (This may be a species limitation or a product of his education. The representational aspects of symbols may be attained through receptive training; the communicative aspects through productive training -- may indicate that learning the representational aspects first leads to different assumption about social interaction, self concept, etc. Intracommunicative symbols are extended into the domain of intercommunication symbols). Savage-Rumbaugh, Rumbaugh, and McDonald (1985) conclude that paniscus possesses a flexible vocal repertoire as well as other primitive language skills, but these skills improve at slower rate than seen in normal child. These findings suggest that before human evolved the capacity to speak they already possessed the capacity to understand speech.
In terms of syntactical processing, early studies of language-instruction in chimpanzees were very disappointing. Chimpanzees could learn to combine two or more symbols in a non-random fashion, but they lacked syntactical comprehension. They often relied on imitation when combining symbols and they often did not appear to comprehend the semantic relationships of individual symbols nor could they extend these combinations to categories of symbols (the basis of syntax). Greenfield and Savage-Rumbaugh (1984) could find no tendency by troglodytes to use consistent symbol order, either syntactically or pragmatically. Needless to say, at the time, no chimpanzees studied had developed or invented their on rules of symbol combination. Kanzi's began to use symbol order to convey particular semantic relationships: he created his own syntax. This emergent syntax involved relations between categories in which each category was composed of different lexical items. For his two-element combinations -- lexigram-lexigram, or lexigram-gesture -- Kanzi learned the rule: action precedes object from environmental models. For instance, he described his own impending action using the statement "HIDE PEANUT." (Of note, troglodytes did not make statements, but they comprise 4% of all Kanzi's utterances.) This symbol order was not an artefact of partial imitation or position preferences for specific lexical items. Semantic miscomprehensions which involved utterances which followed this rule were enlightening. They showed that a lexigram's position in an utterance determined its function. For example, the utterance "HIDE ICE" was made by a human to comment (incorrectly) on a big block of ice that someone was hiding in the ice. Kanzi began searching under some nearby blankets, looking for ice. On his own Kanzi also invented rules for positioning lexigram. They were place gesture last; that is, gestures were made after lexigram(s), which was the opposite ordering strategy employed by his human models. The gesture-after-lexigram rule was not a matter of physical convenience. Greenfield & Savage-Rumbaugh (1990) present an example when Kanzi left a person to go to the lexigram board then returned to the person to gesture. This has the appearance of a formal arbitrary rule. Kanzi invented another syntactical or generative rule: symbol order reflects action order. When he combined two action lexigrams, he did so in order of the behavioral sequence. He had no model for these constructions. Three lexigram (e.g., Action-action-agent) utterances maintained the ordering rules of two-element combinations: two lexigrams (first action, second action in sequence), followed by a gesture indicating agent. In terms of productivity, Kanzi used symbols in a variety of different semantic relations. For example, he used the word "AUSTIN" in twelve different semantic relations: action-agent, action-goal, action-object, affirmation-goal, attribute-entity, conjoined locations, entity-demonstrative, entity-location, goal-agent, goal-instrument, recipient-object (Greenfield & Savage-Rumbaugh,1990). Symbol order is used by Kanzi in his lexigram-lexigram combinations to depict differences in meaning. He tends to place animate beings first when they function as agents and last when they function as objects of actions. For example, "GRAB MATATA" means Matata was grabbed, while "MATATA BITE" means Matata was bitten. Length of combinations were mostly two elements only (90% at time of study, four years of experience). A short utterance length in Kanzi probably does not indicate limited intellectual capacities as much as reflect the number of dimensions of perceived variability in a particular situations (as with children at two-word stage, Greenfield & Savage-Rumbaugh, 1984). Also, his experimenters used few lexigrams (usually one or two) when speaking propositions to him.
Rumbaugh (1988) concludes that chimpanzees have the requisite capacities for mastering several dimensions of language. They may not be natural users of human language, as he puts it (oxymoronically), but chimpanzees demonstrate a capacity for basic semantics and relational information. Like humans, they are oriented toward the detection and/or assertion of cause-and-effect relationships, with the potential to construct and use high-order relational processes to help understand the behavior of fellow apes (and humans, when in captivity). Kanzi understood specific English words produced by human speech, recorded speech, or digitized speech, and demonstrated comprehension of words by readily being able to "translate" between English and lexigram word-symbols. In short, Rumbaugh (1988) points out that Kanzi has learned more of human language than humans have learned of paniscus communications. Premack (1985) argues that creating sentences is not one of the more demanding uses of language; conversing is. Terrace (1985) also makes the point that humans frequently communicate without any apparent attempt to achieve some goal. But considering that all communication occurs within a social context, language, like any form of communication, probably always serves some social purposes, such as attempts to persuade or to change one's social status. The pleasure of naming by children so often remarked upon by Terrace and others probably marks the child's persistent attempts to demonstrate its cognitive, and consequently social, power (cf. Bateson, 1966). There remains one potential change that might be useful for ape studies: phonemes and letters are the basic units in human language, while words or semantic concepts are the basic units chimpanzee utterances (Healy,1980). However, Kanzi does indicate some comprehension about the duality of patterns. Many of the lexigrams used by Kanzi consist of similar features such as circles and wavy lines. And Kanzi is able to understanding many spoken English words, which consists of a limited number of phonemes.
Until recently, animal vocalizations were assumed to be largely indexical -- concerned with the signaler and his subsequent behavior only. Vocalizations of monkeys and apes were assumed to be non-referential, hard-wired, involuntary behaviors, highly tied to the emotional state of the caller (Malmi, 1980). Marler (1976) hypothesized that a crucial step in the evolution of language occurred when early humans were able to perceive graded vocal signals in a discrete manner. However, monkeys (Japanese macaques) appear to perceive a conspecific's vocalizations as consisting of discrete elements (Petersen, 1982). Human language involves an integrated system consisting of numerous elements and many levels of analysis (i.e., syntactical, phonemic, semantic, pragmatic, morphemic). Similarly, comprehension of conspecific vocalizations may require many levels of analysis to encode specific messages with information about the animal's identity, emotional state, sex, age, and species or group membership. Preadaptations for these forms of analysis and potential precursors of elements of this system may be found in the behaviors of primate and other mammals (e.g., Petersen, 1982; Herman, 1986). Petersen (1982) has examined the perceptual phenomena which characterizes of human speech perception: (1) neural lateralization in speech-perception processes; (2) categorical versus continuous perception of acoustic dimensions cueing distinctions among speech sounds; (3) selective attention to linguistically distinctive acoustic features; (4) interactions among different processing stages in the speech perception network; and (5) short-term memory involvement in processing different aspects of speech.
Evidence from neuroanatomy, electrophysiology, and psychology point to a predominant role for the left cerebral hemisphere in managing human speech. Peterson, Beecher, Zoloth, Moody, & Steebins (1978) report a right-ear advantage (REA) for macaques in the discrimination of monaurally presented, species-specific vocalizations. In addition, five of six fuscata and one of five controls (other species) showed a right-ear advantage (REA) for discriminating peak cue (the apparent communicative element of macaque coos) monoaurally presented competing with wideband noise in other ear; and one of two fuscata tested showed a strong LEA for pitch discrimination (this macaque had earlier shown a REA in the peak-cue discrimination task). Dewson (1977) found a more severe impairment of monkey auditory memory following ablation of the left superior temporal gyres than when the corresponding right-lobe tissue was destroyed.
Categorical perception of speech (e.g., voiced vs unvoiced distinction) is found in many mammals, including chinchillas and minks (Kuhl & Miller,1975) Snowdon (1982) reports that the communicative trills of pygmy marmosets are perceived categorically while "paralinguistic" trills perceived continuously. Chimpanzees demonstrate "phonemic-like" variation in certain calls. Whimpers occur in several distinct forms associated with different situations (Goodall, 1986). Also, a fundamental property of speech perception is phonetic constancy: the ability to recognize the phonetic identity of different versions of the same phoneme despite marked variation in acoustic features not relevant to the phonetic identity of the utterance. For instance, when the same phonetic segment is embedded in different contexts (e.g., produced by speakers of different sex, age, or vocal-tract anatomy), its basic acoustic characteristics are altered dramatically, but a phonetic /p/ is still perceived as a /p/ nonetheless. In other primates, discrimination of specific peak or pitch characteristics in macaque calls by many different callers was studied. The rate of acquisition showed that fuscata acquired peak discrimination very quickly, but had serious difficulty with the pitch task; whereas two comparison non-fuscata monkeys showed the opposite latency pattern. Similar perceptual constancy tasks can help distinguish the communicative cues in nonhuman vocalizations from those used more paralinguistically. No studies have yet investigated the fourth and fifth phenomena occur in nonhuman primates.
Evidence that nonhuman primate vocalizations are also modifiable (learnable) like human vocalizations comes from Pola and Snowdon (1975). In one group of pygmy marmosets, twins were born to parents who died soon after. Normal pygmy marmoset infants progress from infant babbling through juvenile stages to essentially an adultlike vocal behavior within five months. In contrast, the parentless twins displayed a much slower rate of development, not reaching the juvenile stage until the age of one, and throughout their lives they never attained a fully adult vocal repertoire. They described this as a "Piagetian-like" progression of vocal development that parallelled the nature of social relationships between an infant and its parent. In the stump-tailed macaque, Lillehei and Snowdon (1978) report a case in which the mother of an older infant died, and the infant regressed to an earlier stage of social dependency and began nursing from an older sister. During this period he also regressed toward a simpler structure in his contact calls rather than maintaining the complex individual coos characteristic of older infants.
Vervet calls provide much insight into the relationship between semantic and associated vocal behaviors. Vervet alarm calls appear to be intention in that vervets can modify their alarm calling rate depending on their audience. For example, upon seeing a predator (a grad student), four females give alarm calls normally given for Maasai tribesmen at significantly higher rates when they were encompanied with their offspring than when they were with unrelated juveniles. In another experiment, subordinate adult males were locked outside in the company of either a female or a dominant male. In all four cases, the males gave more alarm calls when they were with a female than when they were with the dominant male (Cheney & Seyfarth, 1990).Vervet calls and their associations are not constructed by means of onomatopoeia. Leopard alarm calls, for instance, do not sound like the noises made by leopards nor any incidental sounds made during a vervet's escape. Calls are caused by factors other than arousal. Responses that vervet monkeys give to alarm calls are not based on a call's length or amplitude. Alarm calls may be given in more relaxed circumstances and their production can be modified depending on characteristics of audience, both contrary to an arousal hypothesis. Habituation experiments assist categorization of sounds. Depending on vervets' responses, one can determine which sounds are grouped together into common categories by the vervets. Habituation experiments show that vervet calls have referents external to signaler, and that animals make judgments about the meaning of calls on the basis of these referents. Intergroup wrrs and chutters are used in habituation experiments. Although these calls have the same broad referent (another vervet group), they are given in slightly different behavioral contexts: wrrs usually occur when another group has first been spotted; chutters are typically given during an actual fight between members of both groups. The two calls are associated with rather different behavioral responses, but they are treated by the vervets as being similar. Having habituating to one individual's wrr, the monkeys will ignore the same individual's chutter. This is difficult to explain unless wrrs and chutters have the same general referent (Cheney. & Seyfarth, 1990). Further evidence that calls have external referents come from the use of calls in reaction to dogs by different groups in the wild. In the Cameroon savana, vervet are often attacked by feral dogs and they will give a leopard call and climb into a tree when one is spotted. In forests, however, where hunters use dogs and can easily shoot vervets that escape into trees, vervets give short, quiet calls which results in other vervets fleeing into thick brush where the humans can't follow (Cheney & Seyfarth, 1990). Vervets also produce a minor mammalian predator alarm (to jackals, hyenas, lions, and cheetahs) , an unfamiliar human call (to tribesmen) - Maasai), chutters to an observer (once they are accustomed to him), and a baboon call.In terms of evolutionary function, Cheney and Seyfarth (1990) believes that one can easily explain why vervets have so many grunts (given they hear calls as discrete entities), but they cannot explain why vervets have so few.
Human speech and primate vocalizations share many morphological complexities and organization properties. For instance, referential, denotative capacities of animal communication also function in ways other than as alarms. They can be involved in: (1) food location and identification (found in ants, bees, dolphins); (2) predator identification (vervet monkeys, birds); (3) recruitment for assembly at resource sites (wolves, dolphins, birds, bees, ants); (4) coordination of labor (ants, termites, and possibly dolphins); (5) resource prediction; (6) kin identification (possible in dolphins in terms of signature whistles); and (7) status transformation (also possible in dolphins). Human language can perform all seven of these functions. The natural vocalization of nonhuman primates also demonstrate, according to one's definition, some processing analogous to syntax. Snowdon, (1982) defines syntax as any system of rules that allows us to predict sequences of signals. For example, when chirps and whistles are combined by cotton-top tamarins, chirp elements always precede whistle segments and there is a decreasing center frequency across successive elements. Marler (1977) specifies two types of syntax which are possible in communication. "Phonetic" syntax is analogous to the formation of different words through the rearrangement of phonemes. The resulting word has a different meaning and function than do the individual components. "Lexical" syntax is analogous to the combining of individual words into phrases so that the resulting structure is the sum of the meanings of the individual components. Marler argued that instances of phonetic syntax may be common in nonhuman animals, but instances of lexical syntax in animals would be very rare (however, cf. Snowdon, 1982, alerting chirps of pygmy marmosets). Parisi (1983) argues that a syntactic system emerged relatively late in hominid evolution and a fully open lexical and syntactic system was preceded by a semiopen lexical system without syntax. The syntactic system emerges relatively late in human ontogeny. Great apes, young children, speakers of pidgins, and certain aphasics possess semiopen syntactical systems. Parker (1985) notes that there is a natural ordering observed in sign languages, child languages, ideograms, and casual drawings, which reflects the temporal development of the visual event as determined by cognitive and perceptual constraints, and subsequently would not limited to our species. And as with human language, many species may acquire certain vocalizations and vocal traits via protocultural influences (e.g., vocal dialects in Japanese macaques; acquisition of predator alarm calls by vervets; signature whistles of conspecifics in Orcas). There is a richness and complexity of primate vocal behavior, only recently detected.
Human evolution is characterized by more than the development of language. Dramatic shifts in diet, locomotion, social organization, communication, technology, and the brain also took place, resulting in specializations such as pedagogy, social attribution, and consciousness. Given my account of the evolution of language, these developments seem to be a natural continuation of the integration of communicative and representational system. The attribution of mind in others would develop as elements of the natural communications became more involve in one's mental experiences. Dennett (1987) has developed a useful method for investigating communication and the attribution of mental states in nonhuman species. First we assume that a vervet monkey is an intentional system, capable of mental states like beliefs and desires. Next we attempt to determine the level of intentionality of this system. Zero-order intentional systems have no beliefs or desires at all. In this case, a vervet makes a specific alarm calls because it is frightened and aroused; different predators evoke different fears and each fear elicits a characteristic alarm call and a characteristic escape response. Alternatively, a first-order intentional system has beliefs and desires, but it doesn't have beliefs about beliefs (metacognition). In this case, a vervet gives a leopard alarm call because it believes that there is a leopard nearby or because it wants others to run into trees. The vervet has no conception of his audience's mind, nor does it have the ability to makes the distinction between his own and another animal's beliefs. Alternatively, a second-order intentional system has some conception about both his own and other individuals' states of mind. He gives a leopard alarm call because he wants other vervets to believe that there is a leopard nearby. A third-order intentional system makes an alarm call because he wants other vervets to believe that he wants them to run into trees, and so forth. Linguistic communication may require at least third-order intentionality on the part of speaker and listener (Cheney & Seyfarth, 1990); but this might be an overestimation (e.g., cultures in which mental attribution of others is unwelcomed; cf. Jaynes, 1976b). According to Grice (1988), reciprocal altruism and the detection of cheaters that it implies was responsible for the rapid evolution of cognitive capacities exhibited by higher primates. A third-order intentionality is necessary to detect cheaters and to gain advantages as a cheater (e.g., the recipient wants the altruist to believe that he intends to reciprocate).
Other evidence suggests that monkeys possess incomplete theories of mind, if any at all. For example, vervet monkeys will continue to give alarm calls regardless of whether their audience is already aware of danger (i.e., long after everyone has seen the predator). Monkeys often use third parties as social tools, while apes rarely do (Whiten & Byrne, 1988). For example, a baboon may feign injury or insult from another baboon in order to get his alliance partner to attack the supposed attacker. Chimpanzees, on the other hand, may be able to attribute motives to others and therefore recognize that they won't be able to recruit allies on a regular basis using such ploys.
Cheney & Seyfarth (1990) performed a series of ingenious experiments to detect mental attribution in vervet monkeys. A subordinate female and the juvenile offspring of a dominant female were placed in one cell. The dominant female sat in an adjacent cell behind either glass, steel, or a one-way mirror. The presence and knowledge of a dominant female was dissociated in order to determine their individual effects on the behavior of a subordinate older female. In the glass condition, the subordinate showed little agonistic behavior toward the juvenile; in the opaque condition (i.e., the mother was hidden behind barrier), she behaved more agonistically toward the juvenile. Conversely, the juvenile showed more agonistic behavior toward the subordinate female when its mother was visible than when she was not present. Next, the subordinate female and the dominant's juvenile were allowed to familiarize themselves with the properties of one way mirrors (i.e., one was placed in their cell for four weeks). They found that significantly more juveniles showed agonistic behavior in the glass condition than in either the mirror or opaque conditions. In contrast, significantly more subordinate females showed agonistic behavior in the mirror and opaque conditions than in the glass condition. The subordinate also more often avoided the juvenile in the glass condition than in either the mirror or opaque condition. Cheney & Seyfarth (1990) conclude that either this demonstrated a theory of mind on part of the vervet or that vervets are adept at monitoring the observers' apparent attentiveness; subject may have been sensitive to the dominant female's actions, orientation, and direction of gaze.
Interesting case of mental attribution/higher-order intentionality involved a low-ranking vervet monkey named Kitui reported by Cheney & Seyfarth (1990). He would make false leopard calls when new males attempted to joined his group. He was always the lowest rank in his group and would be beneath new interloper if the interloper was able to join the group. So when a new interloper appear close to being accepted into the group, Kitui would make a series of false leopard calls. The interloper would run away and consequently not join Kitui's group. As if to convince his rival of the full import of his calls, Cheney & Seyfarth (1990) report two occasions when Kitui left his own tree and walked across the open plain and entered a tree adjacent to the interloper, alarm calling all the while. Obvious his theory of mind was lacking.
Deception implies that one has attributed a mind to another (one which can be deceived). Deception by means of information concealment is described by Goodall (1986). On one occasion, a 9-year-old chimp, Figan, gave a loud food calls when he was given bunch of bananas. Consequently, the whole group heard the cries and converged on his site, leaving few bananas for him. The next time Figan was given a bunch of bananas, he remained silent (though Goodall reports hearing faint choking sounds in throat) and ate bananas undisturbed. Woodruff and Premack (1979) provides the best example of active deception. In one experiment, a common chimpanzee is shown where in two containers the experimenter has hidden some food. Following this, either a cooperative trainer (who when showed the food by the chimpanzee always shares it) or an uncooperative trainer (who when shown the location of the food, always eats it himself enters the area. The chimpanzee always provides correct information as to the location of the hidden food to the cooperative trainer, however, he acts differently with the uncooperative trainer. First, the chimpanzee withholds information: turns his back and sits motionless so as not to cue trainer to where the food was hidden. Later, after many more trials, some of the chimpanzees will attempt to deceive the trainer and gesture or point to the wrong container. In the wild, chimpanzees often act deceptively to hide their fear-smiles from dominant apes, either by turning away, or pressing their lips together or by covering their mouth with a hand.
Pedagogy or teaching may be a measure of mental attribution. In order to teach correct information to another, one must be attribute false information to that individual. In most mammals, teaching is stereotypic and is not sensitive to a particular audience's ignorance. There are some anecdotes of teaching by chimpanzees (e.g., Washoe was observed molding the hands of her adopted's son for signs, Fouts, Hirsch, & Fouts, 1982). However, chimpanzees do not usually inform others in ignorance of information which they possess, nor do chimpanzees show empathy for grieving animals (Premack, 1986). Premack (1988) interprets their behaviors as if each chimpanzee can create a desired state of mind in others, but they are less adept at recognizing situations when other chimpanzees have beliefs which are incompatible with their own. There is no true negative commands which would be indicative of a different belief in another. The closest approximation of negative commands in monkeys and apes are threat vocalizations. The apparent lack of positive commands in nonhuman primates is stronger evidence that they have difficulty ascribing different beliefs to other minds. The lack of pedagogy among nonhuman primates raises important questions about the animals' ability to attribute states of mind to other
Premack and Woodruff (1978) tested mental attribution in Sarah, a common chimpanzee. They had her view videotapes of human trainers having difficulty trying to solve a variety of problems. For instance, one trainer tried to operate a record player which was unplugged. After each videotape, Sarah was given several photographs, one of which depicted the correct solution to the problem. (a cord plugged in, a cord not plugged in, a cord cut) She consistently chose correct photograph for her favorite trainer. However, for her less favorite trainer, she often chose incorrectly.
Monkeys do seem adept at recognizing each other's social relationships (e.g., matrilineal kind, dominance ranks, and friendships), and adjust behaviors according to who has reciprocated in past, and can predict consequences of their own behavior on the behavior of others. However, all of these abilities do not require attribution of mental states (Gallup, 1982). Individuals who cooperate to solve a problem must recognize each other's aims and purposes in order to arrive at a common goal and consequently they must attribute some beliefs and intentions to others. Free-ranging chimpanzees frequently help each other into trees (Goodall, 1986). Captive chimpanzees hold ladders for each other to climb forbidden trees or to escape from the compound. de Waal (1989) describes a spontaneous game with pygmy chimpanzees in the San Diego Zoo in which individuals deliberately stranded each other in a dry moat by pulling up a chain rope that led down into the moat. Other chimpanzees would then "rescue" their companions by dropping the chain rope back down to them. These apes must have partly understood the purpose of the chain and the need of trapped animal. Cheney & Seyfarth (1990) make the point that there are no examples in primate behavior that cannot be explained except in terms of a theory of mind (see Table 4), which may explain the rise of behaviorism in human psychology.
TABLE 4: Evidence of MIND From Gallup (1985)
| trait | hard-wired analog | self-aware instance of mind |
|---|---|---|
| attribution | unlearned reactions to conspecific threat postures and predators | attribution of intent, and responsibility; anthropomorphism |
| deception | mimicry | intentional distortion &/or withholding of information |
| reciprocal | alarm calls(?) | reciprocal aid giving; altruism, selectively withholding aid from cheaters and stealers |
| empathy | responses to appeasement gestures and infant distress calls | providing solace to injured conspecifics |
| pretending | injury & death feigning | certain forms of deception; multiple representation of event |
Self-awareness is also characteristic of human mental experience and is closely related to the attribution of mind in others (Jaynes, 1976b). Evidence of self-aware behaviors are few but interesting in apes. Kanzi sometimes signed bad before doing something for which he would subsequently be punished (Savage-Rumbaugh et al., 1988). That a chimpanzee can refer to his own name (in solving tests and answering questions) implies some degree self-awareness. Gallup developed a mirror test to detect self-awareness in nonhuman primates. An animal is anesthetize and a mark is placed on his forehead. When he awakens, he is shown his reflection in a mirror. If he react to the mark (rubs or presses it), Gallup (1982) argued that this demonstrated self-awareness. Gorillas fail the mirror test, as do children below the age of two (Gallup, 1982). It could be that gorillas are not interested in superimposed body marks and/or lack motivation to respond. Monkeys can use to learn mirrors to manipulate objects and monitor behavior of others, but they also fail the test. The mirror test may indicate bodily awareness or representation more than conceptualization of a self. Pretend play in younger monkeys and apes indicates some self-recognition and self-awareness in that juveniles must be able to distinguish what is real and what is imagined; they must entertain multiple representations of an object or event at the same time.
Tool use has been implicated in the development of human intellect (e.g., Reynolds, 1982). Goodall (1986) reported tool use in wild chimpanzees (troglodytes) in three main contexts: (1) threatening or attacking intruders (branches & twigs dropped; gorillas throw vegetation); (2) bodily care (chimps wipe blood and feces from hair with leaves; and (3) acquisition and preparation of food (chimps use up to five types of tools to obtain termites, ants, and honey, and break open intractable nuts. In captivity, chimpanzees are known for constructing ladders (de Waal, 1983). Human tool use exhibits constructional advances over those of the apes, however. Humans make tools composed of numerous subcomponents, tools which required advanced visualization of complex three-dimensional shapes and of the behavioral sequences needed to produce these shapes. Humans also can use several tools simultaneously or in succession to achieve a single end. Apes can use several types of tools, but they do not use them in concert, as part of one tool-using sequence, nor do they construct tools of varied components of complex shapes (Gibson, 1981).
It has been hypothesized that primate intelligence originally evolved to solve social problems and was only later extended to problems outside the social domain. (Humphreys 1976). Cheney (1986) argues that natural selection for intelligence may have acted strongest in the social domain. Group life exerted strong selective pressures during primate evolution on the ability to form complex associations, reason by analogy, and make inferences and predictions concerning the behavior of fellow group members. Two concepts underlie intelligent behaviors in primates and other species: (1) the ability to recategorize information, and (2) the capacity to prioritize concepts, to organize concepts hierarchically. Both of these abilities probably developed in primates to solve social problems. Successful behavior in a social group requires one to prioritize individuals as tools, threats, etc. And one attains many perspectives of other individuals in a group through different encounters. A social entity is different from inanimate objects or even other animals, prey or predator, in that its function changes over time. The capability to recategorize the function, threat, affiliations, of other primates in one's group are extremely valuable. Much of this intelligence seems to be directed solely on solving social problems. Despite wide exposure to the behaviors of predators, vervets do not appear to understand that fresh python tracks or a recently killed carcass implies that a predator's proximity. Cheney & Seyfarth (1990) observed baboons inspecting a dead carcass without concern, until one happened to spot a lion and barked alarm calls and fled. They ask the question: is it evolutionarily more useful in terms of survival to develop a conceptual understanding of social relationships than to recognize the track of a python? Vervets and other primates have adaptive specialization in the domain of social behavior that are not extended to interactions with other species, what some have called "laser-beam" intelligence: abilities which are extraordinarily powerful when focused in a single domain but much less developed outside this narrow sphere. The social skills of human, as evidence by the human ability to maintain alliances despite rare contact. Tursiops demonstrate similar intergroup alliances of dyads and triads who cooperate to herd females away from rivals (J. Mann, pers. comm.). These alliances may be more complex than those found in apes and monkeys.
Throughout the years, many scientists have claimed that the development of language influences cognition (e.g., Sapir, Whorf, Bloom). Language is claimed to enhance cognitive abilities, sharpening existing concepts, increasing the type of information which can be categorized. While this possibility is difficult to test in humans, studies of language instruction in animals provide an opportunity to address this question. Premack (1985) concludes that apes exposed to language training are capable of solving problems that nonlanguage-trained animals cannot solve. Language-trained chimpanzees are able to go beyond physical similarity and compared objects on more conceptual grounds. For example, all apes can match half-an-apple to half-an-apple; but only language-trained appear able to match half-an-apple to half-a-cylinder of water. All apes can match X to X (rather than to Y), but language-trained apes can match samples such as XX to YY (vs PQ) and PQ to XY (vs LL). Language-trained chimpanzees can also perform analogies. Premack (1985) notes that there is no control for these studies. One needs to compare cognitive abilities before and after training.
Field observations demonstrate a degree of social organization and cultural transmission that is difficult to explain without some level of language-like communication, though our understanding of their specific communication is not strong (Goodall,1986). It may involve a whole-body gestural system. Lieberman, (1981) sums up many investigators when he says: "chimpanzees raised in laboratory environments thus demonstrate both cognitive and linguistic ability that is qualitatively similar to that of humans, though at reduced levels." In some ways, we may have looked at the problem from the wrong perspective (Homo as most evolved). An interesting hypothesis posits that the capacities of a common ancestor in the Homo-Pan clade has been retained in pygmy chimpanzees, elaborated in Homo, and underwent evolutionary change and specialization in common chimpanzees. Humans and pygmy chimpanzees may have a more "primitive" left hemisphere, in that in certain functional characteristics it is more similar to the ancestral pongid. However, the great encephalization of Homo sapiens sapiens after its split from anthropoids, compared to Pan paniscus, remains unexplained. As stated above, large brains are only able to evolve when the benefits exceed the high costs. And the ancestral species possess evolved metabolic and nutritional systems adequate to support it. The ancestral species must have been successful in its own right. In terms of humans, this suggest that ancestral anthropoids already possessed complex mental representations and sophisticated social behaviors. But what factors allowed ancestral cetacean species to be so successful?
The largest brain ever to appear on the planet belongs to the sperm whale (Physeter catodon), a member of the Cetacean Order, whose brain can weigh up to 9200 g, with an average of 7818 g, (Ridgway, 1986). Lilly (1967) suggests, solely on brain weight, that the sperm whale possesses a higher development of consciousness than humans, despite a relatively low brain-to-body weight ratio of almost 5000:1. Cetaceans are secondary aquatic mammals who, with the sirenians, are the only terrestrial mammals that have made a complete transition back to living in the sea. It is speculated that dolphins evolved from primitive ungulates or perhaps carnivores which foraged in coastal waters until finally adapting to a complete life in the sea between 50 to 70 million years ago (Early Paleocene). Recognizable cetaceans, with dorsal movement of the nares, modified dentition, and the development of the modern odontocete skull (plan), appear in the fossil record as early as the Middle Eocene (Gaskin, 1982). They are a very successful species: cetaceans species occupy nearly every aquatic subzones on the planet. Encephalization appears to have developed rapidly in this Order; as early as 15 - 25 million years ago cetaceans surpassed human levels of brain weight (1400+ g).
Much of what we know about cetacean neuroanatomy and neurophysiology comes from research performed on the Atlantic bottlenose dolphin (Tursiops truncatus). The Tursiops demonstrates great flexibility in behavior, vocal as well as physical, and is believed by many to be the most intelligent cetacean. The resemblance of the dolphin brain to the human brain is obvious and may be unsettling at first sight (see Figure 1 and 2). The great size and vast convolutions of the cetacean brain were noted as early as the late seventeenth century (cf. John Ray, 1671). Tursiops has an average brain size of 1587 g, Lagenorhynchus (white-beaked and Pacific white-sided dolphin) of 1256 g, while Homo sapiens sapiens average around 1370 - 1400 g (Ridgway, 1986a). Encephalization quotients of some dolphins (which takes into account an animal body size, by using what remains after a regression comparing brain to body size) are nearly double that of apes (Gorilla gorilla, 1.76; Pan troglodytes, 2.48; compared to Tursiops, 4.4; Lagenorhynchus, 4.9) and are higher than any other mammal, except for humans (Homo sapiens sapiens; 7.39-7.79; of note, australopithecines fall in the dolphin range: 3.25 - 4.72). The difference between human EQ and dolphin EQ (~3.2) is too unstable, however, to suggest greater encephalization in humans than in dolphins (Jerison,1986).
The odontocete's brain is more convoluted than any other mammal's, having more surface area per unit of volume than even the highly convoluted brain of humans. One index of cortical folding placed humans (2.86) far below Tursiops (4.47), which is due to a larger mean cortical surface area in Tursiops (3745 cm2) than in humans (2275 cm2) humans (Ridgway, 1986a). But the cetacean cortex is relatively thin, ranging from 1.3 mm to 1.76 mm in thickness; whereas the human cortex is almost twice this (2.9 mm at thickest points). Thus, despite its greater mass, the volume of the dolphin cortex is smaller (but comparable) to that of human's: 560 cc compared to 660 cc. Other indices which place cetaceans near humans (and above apes) are so-called Neocortilization I: the volume of neocortex divided by volume of total cortex, and Neocortilization II: the volume of neocortex divided by volume of telecephalon; Cortilization: the volume of total cortex divided by volume of total brain, places Tursiops below all primates, but this appears to be a result of the exceptionally large cetacean cerebellum, Glezer, Jacobs, & Morgane, 1988).
Some investigators have argued that the increased brain (particularly neocortical) size in marine mammals does not imply the same thing as it does in terrestial mammals (i.e., differential enlargement is not equated with increased behavioral capacity) and a number of arguments have been put forward to explain away the unusually large size of the cetacean brain on noncognitive brain functions. For example, the needs during deep diving (prolonged hypoxia) have been implicated (Wilson, 1933, cited in Ridgway, 1986a), but sirenians and seals also dive for long periods and they possess smaller, less convoluted brains (and a thicker cortex), as do dolphins with relatively smaller brains. It has been suggested that the cetacean brain may be metabolically less active and therefore evolutionarily less "expensive" (and subsequently less intelligent). However, dolphins exhibit equal or higher metabolic rates than terrestrial mammals of similar size. Finally, echolocation may require more neural tissue to function equivalently to other sensory systems. But then one must account for the successful echolocation of the very small and primitive brain of echolocating bats. Much of the dolphin brain is dedicated to processing acoustical information. A larger proportion of the diencephalon, for instance, is involved in audition compared to primates. Human language reflects a refinement of the acoustic-vocal system in primates. Similarly, dolphin cognitive capacities may be a refinement of an echolocation-based cognitive system.
Functional and cellular organization of the cetacean cortex is unusually conservative, with some progressive features. There has been a relative decrease in the size of the projection zones, coupled with a substantial increased development of the non-projection or "silent" areas of the brain (as in humans and apes). The cetacean brain exhibits a wide range of sizes and EQs. Dolphins with unusually large brain sizes and EQs, such as Tursiops, have shown little change of size of those brain areas responsible for basic biological functions; most of the increases have occurred in silent regions (Ridgway, 1986b). The relative positions of the projection zones, however, resemble the layout found in insectivores and other primitive brains: the primary projection zones are all forward (rostral) in the brain . The auditory regions occupy broad areas of the suprasylvian gyres which adjoin the visual areas in the lateral gyres and dorsal parietal regions. And there is no intervening cortex between the auditory areas and the visual areas, nor between visual-auditory areas and the sensorimotor areas. Other structural boundaries are indistinct and fuzzy, making the brain appear homogeneous and undifferentiated. Yet the surface of the dolphin cortex consists of the (homologous) neocortex to a greater degree than that of human's: 97.9% to 95.9%; archicortex 0.8% to 2.2%, (Glezer, Jacobs, & Morgane, 1988).
The cytoarchitecture of the cetacean cortex is also conservative. It is relatively agranular, or dysgranular; it lacks laminar layer IV except for the presence of an incipient layer IV in certain small areas ("primary" visual and auditory areas). The neocortex is dominated by phylogenetically older layers I and VI, with an accentuated layer II due to layer I inputs; no true "Betz" cells are present, the largest neurons tend to be in the pyramidal cell layers, III and V, and these neurons do not show the wide variations in size compared to terrestrial mammals; and while there is some range of cell sizes, it is not comparable to that found in terrestrial mammals. Overall, the dolphin displays a monotonous, paleo-archicortical type of organization. Columnar organization exists in the visual areas, for instance, but they are larger in area and fewer than half the number of homologous structures in humans. Morgane, Jacobs, and Galaburda (1985) describe cetacean phylogeny as arrested at a paralimbic/parinsular stage of neocortical evolution. Cetaceans returned to the sea before the lastest steps in sensory and motor cortex evolution had been reached (i.e., the emergence of the koniocortices and area gigantopyramidalis that are found in terrestial mammals). Idioadaptations, specific changes in reaction to unique environmental pressures (e.g., an aquatic existence), have occurred and consequently conclusions regarding intelligence in cetaceans would be premature without taking into account behavioral reports. The peculiarly- arranged cortex in dolphin might be able to handle equivalent kinds and amounts of neural information as is handled by the normally layered neocortex of other mammals. For example, many birds demonstrate primate-like vision with markedly different neural organization than that of a humans. This is not to say that fundamental differences in processing will not be discovered, related to these structural differences.
Despite possible structural primitiveness, the cetacean brain demonstrates hemispheric specialization and, notably, hemispheric independence. Dolphins in captivity tend to swim in a counterclockwise direction, which has been suggested to indicate a right hemisphere (RH) dominance or activation in swimming behavior (Ridgway, 1986a). They also have a "viewing eye", usually a right eye preference for viewing strangers and novel stimuli. The optic chiasms are completely crossed in cetaceans so this observation suggests that cetaceans prefer to initially inspection strangers with the left hemisphere (LH). Also, each eye appears to act independently (disconjugatedly). The absence of uncrossed chiasmal fibers implies some degree of independence in the control and use of data from the two eyes (Jerison, 1986b), which may have resulted in highly independent and specialized hemispheres in the dolphin. Morrell-Samuels, Herman, and Bever (1988) report initial findings of cognitive lateralization in Tursiops which support this hypothesis. The procedure for testing laterality in an Atlantic bottlenose dolphin calls for the dolphin to place her rostrum in a padded cup that is locked into either one of two positions. The cups are positioned so that a TV monitor is viewed monocularly by either the left or right eye. All trials are videotaped for later scoring of reaction time by two judges. They report a significant LH advantage for processing 13 complex signs in a human sign language and a significant RH advantage for processing 13 simple gestural commands in a language-trained dolphin. Given the interaction, any eye acuity differences can be dismissed. In a second experiment, ten gestures are shown seven times to each eye. They find that initial presentations of gestural commands result in a right-hemisphere advantage in reaction time which is slowly replaced by a LH advantage with subsequent presentations (as commands become familiar). The reaction time differential indicated a shift from RHA to a LHA over ten trials, on account of progressively more rapid processing speed of the left hemisphere with subsequent presentations, which may indicate a fundamental change in the cognitive processes used by the dolphin during this test of gestural recognition and comprehension. In a related experiment, Morrell-Samuels, Herman, & Pack, (1990) have a female dolphin run through a delayed match-to-sample (MTS) test of four two-dimensional objects of equal familiarity. A pretest shows that the dolphin could perform binocular delayed MTS. They report that object recognition is more accurate when processed by the RH (left eye presentation). Also, recognition latency is faster in the RH. Latency also declines with familiarity of the test in the RH only; whereas latency remains constant or increases in the LH (right eye presentation) throughout the test. Cetaceans also demonstrate the unusual phenomenon of unihemispheric sleep: when one cortical hemisphere is asleep (in stages 1-3) the other hemisphere is always awake, displaying desynchronized EEG patterns, often with one or both open (and reactive, following) eyes, (Mukhametov, Supin, & Polyakova, 1977). The slow waves registered in the thalamus also appear to be unihemispheric. Cetaceans are voluntary breathers and bihemispheric sleep, which can occur in dolphins only under anaesthesia, would likely result in drowning. This is the greatest demonstration of hemispheric independence in the animal kingdom. Also, some dolphins do not have paradoxical (REM) sleep, such as Tursiops or Phocoena phocoena [harbor porpoise], a phylo-genetically older phenomenon, while other do e.g., pilot whale). Mukhametov et al.. (1977) suggest that the evolution of unihemispheric sleep and the devolution of paradoxical sleep may have a common origin.
Dolphins also demonstrate anatomical and neuroanatomical asymmetries. The RH has a larger surface area than the left (1931 cm2 compared to 1814 cm2; Ridgway & Bronson, 1984). The nasal opening is shifted to the left of the cranial midline and in Tursiops the echolocation beam is projected slightly to the left, which may help explain the RH advantage in cortical surface area. Another unusual property of the dolphin brain is its relatively small corpus callosum (surprising in light of unihemispheric sleep and hemispheric independence). Nieto, Nieto, and Pacheco (1976) found that the ratio of callosal area to brain weight in dolphins (Stenella graffmani) placed it below almost all other groups. For comparison, a man with a 1085 g brain has a callosal area of 991 mm2, and a horse with 385 g brain has 200.8 mm2, but a dolphin with a 832 g brain has a callosal area of only 180.5 mm2. Dolphins may be particularly useful for studying the phylogeny of the corpus callosum and the roles it plays in cognitive function and integration.
The sound production of the bottlenose dolphin is generally classified into two broad types: pure-tone whistles and broad-band clicks (Ralston & Herman,1989). Click sounds include both echolocation signals and the less rapid and audible burst-pulse sounds (heard to human ears as squawks, cracks, or pops), which are used socially. Sperm whales (Physeter) possess no form of whistle and mediate social behavior only by means of these broad-band clicks. Burst-pulse sounds are emitted in what is judged to be highly emotional contexts, such as aggressive episodes as well as play-chases, before copulatory sequences, and occasionally in reaction to novel or unexpected environmental objects or events (Lilly & Miller, 1961; Caldwell & Caldwell, 1967). Whistle ability in cetaceans is correlated with gregariousness to some degree. Individuals of species which may often be found alone or in small groups, such as river dolphins, harbor dolphins, and pygmy sperm whale do not whistle. Whistling species such as Stenella, Tursiops, and the common dolphin are known for assembling into incredibly large herds of hundreds or even thousands. Herman (1980a) suspects that the vocal similarities across species reflects responses to similar ecological niches or pressures: dolphins which forage communally (by echolocating prey) may keep in contact and transmit other social information through whistles. Caldwell and Caldwell (1968) noted a high degree of stereotypy in the whistle contours of individual dolphins. Up to 90% of all whistle contours emitted by bottlenose dolphins in captivity were "signature whistles", made by that individual only. Bottlenose dolphins in the wild also demonstrate a great proportion of stereotyped (signature) whistles, though possibly less than 90% (J. Mann, personal communication). Dolphins can imitate whistle sounds of their tank mates and may do so at times of stress (Ralston & Herman, 1987; Lilly, 1967). However, the complexity and arbitrary of use of whistles in a laboratory setting undermines the view that all or most dolphin whistles are stereotyped (Herman, 1986). The echolocationary skills of the Atlantic bottlenose dolphin are impressive. A dolphin can discriminate auditory information that is correlated with a variety of dimensions, including direction, range, size, shape, and material composition (Nachtigall, 1980). There is some evidence that (spatial) form perception is relatively coarse in that size differences less than 10 to 50% may not be reliably detected (Nachtigall, 1980). But Roitblat, Penner, and Nachtigall (1990) report a very high level of accuracy (94.5 %) by a dolphin wearing eyecups in matching-to-sample discriminations with echolocation only. Spatial mapping of the cetacean environment is performed within the same modality as its communications. Information from sonar probably actively contribute to the representations of objects and events in the dolphin's three-dimensional environment. In terms of language evolution, dolphin may have possessed highly integrated communicative and representational systems long before human language evolved. The rapid encephalization and great success of this order may be explain by such developments.
Tursiops truncatus are acoustic specialist in many ways. Dolphins demonstrate an excellent fidelity of auditory memory. They show the ability to story new auditory information as well as update old information rapidly (Herman & Thompson, 1982). They can listen to up to a total of eight different short sounds and then decide whether a subsequent probe sound was or was not a member of that list (Herman, 1980b). Tursiops display a better short term memory for acoustical information than humans, with possibly more precise representation, as demonstrated by middle- and long-latency components of acoustic ERP (Woods, Ridgway, Carder, & Bullock, 1986). Dolphins are also capable of vocally imitating arbitrary sounds broadcast into its tank through an underwater speaker (Richards, Wolz, & Herman, 1984; Richard, 1986). They accurately reproduce frequency, frequency modulation, amplitude modulation, and pulsed waveforms. The dolphin's ability for vocal mimicry is comparable to the more versatile mimic birds. Few terrestrial mammals, other than humans, demonstrate any skill for vocal mimicry (D'Amato & Colombo, 1986). Akeakamia (Ake for short), a subadult female dolphin at Herman's laboratory, rapidly formed the generalized mimicry concept and could reliably associated computer-generated acoustic labels to different objects (Richards, Wolz, & Herman,1984). Mimicry of another species involves production of behaviors which cannot possibly be innate and requires greater cognitive abilities than those necessary for normal species-typical behaviors (Richards,1986). Many species of dolphins can mimic the signature whistles of conspecifics (Conner & Norris, 1982). Ake's ability to also mimic artificial sounds produced by a computer reveals the great plasticity of the dolphin sound-production system. Dolphins (unless otherwise noted, the term dolphin refers to members of Tursiops truncatus) are capable of producing simultaneous but independent whistles and clicks. A young male (4 years old) dolphin, Hiappo, could correctly discriminate and respond to ten or more vocal commands (such as "jump over hoop" compared to "jump over ball"); however, his behavior consisted of a stereotyped motor response and he required repeated drilling to achieve a high level of accuracy on this task (Pack, personal communication; in comparison, see Savage-Rumbaugh et al., 1988 above). Dolphins readily form and apply rules about relationships between auditory events Herman (1988), as demonstrated by Phoenix (a subadult female Tursiops truncatus at Herman's laboratory). comprehension of imperative sentences in an acoustical artificial language (see below).The performance of monkeys and apes on visual tasks is generally much superior to performance on auditory tasks (D'Amato & Colombo, 1986). With the obvious expectation of humans, few mammals perform well on complex tasks when task information is solely auditory.
How do dolphins use their natural vocalizations? Do the whistles and pulsed sounds of dolphin vocalization possess semantical or syntactical information? Semantics and syntax, however, are subjective concepts which are directly related to human conceptions of self and identity, such concepts which cetacean species may or may not possess (Jerison, 1986). A number of scientists have designed experiments to investigate they role of dolphin vocalization in social behaviors. Lang and Smith (1978) set up an electronic acoustical link between two dolphins (a male and a female) who were housed in different tanks. During the experiment, the animals emitted numerous sounds in sequenced exchanges, rarely overlapping each other's sounds despite the great extent of their vocalizations. This suggest a general turn-taking behavior. Dolphins also demonstrate turn-taking behavior when vocalizing to (with) humans (Herman, 1980a). Lang and Smith (1978) report that dolphins demonstrate antiphonic behavior; i.e., when one dolphin vocalizes, other individuals often call in response. They were no uninterrupted, one-sided vocalization; each dolphin vocalized for no longer than a few seconds at a time, usually followed by a response from the other dolphin. Although they were out of visual and physical contact from each other, these exchanges would often persist for a number of hours. Lang and Smith (1978) recorded female's vocalizations which they later played back in the male's tank. Playback experiments of female's vocalizations in the male's tank resulted in vocalization by the male except when three of six types of vocalizations where played. (The dolphins emitted a variety of whistle types, which the human experimenters classified into roughly six groups.) They concluded that these whistles (C-, E-, and F-type which was least stereotyped of all recorded whistles in terms of frequency variation) were meaningful only in an active two-way exchange. Some whistle types appeared to have been interpreted as being used like a general call signal to localize and identify another dolphin. Another type may have been a simplified signal to maintain acoustic contact.
Whistles have been almost universally regarded as intraspecies communication signals (cf. on possible other functions, Herman 1980a); however the nature of the communication system remains obscure. Duration of whistles can vary from less than 1 sec to several seconds (Herman, 1980a). The emphasis on whistles as the primary mode of acoustic communication may not be justified (Herman, 1980a.) Both Bastian (1967) and Lilly and Miller (1961) observed behavioral correlations with pulse sounds. Attempts to correlate the contour (frequency versus time) of spontaneous whistles to environmental and behavioral events have not been conclusive, except possibly for generalized calls such as signatures and distress calls (Norris, 1983). Increased whistling occur sin unfamiliar surroundings, when dolphins are separated from familiar individuals, stranded, captured or injured (Herman, 1980a). However some confrontation with predators result in whistle suppression (best strategy to elude detection of killer whales); playbacks of predators' calls to experienced prey population , suggests learned component in silencing.
Bateson (1966) takes a less stereotyped view of human and dolphin vocal behavior. He argues that the extraordinary change in the evolution of human language was not the discovery of abstraction or generalization, but the discovery of how to be specific about something other than relationship between audience and "speaker" (prior to vervets). Prior to this, communication primarily concerned social relationships only. Human developed a mode of communication which uses a category system and relationships between categories which are appropriate for the discussion of things that can be handled (a "thing" language) while they are actually discussing the patterns and contingencies of social relationships. Adapting to life in the sea, whales lost external ears to flap and most erectile hairs, and other facial features became unexpressive. Consequently, their natural animal vocalizations had to take over the communicative functions that most animal could perform with facial expressions, or clenched fists, or wagging tails. There was an evolutionary shift from kinesic communication to vocal communication.
Bateson (1966) notes that dolphin appear to lack (or do not use) analogic information to supply emotional information. Theirs may be a purely digital form of communication. Unlike most mammals, including humans, which couple their messages with emotional or paralinguistic content which is conveyed analogically (e.g., increased intensity implies increased urgency or arousal), dolphins may have long lost this ability. Dolphin vocalizations may be digital expressions of social and emotional content. Whereas humans may be less verbally conscious of the social content of our communication -- to the point that we become uncomfortable when someone else verbally interprets our postures and gestures indicative of relational attitudes -- dolphin may constantly and consciously convey their various social attitudes.
Bastian (1967) designed an experiment to detect possible semantic functions in dolphin vocalizations. In his experiment, two dolphins (a male and a female) were housed in the same tank but were physically separated by a net drawn across the center to the tank, which would eventually be replaced by an opaque visual screen. To obtain a food reward each dolphin had to select the correct paddle from a pair of located in its half of the tank, and both dolphins had to select correctly for either to receive a reward. Also, the female had to press her paddle before the male pressed his. Bastian hoped to observed a cooperative transmission of information between the pair of dolphin. The correct paddle was signaled by an out-of-water "cue" light. A flashing light informed the dolphin that s/he was to press the left-hand paddle of the pair; a steadily lit light indicated that the right-hand paddle was to be pressed. A cue light was initially presented in each half of the tank, but as training progressed the male's light was gradually moved into the female's sector and then replaced entirely by the female's light so that the male was now totally dependent on some form of acoustic information from the female about the state of the cue light (or which paddle to press). The transmission of this type of information would imply that dolphins have the ability to create acoustic symbols to represent something as arbitrary as a flashing or steadily lit light.
The male dolphin performed almost flawlessly. However, questions remained as to the specific nature of information that was transmitted by the female. Was the female knowingly transmitting information? Bastian (1967) suggested that the extensive training period for this task may have allowed this to develop. But additional tests indicated otherwise. The female's pretrial vocalizations continued when the visual screen was withdrawn, even though the male was now able to see the cue light again. They persisted even when male was removed from the tank. A change of paddle-cue light associations considerably disrupted the cooperative performance. An attempt to reverse the roles of the two animals by providing the cue light to male was not successful initially and required extensive training. Bastian notes that in a language system, the sender and receiver can changed roles fairly easily.
Only short burst pulses emitted by the female were correlated with male correct response (she also whistled during the task). Playbacks of the short-burst pulses, without the female present, did not influence the male's response. Evans and Bastian (1969) concluded that the female developed stereotyped postures during vocalizations which preceded responding to the left or to the right paddle. Her bodily positions, whether she faced toward male's side of tank or way, may have been detected by the male and used in his decisions. They could not conclude that any natural language features were present in the transactions between the two dolphins.
One should remember that the social functions of the various sounds produced by dolphins are still in an early state of description. The use of playback techniques, improved recording (video) and analysis, may illuminate the role of pulse sounds during social encounters, the structure and variability of whistles, and the character (graded or discrete) of vocalizations. Study of the mechanisms responsible for vocal behavior and observation of the developmental process may prove especially useful.
The visual system of the dolphin has often been underrated by many researchers. Visual acuity in air is comparable to acuity underwater, although the best viewing distances differ in the two media: acuity is best as far distances (2.5 m or greater in air) and at near distances in water (1 m or less), for obvious evolutionary reasons (Herman, Peacocok, Yunker, & Madsen,1975). The ability to understand sentences expressed within a simple, artificial visual/gestural language (Herman, 1986) may be the best evidence of visual perception and processing. Even more impressive, Ake could respond correctly to degraded video displays of her gesture language. She took to television immediately and without training responded correctly to gestural commands made by a trainer on an 8" television screen. Later, these gestures were reduced to simple point-light displays. She performed almost as accurately as fluent humans to these displace, and better than novices to her sign language. Analysis of Ake's processing of gestures indicated that she process signs hierarchically (Shyan & Herman,1987) in which the location of a sign in relation to body and completion of the gesture's full path were the most important features in gesture identification. Herman, Morrell-Samuels, & Pack (1990) credited her high response accuracy to her massive receptive experience. In the eight years of her instruction, she has had to respond to thousands of gestures from numerous signers, many with questionable skills in sign language (personal observation). Visual presentation of objects can be used as substitutes for their symbols in a sentence, with little or no change in the dolphin performance. Dolphins can visually match to sample two dimensional objects, including objects new to its experience (Herman, in prep.). Forestell & Herman (1988) found some difficulty in visual perception or processing. In a matching-to-sample task, if the background on which a figure is to be displayed is not held constant, the figure and background appear to be perceived as unitary percept; the dolphin attends to the whole image rather than to object or picture. If a large common background is used for all trials, performance increases significantly. In a recent study, Herman, Hovancik, Gory, and Bradshaw (1989) report an invariance of cognitive performance involving tasks which either visual or auditory materials. Vision is not rudimentary or secondary sensory system. Visual processing is used to different extents in orientation, navigation, group movements, prey detection and capture, predator defense, and the identification of conspecifics including individuals, gender, and age classes
Bottlenose dolphins are also very curious animals. They will approach and manipulate new objects, opens gates, lifts nets, and even invents games. They often manipulate noncetacean animals, plays with familiar objects, plays chase with cetacean as well as other games such as keep away. According to a survey of 22 trainers, the most manipulative, playful, and curious cetacean species were Tursiops (truncatus and gilli), Pseudorca crassidens (false killer whale) and Orca orcinus (killer whale) (Defran & Pryor,1986). Tursiops truncatus was the best species for vocalizing in air or underwater on command. Only killer whale is less timid (e.g., can easily be segregated from conspecific for trains) and is highly response to humans, for obvious reasons. Bottlenose dolphins also demonstrate observational learning and tool use. Tayler and Saayman (1972) observed a bottlenose dolphin who adopted postures and made movements imitative of cape fur seal (Arctocephalus pusillus) when it was swimming, sleeping, or performing comfort movements (e.g., the dolphin performed seal-like strokes of flippers to propel it along). They also observed dolphin imitations of postures and swimming behavior of fish, turtles, and penguins. Tayler and Saayman (1972) report a case of observational learning and tool use. A diver used a steel hollow scraper attached to suction hose to remove of debris and seaweed growths from the concrete bottom of tank. After several days of watching, a younger dolphin named Haig was found manipulating the apparatus. By lying flat along the hose and using her flippers and rostrum, she could push the scraper in many directions, dislodging seaweed which she ate. Later, after the scraper was removed from the tank, Haig was occasionally observed holding a broken tile in her mouth and swimming with it so that it made contact with the tank bottom, dislodging seaweed which she would eat. Another dolphin, after watching Haig, began to use the same tile piece to dislodge seaweed from the tank bottom. Elementary tool use in the wild is found in dolphins in the Australian Great Shark Bay. Dolphins have been observed transporting dolphins on the end of their rostrum. It is believed they may act as protection against the stings of fish which the dolphins prey on in this area (J. Mann, pers. comm.). Tayler and Saayman (1972) hold that observational learning and imitation may play important roles in the social life of dolphins, particularly in the selection of sexual partners. Herman (1980a) believes that the capability of dolphins for vocal and motor mimicry provides the basis for various types of coordinated group activity, including synchronous swimming, cooperative feeding, and protective behavior.
The work of Herman (e.g., Herman,1980b; Herman, 1986) is especially useful and may clarify issues concerning the cognitive capacities of the Atlantic bottlenose dolphin. Herman has educated dolphins in two simple artificial languages to help determine the cognitive capacities and limitations of this species. He emphasized the comprehension of symbols, given the controversy surrounding ape language studies (which usually relied on productive skills) and findings in human development that word and symbol comprehension precedes production (Greenfield & Smith, 1979). The capacity for understanding information necessarily exceeds the capacity for producing or manipulating information. In this light, the ape language studies grossly underestimated their animals' linguistic and paralinguistic abilities because investigators concentrated on more difficult production task (Herman, 1980b), as is readily demonstrated by Kanzi. Utterances or sentences in this artificial language were produced by humans only and were either gestural (to Ake) or acoustical (to Phoenix). Only imperative or interrogative sentences can be made in these miniature languages. An imperative directed the dolphin to take a named action relative to one or more named objects and their modifiers. A set of syntactic rules governed the grammatical functions of the words so that, as in many human languages, rearrangements of word order created new meanings for lexical items. The dolphins demonstrated they comprehension of an imperative statement by manually carrying out its instructions. The dolphins performed accurately and reliably to the majority of constructions enabled by this artificial language, including responses to sentences with novel lexical items or novel constructions (i.e., additional modifiers or concatenated sentences). Both semantic and syntactic components of a sentence must be understood for a correct response by the dolphin (Herman,1986). For example, the dolphins needed to respond differently to contrasting sentences such as FRISBEE SURFBOARD FETCH and SURFBOARD FRISBEE FETCH. The first sentence instructs the dolphin: "Take the surfboard to the frisbee" and the second, "Take the frisbee to the surfboard." The gestural language employed a grammatical construction which designated the first object an the indirect object and the second as the direct object (on which the action of the sentence would be performed). The acoustic language employed a simpler construction in which object order resembled the behavioral sequence: the first object referred to was taken to the second. Correct responses in the language which had an inverse grammar are not easily explained by stimuli-response chains (Herman, 1980b).
Interrogative sentences posed either yes or no questions about the presence or absence of named objects in the dolphin tank. For example, a dolphin could be asked whether there was a hoop in her tank (which was given as "HOOP + QUESTION" in the gestural format). The dolphin replied "Yes" or "No" by pressing either one of two paddles. Herman and Forestell (1985) demonstrated that dolphins were capable of accurately reporting either presence or absence of named objects in a series of studies. The ability to understand symbolic references to absent objects demonstrates that symbols in this language stood as substitutes or surrogates for the referenced objects. It also demonstrated some degree of displacement (cf. Hockett, 1960). Also, success at this task required that a dolphin make a mental representation of the object which she would compare to those objects present in the tank.
Bottlenose dolphins, one of the many large-brained mammals represented in the aquatic domain, demonstrate understanding of semantic as well as syntactical relationships, independent of modality (Herman, 1986). Dolphins can respond selectively to symbolic information presented in two structural forms, one linear and one more-arbitrarily derived, across two modalities. They demonstrate some recursive processing, reporting (without action), sensitivity to relational features, novel formulations, syntactic order, temporal displacement, all amid context variation, semantic generalization, and sign/symbol interchangeability.
Herman's work and his conclusions have not gone uncriticized. Premack (1985) argues that the performance by Herman's dolphins can be explained by two rules: (1) (Property) Object Action; and (2) (Property) Object Action2 (Property) Object. Human language cannot be framed in terms of such categories as object, property, or action. According to Premack (1985), Herman's dolphins have only shown: (1) discrimination of temporal order, (2) learning rules based on perceptual classes; and (3) possible discrimination of object classes (e.g., can categorize hoops or balls made up of different compositions, etc.). Herman (1986) claims that indicating the presence or absence of an object in a tank requires some rudimentary capacity for displacement in one's mental representations. Premack (1985) argues that displacement does not concern the individual's memory of a verbal statement; it concerns his knowledge of the world, and his ability to use language to access that knowledge. Displacement requires a dolphin to grasp the concept of past and future occurrence of events. As a test of displacement, one could instruct a dolphin to leap into the air, say, if a hoop had been in some location of the tank during the previous day; else roll on your side. Schusterman & Gisiner (1986) argued that sea lions perform comparably with dolphins in a similarly-constructed artificial language; however, Schusterman makes no claims that sea lions understand linguistic rules or categories (e.g., noun phrase).
Atlantic bottlenose dolphins are very social creatures. Close affiliations between mother and infant can be maintained into adulthood (Defran & Pryor, 1986). Other long-term associations between individuals, consisting of dyads or triads, have been observed (HERMAN, 1988). Associated individuals often display a high degree of behavioral synchrony, such as breathing and leaping in unison, swimming in pairs with pectoral fins touching. Bottlenose dolphins may establish temporary alliances with other dyads or triads to herd females (Johnson & Norris,1986; J. Mann, pers. comm.) When alarmed, dolphins tend to bunch together. Most odontocetes (toothed whales) are social predators, displaying various degrees of recruiting schoolmates, aggregating prey, and coordinated capture. With such a degree of mutual dependence, within dolphin societies cetaceans are likely to have complex rules or conventions to govern social relationships and behaviors. Lilly (1967) reports an incident in which he had given a number of injections to a young dolphin, who soon after began aggressive and tried to bite him. The dolphin's mother intervened, however, knocking her infant against the tank wall before he could bite him. Pryor (1981) remarks how she has never seen a porpoise "'go beserk' and attack a human with persistence like a dog or horse." She suggests that a dolphin's social upbringing results in a "fair and firm disciplinarian." Consistent spatial distributions of adults, primarily in terms of verticality (access to surface), interindividual position and distance (kept by agonistic displays by males), and privilege of association, are prevalent in dolphin societies (Evans & Bastian,1967). Variable association patterns are also present in dolphins which do not adhere to strict dominance hierarchies, but demonstrate more symmetrical social interactions (Johnson & Norris, 1986). Johnson and Norris (1986) have observed what they term "echolocation manners" in frame-by-frame analysis of wild spinner dolphins. Dolphins poke and caress one another with sound, but the highest energy emissions can be extremely threatening. Consequently, perpendicular approaches, where the echolocative beams coming from forehead would be greatest, are relatively rare, usually in the case of potentially threatening interactions. A frame-by-frame analysis of a film of wild spinner dolphin turns show that sharp turns of the head are avoided; slight turns of the head are made while each dolphin maintains a basically parallel orientation to the others. Pryor (1981) notes that dolphins make eye contact more often than many other mammals. Dolphins typically show a very strong interest in humans who approach them or enter their tank. They generally draw close and make contact (Pryor, 1981; personal observation). They solicit human contact.
Human language, human cognition, and human culture is the product of extensive integration of the two symbolic systems which most or all most mammals possess to varying degrees: for representation and for communication. Morgan's canon provides a guideline to interpreting cognitive capacities from an individual's behavior are made. Morgan (1894) wrote: "In no case may we interpret an action as the outcome of the exercise of a higher psychical faculty if it can be interpreted as the outcome of the exercise of one which stands lower in the psychological scale" (p. 53). In short, reconstruction of an animals' perceptual world should involve the simplest picture that is consistent with the data. This canon is not an advertisement for parsimonious explanation; in fact, it often violates parsimony (e.g., a convoluted S-R network is look better upon than concepts such as consciousness, etc.). Prescription to this canon is less satisfying for larger-brain animals (e.g., dolphins, apes, humans) than for member of most other species with which this strategy has been used. Given the high metabolic (evolutionary) cost of a large brain, speculations about the behavior of large-brained mammals should invoke higher cognitive capacities; that is, brain mechanisms that are known to require large amounts of nervous tissue (Jerison,1986b). Other explanations violate parsimony.
Human language is a communication system that is also a cognitive system . The first steps toward complete integration of representational and communicative systems in the mammalian brain probably began millions of years ago. Considering general ecological, morphological, and perceptual constraints, cetaceans seem best suited, among all mammals, to develop communal constructions of reality. Strangely, there is little evidence of a theory of mind in dolphins, positive or negative. Accounts of altruistic acts by cetaceans are many, and even some behaviors which border on pedagogy, but few if any empirical tests have been designed to investigate this issue. There is anecdotal evidence of dolphins behaving as if they are aware of the existence of other minds. Do dolphins monitor the effects of signals that they emit and change their behaviors accordingly? It may be premature to answer now. Behaviors which are illicit and punishable are often performed only when a dolphin believes no one is around (e.g., Savage-Rumbaugh and Hopkins, 1986). When a dolphin squirts water at a human (to show annoyance), he will often raise his head out of the water to curiously observe the effect his behavior had on the unsuspecting victim (personal observation). Both examples show an awareness of effects one's behavior has on others. Dolphins demonstrate what Pryor (1986) calls insightful behavior. An experienced animal will "check out" a training criteria by running through a series of variations on a learned behavior. Development of a productive communicative skills in dolphins, such as mimicry, would benefit many areas of investigation. "Future work on artificial systems should pursue the development of phoneme-like set of recombinable sound patterns which optimize perceptual distinctiveness and reproducibility" (Richards, 1986). The integration of representational and communicative systems, as partly demonstrated by the impairment effect (Premack, 1985), have yet to be extensively explored in dolphins. Even instruction in the use of lexigram systems would start to resolve some questions concerning dolphin cognitive and linguistic abilities.
The human brain appears to better organized to impose structure on visual data than on auditory data, and in the dolphins the reverse may be true. Forestell and Herman (1986) report indirect evidence that an object is more an object (e.g., perceived as contiguous) when it is heard than when it is seen by a dolphin. The kind of self that might be constructed by dolphins probably involves acoustic more than visual inputs. Echolocation shares an unusual structural feature with human language: its contribution to the reality constructed by the brain may depend on a signal generated by other animals. Various social interactions in bats, such as foraging and agonistic behaviors, depend on the ability to intercept the vocal signals of others (Fenton, 1980). Why did many species of dolphins attain such large brains? In view of their high cost, we must propose enhancements of data from echolocation (Jerison, 1988). Co-occurrence of communicative and perceptual processes in the same modality would create tremendous pressure for a communally-shared symbol system. Echolocating animals can possibly share raw acoustic information, unprocessed, the very elements from which representational and communicative codes are developed. Integration of representational and communicative systems in dolphins may not be as much a unifying process as less segregation at the outset. Lilly (1967) believed that dolphins were unable to distinguish their sonar from their communications. One's concept of self is tied to the ways and degree of acquiring knowledge. Sharing the very vividness of natural objects would result in intense group cohesion with a reduction of individuating processes. This communal experience may change the boundaries of the self: many members of a group may act as a "decision-making unit" (Jerison, 1986).
Different animals are conscious of different aspects of their world, from proprioceptive body awareness to awareness of agency and social agency (Cheney & Seyfarth). The extent of mental attribution in humans ("linguistic self consciousness", Crook, 1983) may be a consequence of extensive integration of representational and communicative systems, channeled by social demands. During human evolution, elements of intra-individual communication (cognitive structures and processes) became progressively linked to and developed with elements of interindividual communication (displays, calls). This integration requires tremendous expenditures in terms of time and other resources to develop and maintain in the members of a culture. Poets, artists, scientists, and other creative members of society may develop a higher degree of integration of these two symbol systems during ontogeny and work very hard to maintain this desegregration of abilities as adults. "The quality and range of intellectual performance demonstrated by any member of a species is in part a function of the breadth and intensity of the long-term education that individual has received" (Herman,1986). The role education plays in a species' representational and communicative systems cannot be overlooked. Nonlanguage-trained animals perform worse in tasks which require complex manipulation or mental representations than language-trained animals in similar tasks (Premack, 1985, see above). Language-trained dolphin & signing human both process patterns hierarchically, more so than controls (Shyan & Herman,1987). As with apes, we must admit that there is no convincing evidence for a sophisticated language in the natural communications of dolphins. However, we must also admit that the basic component(s) of dolphin signaling is still largely unknown (Smith,1986). Proponents of continuity theory can take comfort from other findings, such as mimicry, observational learning, protocultural influences on behavior, pedagogy, and the attribution of minds in others: all of which are telltale signs of the partial integration of representational and communicative capacities of mammals.
Anderson, R.E. (1987). Cognitive ethology and the study of human cognition. In D. M. Topping, D.C. Crowell and V.N. Kobayashi (Eds.) Thinking across cultures: Third international conference on thinking. (245-253). Hillsdale: Erlbaum.
Bastian, J. (1967). The transmission of arbitrary environmental information between bottlenosed dolphins. In R. G. Busnel (Ed.) Animal Sonar Systems, Vol. II, (803-873). Jouy-en-Josad, France: Laboratoire de Physiologie Acoustique.
Bastian, J. (1968). Psychological perspectives. In T.A. Sebeok, Animal communication: Techniques of study and results of research. (572-591). Bloomington: Indiana University.
Bateson, G. (1966). Problems in cetacean and other mammalian communication. In K. Norris (Ed.) Whales, Dolphins, and Porpoises. (569-582). Los Angeles: University of California.
Bickerton, D. (1986). More than nature needs? A reply to Premack. Cognition, 23, 73-79.
Bogen, G. (1986). On the relationship of cerebral duality to creativity. Bulletin of Clinical Neurosciences, 51, 30-32.
Bogen, J.E. (1985). The callosal syndromes. In K.M. Heilman and
E. valenstein (Eds.) Clinical neuropsychology (2nd ed.). (295- 338). New York: Oxford University.
Bowerman, M. (1982). Reorganizational processes in lexical and syntactic development. In E. Wanner & L. Gleitman (Eds.) Language Acquisition: The state of the art. New York: Cambridge University.
Bradbury, J.W. (1986).Social complexity and cooperative behavior in delphinids. In R.J. Schusterman, J. Thomas, and F.G. Wood (Eds.) Dolphin cognition and behavior: A comparative approach. (361- 372). Hillsdale: Erlbaum.
Bradshaw, J.L. and Nettleton, N.C. (1981). The nature of hemispheric specialization in man. Behavioral and Brain Sciences, 4, 51-63.
Brown, G. (1977). Listening to spoken English. Longdon: Longram.
Bullock, T.H. (1986). Suggestions for research on ethological and comparative cognition. In R.J. Schusterman, J. Thomas, and F.G. Wood (Eds.) Dolphin cognition and behavior: A comparative approach. (207-219). Hillsdale: Erlbaum.
Bullock, T.H. and Gurevich, V.S. (1979). Soviet literature on the nervous system and psychobiology of Cetacea. In J.R. Smythies and R.J. Bradley (Eds.) International review of Neurobiology, Vol. 21, 49-129. New York: Academic.
Caldwell, D.F. and Caldwell, M.C. (1976). Cetaceans. In T.A. Sebeok (Ed.) How animals communicate. (794-808). Bloomington: Indiana University.
Cheney, D. L. & Seyfarth, R.M. (1990). How monkeys see the world. Chicago: University of Chicago.
Chomsky, N. (1980). Human language and other semiotic systems. In T. A. Sebeok and J. Umiker-Sebeok (Eds.) Speaking of apes: A critical anthology of two-way communication with man. (429- 440). New York: Plenum.
Corballis, M.C. (1980). Laterality and myth. American Psychologist, 35, 284-295.
Corballis, M.C. (1989). Laterality and human evolution. Psychological Review, 96, 492-505.
Cornell, E. H. (1984). Developmental continuity of memory mechanisms: Suggestive phenomena. In R. Kail & N.E. Spear (Eds.) Comparative perspectives on the development of memory. (287-316). Hillsdale NJ: Erlbaum.
D'Amato, M.R. & Colombo, M. (1986). Auditory delayed matching-to- sample in monkeys (Cebus apella). Animal Learning and Behavior, 13, 375-382.
de Luce, J. and Wilder, H.T. (Eds.) (1983). Language in primates: Perspectives and implications. New York: Springer-Verlag.
Dennett, D.C. (1987). The intentional stance. Cambridge, MA: MIT/Bradford Books.
Dewson, J. (1977). Preliminary evidence of hemispheric asymmetry of auditory function in monkeys. In S. Harnad, R. Doty, L. Goldstein, J. Jaynes, & G. Krauthamer (Eds.) Lateralization in the nervous system. (63-74). New York: Academic Press.
Dreher, J.J. (1966). Cetacean communication: Small-group experiment. In K. Norris (Ed.) Whales, Dolphins, and Porpoises. (529-544). Los Angeles: University of California.
Dreher, J.J. and Evans, W.E. (1964). Cetacean communication. In W. N. Tabolga (Ed.) Marine Bio-Acoustics. (373-393). New York: Macmillan.
Eglash, R. (1984). The cybernetics of cetacea. Dissertation. University of California, Los Angeles.
Eisenberg, J.F. (1981). The mammalian radiations. Chicago: University of Chicago.
Evans, W.E. and Bastian, J. (1969). Marine mammal communication: Social and ecological factors. In H.T. Anderson (Ed.) The Biology of Marine Mammals. (425-475). New York: Academic.
Falk, D. (1980). Language, handedness, and primate brains: Did the Australopithecines sign? American Anthropologist ,82 72-78.
Fischer, J.L. (1981). Magical imitation in the origin of language. In
E. de Grolier (Ed.) Glossogenetics: The origin and evolution of language. (313-328). New York: Harwood Academic Publishers.
Fouts, R.S., Hirsch, A.D. & Fouts, D. H. (1982). Cultural transmission of a human language in a chimpanzee mother-infant relationship. In H.E. Fitzgerald, J.A. Mullins, & P. Gage (Eds.) Psychobiological perspectives. New York: Plenum Press.
Forestell, P.H., and Herman, L. M. (1988). Delayed matching of visual materials by a bottlenosed dolphin aided by auditory symbols. Animal Learning and Behavior, 16, 137-147.
Gallup, G.G. (1982). Self-awareness and the emergence of mind in primates. American Journal of Primatology, 2, 237-248.
Gallup, G.G. (1985). Do minds exist in species other than our own? Neuroscience and Biobehavioral Review, 9, 631-641.
Gallup, G. G. & Suarez, S. D. (1985) Overcoming out resistance to animal research: Man in comparative perspective. In D. W. Rajecki (Ed.) Comparing Behavior: Studying man studying animals. (5-26) Erlbaum, Hillsdale NJ.
Gardner, H. (1979). The shattered mind. New York: Random House.
Gaskin, D.E. (1982). The ecology of whales and dolphins. London: Heineman Educational Books.
Geschwind, N. & Galaburda, A. (1987). Cerebral lateralization: Biological mechanisms, associations, pathology. Cambridge MA: Bradford/MIT Press.
Gibson, K.R. (1981). Comparative neurobehavioral ontogeny and the constructionist approach to the evolution of the brain, object manipulation, and language. In E. de Grolier (Ed.) Glossogenetics: The origin and evolution of language. (37-61). New York: Harwood Academic Publishers.
Glezer, I.I, Morgane, P.J., and Jacobs, M.S. (1988). Implications of the "initial brain" concept for the brain evolution in Cetacea. Behavioral and Brain Sciences, 11, 75-116.
Glick, S. D. (1985). Cerebral lateralization in nonhuman species. New York: Academic Press.
Greenfield, P.M. and Savage-Rumbaugh, E.S. (in press). Imitation, grammatical development, and the invention of protogrammar by an ape. In N. Krasnegor, D. Rumbaugh, M. Studdert-Kennedy, and R. Schiefelbusch (Eds.) Biobehavioral foundations of language development. Hillsdale: Erlbaum.
Greenfield, P.M. and Savage-Rumbaugh, E.S. (1984). Perceived variability and symbol use: A common language-cognition interface in children and chimpanzees (Pan troglodytes). Journal of Comparative Psychology, 96, 201-218.
Greenfield P. & Smith. J., 1979. The structure of communication in early development. New York: Academic Press.
Griffin, D.R. (1981). Prospects for a cognitive ethology (with commentary) Behavioral and Brain Sciences, 4, 527-538;555- 629.
Griffin, D.R. (1987). Education about animal minds. In D. M. Topping, D.C. Crowell and V.N. Kobayashi (Eds.) Thinking across cultures: Third international conference on thinking. (255-267). Hillsdale: Erlbaum.
Healy, A.F. (1980). Can chimpanzees learn a phonemic language? In T. A. Sebeok and J. Umiker-Sebeok (Eds.) Speaking of apes: A critical anthology of two-way communication with man. (141- 144). New York: Plenum.
Herman, L. M. (1975). Interference and auditory short-term memory in the bottlenosed dolphin. Animal Learning and Behavior, 3 , 43-48.
Herman, L. M. (1980). Cognitive characteristics of dolphins. In L. M. Herman (Ed.) Cetacean behavior: Mechanisms and functions. (363-429). New York: Wiley Interscience.
Herman, L. M. (1986). Cognition and language competencies of bottlenosed dolphin. In R.J. Schusterman, J. Thomas, and F.G. Wood (Eds.) Dolphin cognition and behavior: A comparative approach. (221-252). Hillsdale: Erlbaum.
Herman, L. M. (1986). Contributions of animal cognition to the understanding of human cognition. In S.E. Newstead, S.H. Irvine, and P. L. Dann (Eds.) Human assessment: Cognition and motivation. (95-98). Dordecht: Nijhoff.
Herman, L. M. (1987). Receptive competencies of language-trained animals. Advances in the Study of Behavior, 17, 1-60.
Herman, L. M. (1988). The language of animal language research: Reply to Schusterman and Gisiner. Psychological Record, 38, 349-362.
Herman, L. M. (Ed.) (1980). Cetacean behavior: Mechanisms and functions. New York: Wiley Interscience.
Herman, L. M. (in press). What the dolphin knows, or might know, in its natural world. In K. Pryor and K. S. Norris (Eds.) School structure and social behavior of wild delphinids. Los Angeles: University of California.
Herman, L. M. and Forestell, P.H. (1985). Reporting presence or absence of named objects by a language-trained dolphin. Neuroscience and Biobehavioral Reviews, 9, 667-681.
Herman, L. M. and Tavolga, W.N. (1980). The communication systems of cetaceans. In L. M. Herman (Ed.) Cetacean behavior: Mechanisms and functions. (149-209). New York: Wiley Interscience.
Herman, L. M. and Thompson, R.K.R. (1982). Symbolic, identity, and problem delayed matching of sounds by the bottlenosed dolphin. Animal Learning and Behavior, 10, 22-34.
Herman, L. M., Hovancik, J.R., Gory, J.D., and Bradshaw, G.L. (1989). Generalization of visual matching by a bottlenosed dolphin (Tursiops truncatus): Evidence for invariance of cognitive performance with visual and auditory materials. Journal of Experimental Psychology: Animal Behavior Processes, 15, 124- 136.
Herman, L. M., Morrel-Samuels, P. and Pack, A. A. (1990). Bottlenosed dolphin and human recognition of veridical and degraded video displays of an artificial gestural language. Journal of Experimental Psychology: General, 119, 215-230.
Herman, L. M., Richards, D.G., and Wolz, J.P. (1984). Comprehension of sentences by bottlenosed dolphins. Cognition, 16, 129-219.
Herman, L.M., Peacocok, M.F., Yunker, M.P., & Madsen, C. (1975). Bottlenosed dolphin: Double-slit pupil yields equivalent aerial and underwater acuity. Science, 139, 650-652.
Hockett, C.F.(1960). The origin of speech. Scientific American, 203, 89-96.
Hockett, C.F. and Altmann, S.A. (1968). A note on design features. In T.A. Sebeok, Animal communication: Techniques of study and results of research. (61-72). Bloomington: Indiana University.
Jaynes, J. (1976a). The evolution of language in the late Pleistocene. Annals of the New York Academy of Sciences, 280, 312-325.
Jaynes, J. (1976b). Origins of human consciousness in the breakdown of the bicameral mind. New York: Houghton Mifflin.
Jerison, H. (1988). Evolutionary neurology and the origin of language as a cognitive function. In M.E. Landsberg (Ed.) The genesis of language. (3-9). New York: de Gruyter.
Jerison, H.J. (1986). The perceptual world of dolphins. In R.J. Schusterman, J. Thomas, and F.G. Wood (Eds.) Dolphin cognition and behavior: A comparative approach. (141-166). Hillsdale: Erlbaum.
Jerison, H.J. and Jerison, I. (Eds.) (1986). Intelligence and evolutionary biology. New York: Springer-Verlag.
Krushinskaya, N.L. (1986). The behavior of cetaceans. In G. Pilleri (Ed.) Investigations of cetacea. 19, 211-273.
Kuhl, P.K. & Miller, J.D. (1975). Speech perception by the chinchilla: Voiced-voiceless distinction in alveolar plosive consonants. Science, 190, 69-72.
Laitman, J.T. (1981). The evolution of the hominid upper respiratory system and implications for the origins of speech. In E. de Grolier (Ed.) Glossogenetics: The origin and evolution of language. (63-90). New York: Harwood Academic Publishers.
LeMay, M. & Geschwind, N. (1975). Hemispheric differences in the brains of great apes. Brain, Behavior and Evolution, 11, 48-52.
Lenneberg, E.H. (1980). Of language, apes, and brains. In T.A. Sebeok & J. Umiker-Sebeok (Eds.) Speaking of Apes: A critical anthology of two-way communication with man. (115-140). New York: Plenum Press.
Lenneberg, E.H. (1980). A word between us. In T. A. Sebeok and
J. Umiker-Sebeok (Eds.) Speaking of apes: A critical anthology of two-way communication with man. (71-84). New York: Plenum.
Levy, J. (1986). The evolution of human asymmetry. In H.J. Jerison & I. Jerison. The evolutionary biology of intelligence. (157-173).
Lieberman, P. (1981). On the nature and evolution of the biological bases of language. In E. de Grolier (Ed.) Glossogenetics: The origin and evolution of language. (91-114). New York: Harwood Academic Publishers.
Lilly, J.C. and Miller, A.M. (1961). Vocal exchanges between dolphins. Science, 134, 1873-1876.
Lilly, J.C. (1967). The mind of the dolphin. New York: Doubleday.
Limber, J. (1980). Language in the child and chimp? In T.A. Sebeok & J. Umiker-Sebeok (Eds.) Speaking of Apes: A critical anthology of two-way communication with man. New York: Plenum Press.
Livingstone, F.B. (1981). Evolutionary theory, human adaptation, and the evolution of language. In E. de Grolier (Ed.) Glossogenetics: The origin and evolution of language. (163-184). New York: Harwood Academic Publishers.
Mackay, R. S. and Liaw, H.M. (1981). Dolphin vocalization mechanisms. Science, 212, 676-678.
Marler, P. (1977). The structure of animal communication sounds. In T. H. Bullock (Ed.) Recognition of complex acoustic signals. (17- 35). Berlin: Dahlem Konferenzen.
Menzel, E.W. (1986). How can you tell if an animal is intelligent? In R.J. Schusterman, J. Thomas, and F.G. Wood (Eds.) Dolphin cognition and behavior: A comparative approach. (167-181). Hillsdale: Erlbaum.
Miller, G.A. (1981). Language and speech. San Francisco: Freeman.
Mitchell, R.W., Yao, P., Sherman, P.T., and O'Regan, M. (1985). Discriminative responding of a dolphin (Tursiops truncatus) to differentially rewarded stimuli. Journal of Comparative Psychology, 99, 218-225.
Montagu, A. and Lilly, J.C. (1963). The dolphin in history. Los Angeles: University of California.
Morgan, C.L. (1894). An introduction to comparative psychology. London: Walter Scott.
Morgane, P.J., Glezer, I.I, and Jacobs, M.S. (1988). Visual cortex of the dolphin: An image analysis study. Journal of Comparative Neurology, 273, 3-25.
Morgane, P.J., Jacobs, M.S., and Galaburda, A. (1985). Evolutionary aspects of cortical organization in the dolphin brain. In M. Bryden and R.J. Harrison (Eds.), Research on dolphins. (71-98). Oxford: Oxford University .
Morgane, P.J., Jacobs, M.S., and Galaburda, A. (1986) Evolutionary morphology of the dolphin brain, In R.J. Schusterman, J. Thomas, and F.G. Wood (Eds.) Dolphin cognition and behavior: A comparative approach. (5-29). Hillsdale: Erlbaum.
Morrel-Samuels, P. and Herman, L.M. (in press). Cerebral asymmetry during comprehension of gesture language signs: A brief comparison of humans and dolphins. 24 pp.
Morrel-Samuels, P., Herman, L.M. , and Bever, T. (1989). A left- hemisphere advantage for gesture-language signs in the dolphin. Paper presented at the meeting of the Psychonomic Society, Atlanta GA.
Mukhametov, L.M. (1984). Unihemispheric slow wave sleep in the brain of dolphins and seals. In S. Inoue and A. Borbly (Eds.) Endogenous sleep substances and sleep regulation. (67-75). Tokyo: Scientific Societies.
Mukhametov, L.M., Supin, A. Y., and Polyakova, I.G. (1977) Interhemispheric asymmetry of the electroencephalographic sleep patterns in dolphins. Brain Research, 134, 581-584.
Parker, S.T. (1985). A social-technological model for the evolution of language. Current Anthropology, 26, 617-639.
Parker, S.T. and Gibson, K. (1990). (Eds.) "Language" and intelligence in monkeys and chimpanzees. New York: Cambridge University.
Partan, S.R., Jerger, K.K., Xitco, M.J., Roitblat, H.L., Herman, L. M., Williams, H.N.S., Hoffines, M.A., Gory, J.D., Ralston, J.V.B., (1988) Correlations of dolphin vocalizations with specific behaviors. Poster presented at the 116th meeting of the Acoustical Society of America.
Passingham, H. (1975). Changes in the size and organization of the brain in man and his ancestors. Brain Behavior and Evolution, 11, 73-90.
Petersen, M.R. (1982). The perception of species-specific vocalization by primates: A conceptual framework. In C.T. Snowdon, C.H. Brown & M.R. Petersen (Eds). Primate communication. (171- 211) New York: Cambridge University.
Peterson, M.R., Beecher, M.D., Zoloth, S.R., Moody, D.B., and Stebbins, W.C. (1978). Neural lateralization of species-specific vocalizations by the Japanese macaques (Macaca fuscata). Science, 202, 324-326.
Premack, D. (1985). "Gavagai!" of the future history of the animal language controversy. Cognition, 19, 207-296.
Premack, D. and Woodruff, G. (1978). Does the chimpanzee have a theory of mind? Behavioral and Brain Sciences, 1, 515-526.
Pulleyblank, E. G. (1981). The beginnings of duality of patterning in language. In E. de Grolier (Ed.) Glossogenetics: The origin and evolution of language. (369-410). New York: Harwood Academic Publishers.
Radner, D. and Radner, M. (1989). Animal Consciousness. Buffalo: Prometheus Books.
Ralston, J. V. and Herman, L. M., (1989). Dolphin auditory perception. In J. R. Dooling and S. H. Hulse (Eds.) The Comparative Psychology of Complex Acoustic Perception. (295-328). Hillsdale: Erlbaum.
Reynolds, P.C. (1981). Ape constructional ability and the origin of linguistic structure. In E. de Grolier (Ed.) Glossogenetics: The origin and evolution of language. (185-200). New York: Harwood Academic Publishers.
Richards, D.G. (1986). Dolphin vocal mimicry and vocal object labeling. In R.J. Schusterman, J. Thomas, and F.G. Wood (Eds.) Dolphin cognition and behavior: A comparative approach. (273- 288). Hillsdale: Erlbaum.
Richards, D.G., Wolz, J.P., and Herman, L. M. (1984). Vocal mimicry of computer-generated sounds and vocal labeling of objects by a bottlenosed dolphin, Tursiops truncatus. Journal of Comparative Psychology, 98,10-28.
Ridgway, S.H. (1986). Brain size and symmetry in three dolphin genera. Anatomical Record, 193, 664-665.
Ridgway, S.H. (1986). Dolphin brain size. In M. Bryden and R.J. Harrison (Eds.), Research on dolphins. (59-70). Oxford: Oxford University.
Ridgway, S.H. (1986). Physiological observations on the dolphin brains. In R.J. Schusterman, J. Thomas, and F.G. Wood (Eds.) Dolphin cognition and behavior: A comparative approach. (31- 59). Hillsdale: Erlbaum.
Ristau, C.A. (1983). Symbols and indication in apes and other species? Journal of Experimental Psychology: General, 112, 498-507.
Ristau, C.A. and Robbins, D. (1982). Language in the great apes: A critical review. Advances in the Study of Behavior, 12, 141- 255.
Roitblat, H. S. and Herman, L. M., (1989). Animal thinking: the development and status of research and theory on animal cognition. In D. M. Topping, D.C. Crowell and V.N. Kobayashi (Eds.) Thinking across cultures:Third international conference on thinking. (269-290). Hillsdale: Erlbaum.
Roitblat, H.L., Penner, R.H., and Nachtigall, P.E. (1990). Matching-to- sample by an echolocating dolphin (Tursiops truncatus). Journal of Experimental Psychology: Animal Behavior Processes, 16, 85-95.
Saban, R. (1981). Asymmetry of the middle meningeal veins network in the fossil man and its possible significance. In E. de Grolier (Ed.) Glossogenetics: The origin and evolution of language. (115- 141). New York:Harwood Academic Publishers.
Savage-Rumbaugh, E.S (1986). Ape language: From conditioned response to symbol. New York: Columbia University.
Savage-Rumbaugh, E.S. and Hopkins, W.D. (1986) Awareness, intentionality, and acquired communicative behaviors: Dimensions of intelligence. In R.J. Schusterman, J. Thomas, and F.G. Wood (Eds.) Dolphin cognition and behavior: A comparative approach. (303-313). Hillsdale: Erlbaum.
Savage-Rumbaugh, S., Romski, M.A., Hopkins, W.D., and Servik, R.A. (198?) Symbol acquisition and use by Pan troglodytes, Pan paniscus, Homo sapiens. In Understanding Chimpanzees. (266-295).
Savage-Rumbaugh, E.S, Romski, M.A., Servik, R., and Pate, J.L. (1983). Assessing symbol usage versus symbol competency. Journal of Experimental Psychology: General, 112, 508-512.
Savage-Rumbaugh, E.S., Rumbaugh, D.M., and McDonald, K. (1985). Language learning in two species of apes. Neuroscience and Biobehavioral Reviews, 9, 653-665.
Savage-Rumbaugh, S., Romski, M.A., Hopkins, W.D., and Sevcik, R.A. (1988). Symbol Acquisition and use by Pan troglodytes, Pan paniscus, and Homo sapiens. In P.G. Heltne & L.A. Marquandt (Eds.)Understanding chimpanzees. (266-295). Cambridge: Harvard University.
Schusterman, R.J., and Gisiner, R. (1986). Animal language research: Marine mammals re-enter the controversy. In H.J. Jerison and I. Jerison (Eds.) Intelligence and evolutionary biology. (317- 350). New York: Springer-Verlag.
Schusterman, R.J., and Gisiner, R. (1988) Artificial language comprehension in dolphins and sea lions: The essential cognitive skills. Psychological Record, 38, 311-348.
Shyan, M.R., and Herman, L. M. (1987). Determinants of recognition of gestural signs in an artificial language by Atlantic bottlenosed dolphin (Tursiops truncatus) and humans (Homo sapiens). Journal of Comparative Psychology, 101, 112-125.
Smith, H.A.P. (1965). Communication between dolphins in separate tanks by way of an electronic acoustic link. Science, 150, 1839-1844.
Smith, W.J. (1986). Signaling behavior: Contributions of different repertoires. In R.J. Schusterman, J. Thomas, and F.G. Wood (Eds.) Dolphin cognition and behavior: A comparative approach. Hillsdale: Erlbaum.
Snowdon, C.T. (1982) Linguistic and psycholinguistic approaches to primate communication. (212-238). In C.T. Snowdon, C.H. Brown & M.R. Petersen (Eds). Primate communication. (171- 211) New York: Cambridge University.
Sugarman, S. (1983). Why talk? Comment on Savage-Rumbaugh
et al., Journal of Experimental Psychology: General, 112, 493- 497.
Terrace, H.S. (1979). Is problem-solving language? Journal of the Experimental Analysis of Behavior, 31, 161-175.
Terrace, H.S. (1985). In the beginning was the "name". American Psychologist, 40, 1011-1028.
Terrace, H.S. (1985). On the nature of animal thinking. Neuroscience and Biobehavioral Review, 9, 643-652.
Thompson, R.K.R. and Herman, L. M. (1977). Memory for lists of sounds by the bottlenosed dolphin: Convergence of memory processes with humans? Science, 195, 501-503.
Thompson, R.K.R. and Herman, L. M. (1981). Auditory delayed discrimination by the dolphin: Nonequivalence with delayed matching performance. Animal Learning and Behavior, 9, 9-15.
Thorpe, W.H. (1972). The comparison of vocal communication in animals and man. In R. A. Hinde (Ed.) Non-Verbal Communication. (27-47). Cambridge: Cambridge University.
Tobias, P.V. (1987). The brain of Homo habilis: A new level of organization in cerebral evolution. Journal of Human Evolution, 16, 741-761.
Turner, R.N. (1964). Methodological problems in the study of cetacean behavior. In W. N. Tabolga (Ed.) Marine Bio-Acoustics. (337-351). New York: Macmillan.
Waldrop, M.M. (1984). The necessity of knowledge. Science, 223, 1279-1282.
Wellman, H.M. (1988). First steps in the child's theorizing about the mind. In J.W. Astington, P.L. Harris, & D.R. Olson (Eds.) Developing theories of mind. Cambridge: Cambridge University Press.
Wind, J. (1982). On sapientization and speech. Current Anthropology, 23, 107.
Wind, J. (1981). Primate evolution and the emergence of speech. In E. de Grolier (Ed.) Glossogenetics: The origin and evolution of language. (15-35). New York: Harwood Academic Publishers.
Wolpoff, M.H. (1985). Ramapithecus and Hominid Origins. In R.L. Ciochon, and J.G. Feagle (Eds.) Primate evolution and human origins. (244-253). Benjamin/Cummings Publishing, Menlo Park CA.
Woods, D.L., Ridgway, S.H. and Bullock, T.H. (1986). Middle- and long- latency auditory event-related potentials in dolphins. In R.J. Schusterman, J. Thomas, and F.G. Wood (Eds.) Dolphin cognition and behavior: A comparative approach. (61-77). Hillsdale: Erlbaum.
Worthy, G.A.J. and Hickie, J.P. (1986). Relative brain size in marine mammals. American Naturalist, 128, 445-459.
Wursig, B. (1979). Dolphins. Scientific American, 240, 136-148.
Wursig, B. (1989). Cetaceans. Science, 244, 1550-1557.
Zaidel, E. (1979). Of apes and hemispheres. Commentary on the special issue on cognition and consciousness in nonhuman species. Behavioral and Brain Sciences, 4, 607-609.
Zaidel, E. (1983). A response to Gazzaniga: Language in the right hemisphere, convergent perspectives. American Psychologist, 38, 542-546.