Darwin, Tinbergen, and the Evolution of Comparative Cognition

Home , Comparative cognition, Comparative psychology, Tinbergen Lecture

OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

CHAPTER Darwin, Tinbergen, and the 2 8 Evolution of Comparative Cognition

Sara J. Shettleworth

Abstract Darwin and Tinbergen represent two enduring contrasts in comparative cognitive psychology: in the types of behaviors studied and the kinds of explanations sought. Darwin encouraged the search for human-like behaviors in animals as evidence for evolutionary continuity of mental processes. Tinbergen encouraged the careful causal analysis of animal behaviors as such and eschewed interpretations in terms of anthropomorphic processes. The Darwinian program has reemerged in contemporary research on comparative cognition. Its development and relationship to other areas of behavioral biology are traced. In using behavior as a window onto the animal mind, it is important to remember the lessons of Tinbergen and like-minded behaviorists in psychology. Several of the challenges that arise in attempting to show that other species share complex cognitive processes with humans are discussed in the light of the contrast represented by Darwin and Tinbergen, as are examples of how these approaches are being productively integrated. Key Words: Darwin, ethology, behavioral ecology, comparative psychology, history, cognitive ethology, associative learning

Introduction he set an agenda for studying animal minds that to Darwin and Tinbergen in the title of this chap- a remarkable degree is still being played out. By ter represent two enduring contrasts in comparative focusing on human-like behaviors in other species, cognitive psychology: in the types of behaviors stud- it encouraged explanations that were often danger- ied and the kinds of explanations sought. Th is chap- ously anthropomorphic in interpreting human-like ter begins with some historical background, from the behaviors as produced by human-like thought with- line of research that began with Darwin to a sketch out properly considering alternatives. Tinbergen of how several subﬁ elds in the biology of mind and represents the big chunk of the twentieth century in behavior are converging in the contemporary com- which the prevailing approach in both biology and parative study of cognition. Th is convergence is a psychology was a reaction to anthropomorphism source of much that is exciting and new, but it can in its devotion to careful causal analyses of animal also be a source of misunderstanding and contro- behavior as such, much of it not terribly human- versy. Some of the challenges that result have their like but nonetheless important in the lives of the roots in the contrasting approaches identiﬁ ed in the creatures concerned—homing in wasps or court- title, as discussed in the second half of the chapter. ship in gulls and sticklebacks. Recently there has Darwin is in the title because, in Th e Descent of been a big swing back toward the Darwinian pro- Man and Selection in Relation to Sex (Darwin, 1871), gram of looking for human-like cognitive abilities

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 1 99/22/2011/22/2011 12:23:1912:23:19 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

in other species—tool use, planning (chapters 12 “killjoy” behaviorist, or “higher vs. lower” expla- and 13 of this volume), social deception, and so on. nations for behavior. Th is tension was already well Th is approach is fraught with old intractable prob- developed when people began studying animal cog- lems. Some of them arise because, in the excitement nitive processes in the late nineteenth century, as of using behavior as a window onto the mind, it’s discussed in the next section. easy to forget the lessons of Tinbergen and Skinner about the importance of present cues and past his- Th e Evolution of Comparative Cognition tory in controlling it. Darwin and the Anthropocentric Approach Th e contrast represented here by Darwin and In On the Origin of Species, Darwin (1859) Tinbergen is sometimes referred to as that between largely steered clear of the touchy topic of human anthropocentric and ecological approaches (Kamil evolution. However, 12 years later, in Th e Descent & Mauldin, 1988; Shettleworth, 1993), that is, of Man and Selection in Relation to Sex, (Darwin, the study of animal behaviors in relation to human 1871), man—or humanity—was in the forefront, behavior vs. the study of behavior in its ecological and right up front he devoted two chapters to com- and evolutionary context. Of course, as a keen natu- paring human mental powers to those of other spe- ralist, Darwin recorded many observations of behav- cies. At the beginning of chapter 2 he writes…. “My ior that are simply fascinating in their own right, as object in this chapter is solely to shew that there in his experiments on the sensory abilities of earth- is no fundamental diff erence between man and the worms (Darwin, 1881). Nevertheless, as discussed higher mammals in their mental faculties” (Darwin more in a moment, his role in the history of com- 1871, p. 35). Even though he acknowledges that parative cognition was to focus it on human-like “no classifi cation of the mental powers has been accomplishments of other animals. Tinbergen and universally accepted” he does a great job of provid- other ethologists, in contrast, were primarily con- ing one. Th e topics in chapter 2 include “Certain cerned with how animals do what they do in nature. instincts in common, emotions, curiosity, imitation, Indeed, one of the ways in which early ethologists attention, memory, imagination, reason, progressive defi ned their fi eld as the biological study of behavior, improvement, tools, and weapons used by animals, distinct from animal psychology, was by its focus on language, self-consciousness, sense of beauty, belief natural behaviors of diverse species—insects, birds, in God, spiritual agencies, superstitions.” In chapter and fi sh as much as mammals (Burkhardt, 2005). 3 he goes on to “the moral sense and the qualities of In addition, Tinbergen (1963) famously identifi ed social animals.” Maybe minus the part on religion, four kinds of questions that can be asked about any this could be a tour of contemporary research. behavior: its proximal cause, as in the cues that elicit Darwin concludes his review of the evidence as it; its developmental history; its current function, follows: “Th e diff erence in mind between man and as in its contribution to survival and reproduction; the higher animals, great as it is, is certainly one of and its evolution. Th e distinctions between these degree and not of kind.” (Darwin, 1871, p. 105). questions need to be kept clearly in mind so that, Proving this had to be central to Darwin’s agenda for example, understanding the function of a behav- because it seems to clinch the argument that humans ior is not mistaken for understanding its proximal are part of the same evolutionary tree as all the other cause, but Tinbergen emphasized that a complete animals. Th is is an anthropocentric agenda, but understanding of any behavior includes answers to Darwin is often accused (e.g., by Bolhuis & Wynne, all of them. Moreover, the answer to one may illu- 2009) of being anthropomorphic as well, in the minate answers to the others (Sherry, 2005). sense of interpreting animal behaviors as refl ecting Much of the richness and excitement that char- thoughts similar to those a person would have when acterizes the study of animal or comparative cogni- engaging in comparable behaviors. Because his goal tion in the early twentieth century has come from was to prove the anthropomorphic case, this is not the participation of people who bring knowledge of surprising, but often Darwin was admirably cir- diff erent species and study them from diff erent per- cumspect. For example, in a passage foreshadowing spectives, but the contrasts symbolized by Darwin contemporary speculations that animals engage in and Tinbergen endure. Th ey are expressed, not only mental time travel (chapters 12 and 13 of this vol- in the pervasive tension between anthropocentric vs. ume) he writes, “But can we feel sure that an old dog ecological or psychological vs. biological approaches with an excellent memory and some power of imagi- to choice of species and problems, but in a perhaps nation, as shewn by his dreams, never refl ects on his more basic tension between anthropomorphic vs. past pleasures in the chase?” (p. 62, Darwin, 1871).

2 darwin, tinbergen, and the evolution of comparative cognition

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 2 99/22/2011/22/2011 12:23:2012:23:20 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

He thus raises the possibility of what we would now “lower,” and accepting the principle of parsimony call episodic memory without claiming we can feel in psychological processes (Mitchell, 2005; Sober, sure from the dog’s behavior that he is refl ecting on 2005). One reasonable modern interpretation of his past pleasures. Elsewhere, with regard to the pos- the Canon derives from the widely accepted princi- sibility that animals have abstract concepts, he refers ple of cladistic parsimony (see Sober, 2005): the fact to “the impossibility of judging what passes through that simple forms of memory and associative learn- the mind of an animal” (Darwin, 1874, p. 83). ing have been found in all species tested, including Some of Darwin’s most enthusiastic followers even simple invertebrates (Papini, 2008) justifi es were not always so circumspect. Gathering anec- claims that these processes evolved very early, per- dotes about clever animals seemed to be an impor- haps refl ecting universal causal processes and/or tant way to support Darwin, and some of their properties of nervous systems, and, therefore, are excesses of anthropomorphism had an important present throughout the animal kingdom. Th us, no role in the early history of comparative psychol- special evidence should be required to invoke them ogy (Boakes, 1984). Anecdotes about animals who for explaining behavior of a previously unstudied could open gates and doors were prominent here, species such as New Caledonian crows or bonobos. and they invited anthropomorphic explanations, as However, one can just as reasonably use such when Romanes (1892, p. 421) described how his phylogenetic reasoning to argue that species suf- coachman’s cat came to open a door. “First the ani- fi ciently closely related to humans should share mal must have observed that the door is opened by cognitive processes other than associative learning the hand grasping the handle and moving the latch. such as reasoning and imagination (Sober, 2005). Next she must reason, by ‘the logic of feelings’—If Evolutionary thinking also supports suggestions that a hand can do it, why not a paw?” Th e tendency even species distantly related to humans and apes to jump to conclusions about how a clever-looking but facing similar environmental demands may have behavior has developed in a single subject or what convergently evolved similar cognitive mechanisms mechanisms underlie it without much evidence one (Emery & Clayton, 2004). Nevertheless, the default way or another is still with us. For example, the in contemporary research continues to be the killjoy now world-famous chimpanzee in a Swedish zoo assumption that, in the absence of good evidence to who regularly piled up stones in the off hours and the contrary, behavior should be explained by asso- threw them at spectators when the zoo was open ciative learning plus species-typical behavioral pre- is said to plan (Osvath, 2009). Planning suggests dispositions (Wasserman & Zentall, 2006). Given that gathering stones and piling them on the side of the ambiguities of even a modern interpretation of the enclosure nearest the spectators developed after Morgan’s canon, a better-justifi ed approach may the animal discovered the rewarding eff ect of throw- be evidentialism: don’t accept “lower” or “higher” ing the stones. But the piles of stones and the way explanations of behavior without good evidence in which the chimpanzee replaced them when they (Fitzpatrick, 2008; Sober, 2005). Accordingly, as had been removed were apparently not documented discussed later in this chapter, one of the biggest until after stone throwing became a problem for the challenges for contemporary research is formulating zoo keepers. Th e possibility that the animal origi- unambiguous tests for alternative explanations of nally began collecting stones for some reason other behaviors suggestive of “higher” processes, such as than planning to throw them is now untestable, but mental time travel, metacognition, theory of mind, it is dismissed nonetheless (see also Suddendorf & and physical understanding (Heyes, 2008). Corballis, 2010). E. L. Th orndike (1911/1970; Galef, 1998) was Th e excesses of Darwin’s supporters stimulated a pioneer in testing alternatives to anthropomor- a killjoy backlash, with eff ects felt through most phic explanations. His studies of how dogs and of the twentieth century. Th ese include Lloyd cats learned to escape from “puzzle boxes” were Morgan’s canon (Morgan, 1894, p. 53): “In no case directly inspired by anecdotes like those purveyed may we interpret an action as the outcome of the by Romanes. His contribution was to do experi- exercise of a higher psychical faculty, if it can be ments and to show that performances like open- interpreted as the outcome of the exercise of one ing latches can be accounted for by simple trial which stands lower in the psychological scale.” Th e and error learning, with no evidence for insight or problems with Morgan’s canon include its assump- imitation. Subsequently, with the notable excep- tion of a scala naturae of “psychical faculties,” the tions of Kohler’s (1925/1959) studies of apes and implication that we can judge what is “higher” and Tolman’s (e.g., 1948) writings, the tendency to look

shettleworth 3

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 3 99/22/2011/22/2011 12:23:2012:23:20 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

for human-like thought and reasoning in animal independently. In the early 1970s, some of these problem solving went largely underground for most researchers began trying to catch up with the cog- of the rest of the twentieth century. In the history nitive revolution that was sweeping the study of of comparative psychology, the resulting emphasis human psychology. Cognition in this context refers on studying causes of behavior without speculating to the mind as a receiver and processor of informa- about internal mental processes is usually identifi ed tion, whether conscious or not. Indeed, until more with the growth of behaviorism (Boakes, 1984), but recently, consciousness was not an issue even in it was just as characteristic of ethology. For exam- research on human cognition, which looked mostly ple, at the very beginning of Th e Study of Instinct at input-output relations and treated the mind as a Tinbergen writes, “Because subjective phenomena black box or a computer. Most of the early leaders of cannot be observed objectively in animals, it is idle the revolution in animal experimental psychology either to claim or to deny their existence. Moreover were trained in good behaviorist methods, which to ascribe a causal function to something that is they now began using to test rats or pigeons for pro- not observable often leads to false conclusions.” cesses being studied in humans. Th e book Cognitive (Tinbergen, 1951, p. 4). Processes in Animal Behavior (Hulse, Fowler, & Honig, 1978) that is often seen as proclaiming the Th e Development of Evolutionary beginning of the fi eld of comparative cognition Comparative Cognition included a pretty narrow range of topics: condi- A few years after Tinbergen was writing, we began tioning, memory, attention, serial learning, space, to hear about something called cognitive ethology time, concepts. Th e range of species was even nar- from the distinguished biologist Donald Griffi n rower: rat, pigeon, chimpanzee. Th is research was (1976; 1978). “Th e basic goal of … cognitive ethol- thoroughly anthropocentric. Encouragement from ogy will be to learn as much as possible about the the growing fi eld of behavioral neuroscience helped likelihood that nonhuman animals have mental to perpetuate this focus, as paradigms for testing experiences…..” (Griffi n, 1978, p. 528), in other memory and the like provided “animal models” for words, to bring animal consciousness into ethology. use in studies of neurobiology and ultimately clini- Although Griffi n updated the information inform- cal applications (chapter 14 this volume). ing his position until almost the end of his life Even as this research developed, however, some (Griffi n & Speck, 2004), his emphasis on mental- psychologists complained, as others had in the past istic interpretations of animal communication and (Beach, 1950; Hodos & Campbell, 1969), that other behaviors found many detractors (e.g., Yoerg “comparative” cognition research was not genuinely & Kamil, 1991). After a couple of lively debates in comparative. Th ey called not only for looking at more Th e Behavioral and Brain Sciences (Dennett, 1983; species but also for studying cognitive processes used Griffi n, 1978) cognitive ethology as such was short to solve information processing problems in nature. lived (see Allen, 2004), inspiring one published A leading example of this synthetic (Kamil & conference (Ristau, 1991a) and some new research Mauldin, 1988) or ecological (Shettleworth, 1993) on piping plovers (Ristau, 1991b). Arguably, how- approach was the comparative study of spatial ever, Griffi n’s exhortations to animal behaviorists to memory in birds that do and do not store food. throw off the shackles of behaviorism and consider Wild birds were studied both in the fi eld and in the possibility of human-like thinking even in assas- standard laboratory paradigms to test the hypoth- sin bugs have stimulated much of the research dis- esis that reliance on stored food is associated with cussed in this Handbook. precise long-lasting spatial memory. Th is research To some (Allen, 2004; Kamil, 1998), cognitive soon converged with studies of the neurobiology of ethology ought to mean something rather broader spatial memory and the hippocampus to become a than anthropomorphic considerations of animal leading example of so-called neuroecology, or the consciousness, namely the role of cognition as comparative study of brain mechanisms in relation information-processing in the kinds of behaviors to diff erences in ecology (Brodin, 2010; Sherry, of interest to ethologists, what is now referred to 2006). As species that might be expected to show as cognitive ecology (Dukas & Ratcliff e, 2009; evidence of episodic memory and planning, food- Real, 1993). Cognition as information processing storing birds have subsequently contributed to was what psychologists studying animals were get- research on more general questions about memory ting interested in at about the same time as Griffi n and its neural substrate (chapter 12 of this vol- began promoting cognitive ethology, but quite ume; Pravosudov & Smulders, 2010). Similarly,

4 darwin, tinbergen, and the evolution of comparative cognition

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 4 99/22/2011/22/2011 12:23:2012:23:20 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

another specialized behavior shown by birds in the learning (Galef & Heyes, 2004; Heyes & Galef, fi eld—tool using by New Caledonian crows and 1996; Laland & Galef, 2009) are also subareas of woodpecker fi nches—has led to experimental anal- evolutionary comparative cognition with their own yses of more general processes, in this case mecha- thriving interdisciplinary research communities and nisms involved in solving physical problems, possibly conferences. Finally, as evidenced by several chap- including insight and physical understanding (Bluff , ters in this Handbook, long-term observations and Weir, Rutz, Wimpenny, & Kacelnik, 2007). clever fi eld experiments on primates and some other Another strand in the developing synthetic or, species (e.g., Cheney & Seyfarth, 2007; de Waal & to adopt the title of this Handbook, evolutionary Tyack, 2003) have raised important questions about comparative approach, was the growth of behav- the cognitive processes involved in communication, ioral ecology, the subfi eld of ethology devoted to tool use, social transmission of behavior, and so on. Tinbergen’s questions about function and evolu- Laboratory studies of chimpanzees and other pri- tion, now incorporating mathematical models of mates suggested by such observations sometimes the eff ects of behavior on fi tness (Cuthill, 2005; include human children, thus addressing shared Danchin, Giraldeau, & Cezilly, 2008). Perhaps the mental powers very directly (e.g., Herrmann, Call, biggest success in the early days of behavioral ecology Hernández-Lloreda, Hare, & Tomasello, 2007). was optimal foraging theory (see Krebs & Davies, Indeed, the recent explosion of data on all aspects of 1981). Predictions about fi tness-maximizing deci- animal cognition underpins several recent reexami- sions concerning where to forage and for how long, nations of Darwin’s claim that humans diff er men- what to eat, and so on were often tested in situ- tally in degree but not in kind from other species ations resembling psychological studies of learning (Penn, Holyoak, & Povinelli, 2008; Premack, 2007; and choice. Th is resemblance was exploited in some Tomasello, Carpenter, Call, Behne, & Moll, 2005). productive collaborations between behavioral ecol- As a result of all of these developments, the ogists and experimental psychologists (Kacelnik, study of comparative or animal cognition in the Brunner, & Gibbon, 1990; Shettleworth, Krebs, broadest sense now embraces three major sets of Stephens, & Gibbon, 1988). One common fi nd- cognitive processes about equally, addressing them ing was that, although observed behavior might in species from ants to chimpanzees and humans conform roughly to predictions of foraging theory, (Shettleworth, 2010). Basic mechanisms of percep- psychological mechanisms such as variable percep- tion, memory, associative learning, discrimination tion and memory for time intervals did a better job learning, and categorization cut across all kinds of of accounting for the details. More recently, some content. Although comparative psychologists have of this research has converged with behavioral eco- been studying some of them since Th orndike’s nomics in looking at simple mechanisms of eco- time, research in these areas continues to develop. nomic decision making that may cut across species Understanding these basic mechanisms is essential (Kacelnik, 2006; Rosati, Stevens, Hare, & Hauser, to appreciating how cognition in specifi c physical 2007; Schuck-Paim, Pompilio, & Kacelnik, 2004). or social domains may or may not be specialized. Other subfi elds in the broad study of animal Th e latter processes of acquisition, representation, behavior have embraced or converged with areas of and behavioral control, or cognitive modules, are comparative cognition. Cognitive ecology (Dukas defi ned largely in terms of their functions, what & Ratcliff e, 2009; Real, 1993) has already been aspects of the world they are about. Physical cog- alluded to as the study of how cognitive mecha- nition includes time, space, number, and instru- nisms evolve and are used in their natural con- mental learning, topics that have been studied for text. Similarly, sensory ecology (Dusenbery, 1992; quite a long time within experimental psychology, Endler & Basolo, 1998) is the long-standing and as well as tool using and planning. Social cogni- well-developed study of animal sensory systems in tion includes social knowledge, that is, what ani- relation to ecology. Th e development, use, and neu- mals know about their social companions and how robiology of bird song is a subject large and lively they come to know it, and the cognitive processing enough for its own conferences and books (e.g., evident in animal communication systems, top- Marler & Slabbekoorn, 2004; Zeigler & Marler, ics discussed in several chapters in this Handbook. 2004) but with many links to broader issues in Th us, compared to 25 or 30 years ago, the whole comparative cognition, especially language evo- fi eld of comparative cognition encompasses a much lution (Margoliash & Nusbaum, 2009). Spatial broader set of problems, being studied in a very behavior (Gallistel, 1990; Jeff ery, 2003) and social wide sample of species. Th ey are also being studied

shettleworth 5

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 5 99/22/2011/22/2011 12:23:2012:23:20 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

by a very diverse community of researchers—people behaviors and independent variables are related in trained not only in psychology but behavioral ecol- the same way across species. Obtaining such data ogy, ethology, anthropology, sometimes human may require testing humans in novel ways. Th us, cognitive psychology, and child development, not to discover whether monkeys and pigeons show to mention philosophy. Th is diversity undoubtedly serial position eff ects in memory for lists of visual contributes to making the fi eld so rich and interest- items, the animals were presented with four trial- ing, but it also can make for misunderstanding and unique images in sequence. Memory was probed controversy. Some of the challenges are discussed in with a single item, and reinforcement was given the next part of this chapter. for correctly indicating whether it had been in the list or not. Because the items had no meaning for Challenges the animal subjects, comparable data were obtained From Anthropomorphism from human subjects by testing them with kaleido- to Behavioral Tests scope patterns. All three species and others tested When a chickadee stores a sunfl ower seed, is it later show dynamic serial position eff ects in serial planning ahead? When a vervet monkey emits an probe recognition tests, with memory for the most alarm call, does it want to inform its companions a recent items dominating at short retention intervals leopard is near? More generally, when another animal and memory for the earliest ones at longer intervals. does something that looks human-like, is that behav- Although the precise retention intervals at which ior evidence for the same cognitive processes under- recency and primacy are seen vary with species, the lying analogous behavior in people? How would we common pattern is evidence for a common memory tell? Th is last question is especially challenging when process (Wright, 2006). Here, nothing need be said dealing with processes documented in humans pri- about consciousness. marily by verbal report and assumed to be associated with particular states of consciousness (Heyes, 2008; Metacognition and awareness Suddendorf & Corballis, 2010). Th e challenges Th e comparative study of metacognition, or range from minimal to virtually intractable. Many awareness of one’s own cognitive processes (see are illustrated by research discussed in other chapters chapters 12 and 15 of this volume) contrasts with of this Handbook, which can be consulted for thor- the comparative study of serial position eff ects ough reviews of topics touched on here. in that it is addressed to a process that is usually conscious in humans. Metacognition, strictly Animal memory and functional speaking, implies second-order representation, similarity cognition about cognition, rendering controversial People often report what they remember by talk- any claim to demonstrate it in a nonhuman spe- ing about it, but all species exhibit memory non- cies (e.g., Carruthers, 2008; Penn et al., 2008). As verbally. In general, if behavior depends on past with serial-order eff ects, studies of metacognition experience of a particular object or event, the ani- include numerous reports of identical patterns of mal is said to have a memory for it. For instance, data from human and nonhuman (usually rhesus Darwin (1871) supported his argument for long- monkey) subjects in identical tests (e.g., Kornell, lasting memories in animals with the observation 2009; Shields, Smith, Guttmanova, & Washburn, that when he returned home from his fi ve-year 2005; Smith, Shields, Allendoerfer, & Washburn, voyage on the Beagle, his dog greeted him as if he 1998; Smith, Shields, Schull, & Washburn, 1997). had never been away. Some contemporary research Many such tests involve near-threshold perceptual similarly exploits the subtle and detailed memories discriminations in which subjects may either clas- that animals form spontaneously, as in rats’ memory sify stimuli into one of two categories or opt out for the location and context where they encoun- of trials by choosing a third, “uncertain,” response. tered particular objects (Eacott & Norman, 2004). Monkeys, dolphins (Smith et al., 1995), pigeons Even when testing behaviors requiring extensive (Sole, Shettleworth, & Bennett, 2003), and people training, as in comparisons of serial position eff ects typically make this third response most often when across monkeys, pigeons, and people (Wright, the stimulus is near the threshold of discriminabil- 2006), the goal is to examine functional similarities ity. Th is pattern can be explained by signal-detec- among patterns of data resulting from comparable tion theory and/or responding based on the relative manipulations (Heyes, 2008). Looking for func- probabilities of reward for the three options at each tional similarities means looking for evidence that point along the stimulus continuum (Hampton,

6 darwin, tinbergen, and the evolution of comparative cognition

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 6 99/22/2011/22/2011 12:23:2012:23:20 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

2009; Jozefowiez, Staddon, & Cerutti, 2009; Smith, called what-where-when memories (see Crystal, Beran, Couchman, & Coutinho, 2008; Sole et al., 2010). Accordingly, when Clayton and Dickinson 2003). Th us, although people say they feel uncer- (1998) reported the fi rst demonstrations of such tain when they opt out of trials, the functionally memory, using Western scrub jays’ memory for their similar behavior of other species need not be medi- food caches, they scrupulously referred to the birds ated by a conscious state of uncertainty. Turning the as having “episodic-like” memory. Clayton and col- argument around, on some accounts, human meta- leagues went on to show that the birds’ memory for cognitive responses may not be mediated by con- what they had cached where and how long ago had sciousness either (Carruthers, 2008; Koriat, Hilit, & other properties attributed to episodic memory. For Nussison, 2006). instance it is a single integrated representation of Th e existence of “simpler” explanations here the caching episode that can be used in a fl exible does not rule out the possibility that some sort of way (see chapter 12 of this volume). metacognitive awareness plays a role since both are However, not all agree that the essential features compatible with the observed results (Hampton, of episodic memory are well captured by the stud- 2009). Arguments like those just sketched have, ies with scrub jays. For instance, the animals were therefore, challenged investigators to devise other trained, albeit not very extensively, to expect certain tests to isolate evidence of processes closer to what is kinds of tests, making the paradigm essentially one implied by metacognition in people, that is, to fi nd of conditional discrimination learning (Roberts, a situation in which the anthropomorphic “higher” 2002). How long ago a caching episode had explanation predicts a diff erent outcome from one occurred was a cue to what items would be edible based on sensitivity to reward rates and external (or at test, whereas “when” in human episodic memory public, Hampton, 2009) cues. Th e current best can- is typically a defi nite time in the past such as “last didate is the test of metamemory used by Hampton Monday.” In a paradigm logically equivalent to that (2001) with rhesus monkeys (see Hampton, 2009), developed by Clayton and Dickinson, rats indeed although some claim that other approaches qualify encode “how long ago” rather than “when” (time of (e.g., Smith, Beran, Redford, & Washburn, 2006). day; Roberts et al., 2008), but they can also use time In any case, since we cannot get from other species of day when required to (Zhou & Crystal, 2009). It the sorts of verbal reports that we commonly rely on has also been argued (e.g., by Eichenbaum, Fortin, for evidence of consciousness in humans, cleverly Ergorul, Wright, & Agster, 2005) that the ”when” delineating functional similarities may be as far as in episodic memory need not be the time at which we can go. Even in the rare cases when other forms something occurred but its entire spatio-temporal of behavioral control are ruled out, we still can- context. Accordingly, a rat’s memory for the loca- not know how closely other species’ mental states tion of an object in a particular spatial context resemble our own (see e.g., Hampton, 2005; Heyes, (Eacott & Norman, 2004) qualifi es as episodic-like, 2008). as does memory for the position of an odor in a sequence of odors (Eichenbaum & Fortin, 2005). Episodic memory: The challenge Performance in tasks of the latter type, which of multiple definitions exploit rats’ outstanding ability to discriminate and When other species cannot be tested in the same learn about odors, shares an impressive number of way as humans, new challenges arise. Research on other properties with verbal memory, including episodic memory in animals (see chapters 12 and 13 eff ects of hippocampal lesions (Fortin, Wright, & of this volume) provides examples. Episodic mem- Eichenbaum, 2004). ory was originally (e.g., Tulving, 1972) defi ned as Another issue is that with human subjects epi- memory for one’s own experiences, as distinct from sodic memory is typically probed unexpectedly, so memory for facts and ideas (semantic memory). subjects cannot prepare for the test at the time of But the defi nition evolved to include autonoetic con- memory encoding. Accordingly, some clever tests sciousness, the sense of reexperiencing an episodically with pigeons have shown that they can “answer unex- recollected event, mentally traveling back in time pected questions” about what they just did (Zentall, to it (Tulving, 2005). Since we are unlikely ever to Clement, Bhatt, & Allen, 2001). However, because know if animals travel anywhere mentally, com- these tests probe only very short term memory, they parative researchers have fallen back on the original in turn lack another property generally attributed defi nition and sought evidence that animals have to episodic memory, namely, that it is a species of integrated memories for unique events, sometimes long-term memory (cf. Hampton & Schwartz,

shettleworth 7

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 7 99/22/2011/22/2011 12:23:2112:23:21 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

2004). As another approach, it may be possible to psychology to guide the formulation of behavioral study naturalistic examples of the spontaneous use defi nitions. As a result, the crucial features distin- of episodic-like memory that do not involve food guishing behavior based on some sort of representa- caching. One candidate is the male meadow vole’s tion of the future as such from other kinds of future ability to recall where and when he encountered a oriented behavior are still debated. Suddendorf and female about to come into estrus so he can time his Busby (2005) proposed that a behavior that was next visit appropriately (Ferkin, Combs, delBarco- novel and that was performed in the service of a Trillo, Pierce, and Franklin, 2008). In any case, a motivational state other than the one the animal wide variety of tests for episodic memory is used by was in at the moment, as in gathering food when students of human cognition. Studies of compara- not hungry, would be evidence of future planning. tively short-term memory for stimuli encountered But with the accumulation of fi ndings fi tting these in the laboratory may be labeled studies of episodic criteria, some have seen a need for additional speci- memory (Kohler, Moscovitch, & Melo, 2001) fi cations. For instance, domain generality (Premack, equally with studies of subjects’ rich recollections 2007) rules out adaptations of particular behavior of personal experiences (Addis, Wong, & Schacter, systems, such as food storing. Being performed on a 2007). One potential contribution of research with single test (Suddendorf & Corballis, 2010) rules out nonverbal species is to focus attention on behavioral gradual learning through delayed reward. Specifi city criteria for episodic memory in humans (Crystal, to a particular future time (Roberts & Feeney, 2009) 2010). Th e best current conclusion from compara- implies a conception of the future as such. tive research seems to be that memory in nonhu- Such criteria seem to rest more on appeal to man species shares numerous features with episodic folk psychological, anthropomorphic, conceptions memory in humans, but no single example or spe- of planning than on parallel data from humans. cies necessarily shares all of them, and autonoetic Similarly, in a purported demonstration of planning consciousness may be unique to humans (Crystal, by apes (Mulcahy & Call, 2006), the animals had 2009; Suddendorf & Corballis, 2010). multiple opportunities to collect a tool that could not be used right away, but in most cases successful Future planning, tool use, trials were scattered more or less randomly across all and folk psychology trials. Folk psychology would seem to suggest that Memory presumably evolved not so creatures once an animal had discovered the rules of the game could ruminate about their pasts but so they could it would plan on every trial, so is this or is this not use past experience to determine future behavior. evidence of future time travel (Shettleworth, 2010; Reasonable as this idea seems, only recently has it Suddendorf & Corballis, 2008)? In any case, folk been incorporated into conceptions of human epi- psychology is not necessarily a good predictor of sodic memory. Planning, or mental time travel into how people behave when tested like other animals. the future, is now seen as the fl ip side of mental time For instance, when people are given tests similar to travel into the past. Both involve autoneotic con- those failed by apes, they may behave irrationally sciousness and depend on closely similar parts of the too. Human subjects avoid a tool that brings a brain (Addis et al., 2007). Research with other spe- reward close to a hole or trap, even when it will not cies has followed, with attempts to demonstrate that fall in (Silva, Page, & Silva, 2005), and like apes animals show future planning (chapters 12 and 13 they may choose to pull a string contacting but not of this volume, and Suddendorf & Corballis, 2010). actually connected to a reward (Silva, Silva, Cover, Two problems bedevil this research. Th e fi rst is that Leslie, & Rubalcaba, 2008). animals have many kinds of behavior that prepare them for the future but which can be accounted for Varieties of Proximal Cause without assuming any representation of the future Translating anthropomorphism into unambigu- as such or even requiring specifi c past experiences ous behavioral tests challenges researchers to isolate (Roberts, 2002). Examples include migrating, the specifi c cognitive mechanism of interest from hibernating, storing food, building nests, respond- other possible causal factors for observed behavior. ing to learned signals about delays to food. Th e sec- Th is can be entirely straightforward and uncontro- ond is that relevant research on planning or future versial, as in research on nonverbal numerosity dis- time travel in people, including how it develops in crimination, that is, discrimination among arrays of young children, is at an early stage, leaving research- visual stimuli, strings of sounds, and so on, accord- ers studying animals with little more than folk ing to the numbers of items they contain. Here it

8 darwin, tinbergen, and the evolution of comparative cognition

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 8 99/22/2011/22/2011 12:23:2112:23:21 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

is necessary to rule out reliance on some correlated then allowed to tip up one tube so the reward slides feature such as the amount of space or time the items out. By bending over, the monkey can peer into the occupy (for review see Shettleworth, 2010). Only if tubes before choosing, a slightly eff ortful behavior discrimination remains accurate with novel items that is unnecessary when the opportunity to choose varying in shape, size, color, contour length, and a tube comes immediately after witnessing baiting. so on, are subjects said to be displaying evidence of However, when baiting is done behind an opaque sensitivity to numerosity. Th is need not mean that screen, the task becomes a test of metamemory: if control by nonnumerical features is entirely absent they are aware they have no memory of a recent (see Cantlon & Brannon, 2005), but the strategy of baiting, monkeys should more often look before establishing one sort of control by ruling out alter- choosing. Th e possibility that looking on trials with natives works well here. Sometimes, however, in the screen was acquired because it was rewarded the enthusiasm for using behavior as the readout of more frequently than not looking under this con- cognition, other kinds of proximal causes for a tar- dition was ruled out by familiarizing the monkeys get behavior can be overlooked. What Skinner and with the elements of the task, pulling and looking, Tinbergen taught us about the importance of past before introducing the opaque screen. When the history and present cues is not obsolete just because opaque screen was then introduced, most monkeys cognition is being tested. In the spirit of Tinbergen’s looked more often right away. four kinds of causes of behavior, even when looking But the proximal cause of looking here is not nec- only at proximal causes, there may be more than essarily awareness of memory strength, that is, of one kind of answer. metamemory. An old-fashioned ethological causal A classic example comes from an early experi- analysis readily shows that the monkeys’ behavior mental study of theory of mind in chimpanzees. can be accounted for in terms of confl ict between Th e animals could choose to beg for hidden food two competing behavioral tendencies (see Hampton, from one of two people when one had seen it hid- 2009). Of the two responses learned before the test, den and one had not. To test whether their choice pulling and looking, correct pulling leads to reward of the former individual was based on sensitivity to with a shorter delay. Because monkeys were well the person’s knowledge state (i.e., theory of mind), trained on this response initially, it is likely to be the animals were given a series of rewarded tests in elicited when baiting is seen. With baiting unseen or which one observer had a paper bag over his head, forgotten, the tendency to pull should be weaker, if a novel cue to ignorance (Povinelli, Nelson, & only because the normal cues that elicit it are absent. Boysen, 1990). Overall, the animals tended to beg Additionally, regardless of whether baiting was seen from the observer who could see, a result consistent or not, looking is rewarded only after the delay that with theory of mind. However, rewarded tests are is necessary to locate the baited tube. However, learning trials, and closer analysis showed that the when baiting is unseen and the tendency to pull is, animals chose randomly on the fi rst two tests but therefore, weak, looking will emerge as the stronger learned to use the novel cue during the next few tri- response. In this way appropriate information seek- als (Povinelli, 1994). Similarly, in recent studies of ing—metacognitive behavior in a sense—need not dogs’ sensitivity to human social cues, the possibil- depend on explicit awareness of memory strength. ity of learning during a short series of test trials may Th e same may be true of people in similar situa- be dismissed too readily (Udell, Dorey, & Wynne, tions (Carruthers, 2008; Hampton, 2009; Kornell, 2010). Th is is especially so because a plausible alter- 2009). Th e requirement for viewing tests of “higher” native to a preexisting or very early-developing cognitive capacities from the animal’s point of view, theory of mind in dogs is that during domestica- in terms of the cues present, their past histories, tion they have evolved an exceptional sensitivity to and the relative strengths of responses they elicit by and ability to learn about human behavior (Reid, virtue of training and/or evolution transcends these 2009). few examples. In such cases it may not be necessary Another example comes from a simple test for the to assume any representations of memory strength spontaneous use of metamemory devised by Call and or similar higher-order causal factors are present. Carpenter (2001) for apes and children and then used Th is is similar to the point made repeatedly by by Hampton and colleagues (Hampton, Zivin, & Povinelli and his colleagues about tests of theory Murray, 2004) with rhesus monkeys. In the latter of mind or physical understanding (e.g., Penn & version, the monkey watches as an experimenter Povinelli, 2007): animals may behave eff ectively places a treat in one of four horizontal tubes and is on the basis of representations of observable cues,

shettleworth 9

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 9 99/22/2011/22/2011 12:23:2112:23:21 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

without representing unseen mental or physical scorpions. As the pups mature (again indicated by causes. Examples from animal problem solving are their begging calls) adults bring scorpions that need discussed later in this chapter. more and more processing. Adults also engage in other time-consuming behaviors that ensure the Function and Mechanism and Diff erent young learn to catch and disable scorpions for Ways of Using Language themselves. Th at the young do in fact learn from Many interdisciplinary fi elds bring together the experience so provided was shown by the ben- people who use the same words in diff erent ways. efi cial eff ects of experimentally providing extra scor- In comparative cognition, this often means using pions. Th e adult meerkats’ behavior thus meets all terms that have both functional and mechanistic the accepted functional criteria for teaching, but for defi nitions, leading to disagreement and debate those to whom “teaching” means human pedagogy, when people understand diff erent things by them. labeling it as such is deeply controversial (Csibra, In cognitive science, this is the problem of confus- 2007; Premack, 2007). In the context of behav- ing functional and representational explanations ioral ecology, the adult meerkats’ behavior could be (Penn & Povinelli, 2009). In various guises, the described as governed by “rules of thumb” such as same problem is widespread in the study of animal “when hearing calls of very young pups, bring dead behavior. In Tinbergen’s writings and subsequent scorpions.” In ethological terms, the calls are sign discussions (Bolhuis & Verhulst, 2009), it is the stimuli, but either way the adults teach eff ectively confusion of function with cause. As discussed by in the absence of any representation of the pups’ John Kennedy (1992) in his insightful little book understanding or need to know. Th e New Anthropomorphism, such confusion is par- Sampling in optimal foraging theory also exem- ticularly insidious in behavioral ecology, where the plifi es the sort of term discussed by Kennedy tendency to label behaviors anthropomorphically— (1992) in that it has a functional meaning in a for example “rape,” “deception,” “strategy”—creates certain context but implies an anthropomorphic the illusion that they can be explained by the same mechanism. Classic models of optimal foraging conscious processes that accompany analogous indicate that, to choose the best food item or for- behavior in people. Even when such behaviors have aging patch, individuals should sample, gathering a formal defi nition within behavioral ecology, label- information that might be useful when conditions ing them with terms current in ordinary language change. even if that sometimes means abandoning and/or other fi elds can lead to unnecessary disagree- the currently best option. Th e notion that animals ments. One example comes from recent claims to might sample in this way seems to fl y in the face demonstrate animal teaching (Hoppitt et al., 2008; of the well-established fi nding from psychology Shettleworth, 2010). laboratories that animals strongly prefer immediate Teaching has had an accepted functional defi ni- rewards. In terms of reinforcement theory, if patch tion in evolutionary studies of behavior for nearly A is currently paying off at the highest rate avail- 20 years (Caro & Hauser, 1992). To qualify as able, animals should never choose patch B except teaching, an animal must modify its behavior in the by mistake. Shettleworth, Krebs, Stephens, and presence of naïve individuals in a way that makes Gibbon (1988) pitted these accounts against each possible or speeds up their learning, and it must do other in an operant study with pigeons that simu- so at some immediate cost to itself. Th is last criterion lated two foraging patches, one of which was stable suggests that, as seems to be the case, teaching may and one of which changed abruptly in payoff rate be rare but that, when it is seen, those taught will from time to time, sometimes becoming much bet- be related to the teacher, thus eventually conveying ter than the stable patch. Th e optimal strategy is to a fi tness benefi t to the teacher. Notice that this defi - sample this fl uctuating patch with a specifi able fre- nition implies nothing about whether the teacher quency depending on such things as the reward rate understands the pupils’ need to learn, only that in the stable patch (Stephens, 1987). Th e birds did, some cues from potential pupils elicit appropriate in fact, sample at roughly the predicted frequency and costly behavior. Th is is exactly what was shown in several experimental conditions, but their behav- in an elegant demonstration that meerkats teach ior was best accounted for by a general model of their young how to handle scorpions (Th ornton & instrumental choice (Gibbon, Church, Fairhurst, & McAuliff e, 2006). When pups are very young (and Kacelnik, 1988), in which animals always choose hence inexperienced), as indicated by their begging the option perceived to predict the shortest delay calls, adults present them with dead or disabled to reward based on a fuzzy memory of the delays

10 darwin, tinbergen, and the evolution of comparative cognition

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 1010 99/22/2011/22/2011 12:23:2112:23:21 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

associated with each option present. Th us animals sight of a predator refl exively elicits calling. One of can “sample” approximately optimally but without Dennett’s central points was that because, in princi- doing so in the anthropomorphic sense of deliber- ple, they can be put to the test with experiments and ately rejecting the shortest current delay to reward opportunistic observations, the mentalistic claims in order to collect information that is useful in the of the cognitive ethologists should not be dismissed long term. out of hand. Characterizing fi rst- and higher-order Th is same timing model of choice between rein- intentionality as “exciting” explanations, in contrast forcement schedules accounts for behavior in a with “killjoy bottom-of-the-barrel” zero-order or variety of other laboratory tests of optimal forag- refl exive explanations may have been no more than ing (e.g., Kacelnik & Bateson, 1996; Kacelnik et al., a device for emphasizing the important potential 1990). More recently the analysis of animal forag- contribution of cognitive ethology. However, the ing decisions has evolved to connect with studies of implication that complex or higher-order explana- economic decision making. In studies of analogues tions of apparently cognitively demanding behavior to the sunk costs eff ect and infl uences of irrelevant should be valued more than simpler ones encourages alternatives on choice, other species make the same a tendency implicit in some contemporary research sorts of “irrational” decisions as do humans, and reports to dismiss accounts based on simple learn- again the results can be explained by general mecha- ing mechanisms before thoroughly testing them. nisms of reinforcement and choice (Bateson, Healy, Research on the possibility of insight in ani- & Hurly, 2003; Kacelnik & Marsh, 2002). As dis- mals provides some illustrations. Controversy over cussed in depth by Kacelnik (2006), such work inte- whether animals solve problems insightfully dates grates the contrasting but interrelated approaches to from the earliest days of comparative psychology rationality in evolutionary biology, economics, and (Boakes, 1984). Th orndike’s (1911/1970) conclu- psychology. sion from his puzzle box experiments—that only trial and error learning was involved—was in turn Clever Animals and Killjoy Explanations put to the test by Wolfgang Kohler (1925/1959) in Explanations of apparently complex processes as his studies of chimpanzees using sticks or climbing the product of simpler ones, as in animal teaching on boxes to reach inaccessible food. Some animals and sampling behavior, sometimes seem to have a did appear to solve problems insightfully, but there diff erent status in comparative cognition than else- was much evidence for trial and error learning too where in biology. As Darwin so eloquently argued (Povinelli, 2000). Subsequent research suggested in Th e Origin of Species, the whole wondrous com- that experience with sticks, boxes, and the like con- plexity of organic evolution can be explained as the tributed to their later successful use (Birch, 1945; outcome of simple processes, here variation, selec- Schiller, 1957). Current interest in the possibility of tion, and inheritance. Similarly, collective behavior animal insight (e.g., Bird & Emery, 2009a; Heinrich, of animal groups as in the exquisite regulation of 1995) is thus one more swing of the pendulum. a honeybee colony or the coordinated movements Th e defi nition of insightful behavior most often of schooling fi sh, is increasingly well understood as referred to in this context is that formulated by the the result of local responses by myriads of individu- British ethologist W. H. Th orpe (1956): “the sudden als (Couzin, 2009). It is hard not to be in awe of production of a new adaptive response not arrived at how a perfectly air-conditioned termite mound can by trial behavior” (i.e., trial and error; Th orpe, 1956, be constructed without any instructions from a ter- p. 100; italics in original). Similarly, in human mite architect. However, explaining behaviors that problem solving (Sternberg & Davidson, 1995; look as if they require human-like thought in terms Weisberg, 2006) insightful solutions have three of simple processes such as associative learning and general properties: species-typical predispositions seems tantamount to denial of mental continuity between humans and 1. Th ey appear suddenly, accompanied by a other animals. Labeling such explanations as “kill- distinctive subjective experience of surprise and joy” originated in Dennett’s (1983) analysis of levels delight—the “aha moment” (Kounios & Beeman, of intentionality in cognitive ethology. In his prin- 2009). cipal example, a monkey’s alarm call might refl ect 2. Th ey usually appear after an impasse, that wanting its fellows to move away, wanting them to is, after a period of unsuccessful attempts—rather know a predator is approaching, or even higher lev- than by gradually homing in on the solution els of intentionality. Th e killjoy alternative is that the through an analytical approach (Weisberg, 2006).

shettleworth 11

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 1111 99/22/2011/22/2011 12:23:2112:23:21 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

3. Th ey involve restructuring the problem, that absent. In addition, in interleaved sessions without is, approaching it in a new way. the banana, they had been rewarded with food for pushing the box toward a spot at varying locations Other than “aha moments,” these criteria corre- on the wall. In the test, they looked back and forth spond well with what is implied by Th orpe’s defi - between banana and box at fi rst, appearing con- nition. Th e question with both animal and human fused, but then approached the box and pushed it subjects is whether behavior meeting these criteria toward the banana. involves a distinctive cognitive process, namely Directed pushing and climbing to peck contrib- insight. As originally conceived by Kohler and other uted to success through a few simple mechanisms Gestalt psychologists, insight consists of suddenly (Epstein, 1985). First, a basic principle of ethol- “seeing” the solution, a perceptual process like see- ogy is that various cues in a situation may control ing the alternative form of a reversing fi gure (Mayer, incompatible behaviors in ways that may change 1995). In contemporary discussions of human dynamically with behavior. Here, when the pigeons problem solving (e.g., Kounios & Beeman, 2009; fi rst confronted the banana and the displaced box, Weisberg, 2006) insight contrasts with the analyti- looking back and forth was the expression of two cal approach, examining possible solutions in the learned behaviors that were both elicited but could mind’s eye or in actuality until an eff ective one is not be completed. Because the birds had been found. People readily report whether they have used explicitly extinguished for fl ying and jumping at the insight or analysis to solve even simple verbal prob- banana, approach to the box and then pushing it lems (Bowden, Jung-Beeman, Fleck, & Kounios, soon predominated. Th e banana became the target 2005). Some contemporary students of animal of pushing in the absence of the spot because both problem solving (e.g., Bird & Emery, 2009b) have banana and spot had been associated with food, incorporated mental analysis into their concep- that is, through functional (also called mediated) tion of insight, but because we cannot ask animals generalization (Urcuioli, 2006). Finally, by pushing whether they mentally tried possible solutions or the box toward the banana, the “insightful” pigeons simply “saw” how to succeed, this is not only histor- produced the situation (box under banana) in which ically and comparatively inaccurate but behaviorally climbing had been reinforced. Th e latter process is meaningless. In any case, not all agree that subjec- what Epstein calls automatic chaining. Epstein (e.g., tive experience is the best guide to mechanism even 1985) claims that such combining of old behaviors in human problem solving (cf. Weisberg, 2006). in new ways underlies human as well as animal With both animal and human subjects, discov- creativity. Th is is a strong claim to be put against ering whether behaviorfi tting criteria 1–3 from the claims that insightful problem solutions involve a preceding list refl ects a special cognitive mechanism special process of insight or physical understanding. typically consists of ruling out alternative mecha- Taking Epstein’s claim seriously means deconstruct- nisms. For nonhuman species the generally accepted ing situations in which animals solve problems in alternative is associative learning, but the subtle ways apparently novel ways into elements and testing in which past learning together with species-typical whether experience with these elements can explain behaviors and present cues might generate complex the performances of interest. and apparently novel behavior are not always fully Several recent examples involve tool using by birds appreciated or rigorously tested. A well-known illus- of the crow family (corvids), both New Caledonian tration of how simple forms of learning and other Crows, which naturally make and use tools, and basic behavioral principles can produce apparently non-tool-using species, such as rooks and ravens, insightful behavior is the demonstration (Epstein, which can solve similar tasks when tested appro- Kirshnit, Lanza, & Rubin, 1984) that pigeons with priately. In one case (Bird & Emery, 2009a) rooks appropriate past experience solve a novel “banana were shaped to nudge a stone so it fell down a tube and box” problem, just as Kohler’s (1925/1959) and released food from an apparatus at the bottom. apes did, by moving a box and climbing on it to Remarkably, when stones were no longer provided reach a hanging (toy) banana. All pigeons in the at the mouth of the tube, rooks brought stones study had previously been rewarded with food for from nearby. What experience might contribute climbing onto the box and pecking the banana, to the apparently spontaneous fetching and drop- but the only birds that “insightfully” solved the ping of stones remains to be determined. A further problem had also been extinguished for jumping study using New Caledonian crows (von Bayern, and fl ying toward the banana when the box was Heathcote, Rutz, & Kacelnik, 2009) showed that

12 darwin, tinbergen, and the evolution of comparative cognition

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 1212 99/22/2011/22/2011 12:23:2112:23:21 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

once they are experienced pushing stones down the when only one potential tool is off ered, it should tube, these birds also bring stones to the apparatus, not be surprising if the animal tries to use that tool but they do not require previous experience see- in some way, especially if the tool is a stick and ing stones release the food. Birds that had pecked the animal has a natural or learned propensity to or used a stick to operate a similar apparatus fi tted manipulate sticks. Th us, tests presenting only one with a short tube became more likely to bring and option, in eff ect asking whether the animal uses the drop stones through the long tube. Exactly what the tool or not, may be less informative than tests with birds need to learn here remains to be determined. several diff erent tools, asking whether the animal For instance, they may learn that contact with the chooses the best tool for the job. In such a test with mechanism at the bottom of the tube releases food New Caledonian crows (Wimpenny et al., 2009), and for still-unknown reasons generalize this to pro- inexperienced birds did not always choose the cor- ducing contact with stones. rect length stick. Related research is ongoing with More extensive analyses of the experiences neces- primates as well as birds. Even with chimpanzees, sary for New Caledonian crows to use tools in novel how a task requiring tool use is presented may infl u- ways have been done with metatool use, that is, using ence whether the animals appear to understand the one tool to obtain another with which food can be solution or not. For instance, chimpanzees may be obtained. Metatool use has been claimed to be espe- able to avoid losing food into a trap when allowed cially human-like (see Taylor, Hunt, Holzhaider, & to slide it along with a fi nger but fail when required Gray, 2007), and various species of primates can to use a stick tool (Seed, Call, Emery, & Clayton, show it (see Wimpenny, Weir, Clayton, Rutz, & 2009; also see Girndt, Meier, & Call, 2008). Th e Kacelnik, 2009). Using a metatool eff ectively on examples sketched here indicate that progress in the fi rst opportunity, before the required chain of understanding how animals solve novel problems behavior could be primarily reinforced, has been is more likely from examining the contributions suggested to reveal insight or reasoning (Taylor, of simple learning mechanisms and species-typical Elliff e, Hunt, & Gray, 2010; Taylor et al., 2007). behavioral predispositions than from attempt- In the fi rst demonstration of such behavior (Taylor ing to prove the existence of reasoning or insight. et al., 2007), New Caledonian crows provided with Discovering how animals successfully interact with a stick too short to reach a reward used that stick to their social and physical worlds through such simple obtain a long stick, which they then used eff ectively. mechanisms should be as much as cause for joy as Notwithstanding the claim in the article’s title, this any validation of anthropomorphism. behavior was “spontaneous”; in fact, the birds were trained beforehand on several elements of the prob- Cognition from the Bottom Up lem: obtaining meat with the long stick, extracting Th e approach to tool use represented by the a long stick from a container, and failing to obtain study of the “insightful” pigeons illustrates a central meat with the short stick. But even though use of theme of this chapter: complex behavior often arises the short stick to obtain meat directly was extin- from simple elements. Th is bottom-up approach is guished during this pretraining, both the crows’ nat- becoming increasingly common in the comparative ural predisposition to handle sticks and stimulus (or study of cognition. Some of the oldest questions mediated) generalization from the long stick would about the cognitive abilities of nonhuman spe- have been expected to support the observed use of cies, such as “Do animals count” “Do they talk?” the short stick. No special understanding of meta- “Or have cognitive maps?” are being replaced by tools need be invoked. In an extension of this study questions such as, “What are the elements of this (Taylor et al., 2010), the same group of researchers cognitive ability?” “How are they shared across spe- compared two groups of birds with diff erent experi- cies and why?” Th is approach to episodic memory, ences of the elements in a sequence of two meta- planning, and metacognition has been mentioned. tools. Although they concluded that the successful Other examples include the study of numerical cog- birds in this study used causal reasoning rather than nition (Brannon, 2006; Nieder, 2005), language solely the mechanisms identifi ed by Epstein, further (Hauser, Chomsky, & Fitch, 2002; Margoliash & research is needed to test this conclusion. Nusbaum, 2009), same-diff erent conceptualization Epstein’s approach implies that in cognitive tests (Wasserman & Young, 2010), spatial behavior (see we need to make a careful ethological analysis of chapter 8 of Shettleworth, 2010, ), other-regarding the cues available and their past histories and/or the behavior (de Waal, 2008; Warneken & Tomasello, animal’s predispositions toward them. For example, 2009, chapter 20 of this volume), and theory of

shettleworth 13

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 1313 99/22/2011/22/2011 12:23:2112:23:21 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

mind (Call & Tomasello, 2008). Sometimes this in humans, but other unconscious processes lead to “bottom-up” approach means looking for shared more “animal-like,” less-rational outcomes such as neural mechanisms (de Waal & Ferrari, 2010), but preference for immediacy and other biases in eco- more often elemental cognitive processes are defi ned nomic decision making. Presented appropriately, at the level of behavior (Shettleworth, 2010). Either abstract transitive inference problems are solved way, this approach compares human cognition to by human subjects in the “stupid” way typical of that of other species in terms of an interplay of pigeons (Frank, Rudy, Levy, & O’Reily, 2005). shared and unique processes. In numerical cogni- Evidence that we respond unconsciously to simple tion, for example, even human babies share with social cues is widespread even though it violates our other species an ability to discriminate approxi- intuitions about how we behave (Goddard, 2009). mately among diff erent numerosities, with accuracy As one example, when images of eyes were “watch- dependent on their ratios as described by Weber’s ing” the box for contributions to a coff ee pool, aver- Law. However, precise discrimination among quan- age payments more than doubled over weeks when tities greater than about four appears to depend on fl ower images were present instead (Bateson, Nettle, number language and thus be confi ned to numer- & Roberts, 2006). Andin some cases, processes ate humans. A more sweeping claim is that, despite studied fi rst in other species have later been demon- all the cognitive abilities shared by humans and strated in people, using nonverbal tests. For instance, other animals, only humans have higher order or the ability of disoriented rats to reorient using the relational concepts (Penn et al., 2008). Only in the geometry of a surrounding space is shared not only light of so much new research on cognitive abilities with birds, fi sh, and monkeys, but also with young in nonhumans is it possible meaningfully to address children and, under appropriate conditions, adults such claims. (Cheng & Newcombe, 2005), leading to a proposal Further contributions to a bottom-up analysis that in human spatial cognition insights from ani- of cognition come from developments in the study mals should replace the anthropomorphic concept of human psychology that are revealing an unex- of the cognitive map (Wang & Spelke, 2002). Such pected role in human behavior for simple, largely developments seem to promise convergence of bot- unconscious, nonverbal processes, such as those tom-up analyses of human behavior with similar found in other animals. Such work contrasts with analyses of other species into a richer comparative the traditional Darwinian search for the human-like psychology more deeply rooted in an understanding in animals in being a search for the animal-like in of how fundamental evolutionary processes produce humans, but it is equally Darwinian. Although, as complexity from simple species-general elements. discussed at the beginning of this chapter, emphasis on demonstrating human-like mechanisms in ani- Future Directions mal behavior can be traced to Th e Descent of Man As this Handbook illustrates, these are exciting and Selection in Relation to Sex (Darwin, 1871), in times for comparative cognitive psychology, but Th e Expression of the Emotions in Man and Animals the fundamental tension represented in this chap- Darwin (1872/1965) emphasized the opposite. ter by Darwin and Tinbergen is still with us. Th e For instance, in the Introduction he writes, “With participation in the fi eld of people with diverse mankind some expressions, such as the bristling of backgrounds and interests maintains this tension. the hair under the infl uence of extreme terror, or It can be negotiated productively to the extent that the uncovering of the teeth under that of furious researchers with diff erent approaches continue to rage, can hardly be understood, except on the belief be engaged with one another. Th ere is a fi ne line that man once existed in a much lower and animal- between dismissing “higher” processes as excessively like form” (p. 19, (Darwin 1872/1965). Darwin’s anthropomorphic and dismissing “lower” ones as ideas about the origins and functions of emotional not conceivably capable of doing the job. A priori, expression are still being tested (e.g., Susskind et al., Morgan’s canon or evolutionary theory does not 2008). Many examples of unconscious and unex- necessarily dictate either. Solid evidence support- pected infl uences on human behavior come from ing one or another explanation for a given behavior classic evolutionary psychology (see chapter 4 of depends to some extent on what alternative expla- this volume). nations are entertained, and that often depends on In cognitive psychology, implicit memory and deep and imaginative understanding of the species automatic processing of some fundamental infor- under study. As emphasized in this chapter, it also mation (Hasher & Zacks, 1984) are well studied depends on an appreciation of the contributions

14 darwin, tinbergen, and the evolution of comparative cognition

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 1414 99/22/2011/22/2011 12:23:2212:23:22 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

made by ethology and basic learning theory as well Cantlon, J. F., & Brannon, E. M. (2005). Semantic congruity as the many contemporary fi elds that converge in aff ects numerical judgments similarly in monkeys and the comparative study of cognition. humans. Proceedings of the National Academy of Sciences (USA), 102, 16507–16511. Caro, T. M., & Hauser, M. D. (1992). Is there teaching in Acknowledgments nonhuman animals? Th e Quarterly Review of Biology, 67, I thank the graduate students in the Oxford University 151–174. Department of Zoology for the invitation to give the inaugu- Carruthers, P. (2008). Meta-cognition in animals: A skeptical ral Tinbergen Lecture in February 2010, on which much of this look. Mind & Language, 23, 58–89. chapter is based. My research has been supported for nearly 40 Cheney, D. L., & Seyfarth, R. M. (2007). Baboon metaphysics: years by the Natural Sciences and Engineering Research Council Th e evolution of a social mind. Chicago: University of Chicago of Canada. Press. Cheng, K., & Newcombe, N. S. (2005). Is there a geometric References module for spatial orientation? Squaring theory and evi- Addis, D. R., Wong, A. T., & Schacter, D. L. (2007). dence. Psychonomic Bulletin & Review, 12, 1–23. Remembering the past and imagining the future: Common Clayton, N. S., & Dickinson, A. (1998). Episodic-like memory and distinct neural substrates during event construction and during cache recovery by scrub jays. Nature, 395, 272–274. elaboration. Neuropsychologia, 45, 1363–1377. Couzin, I. D. (2009). Collective cognition in animal groups. Allen, C. (2004). Is anyone a cognitive ethologist? Biology and Trends in Cognitive Sciences, 13, 36–43. Philosophy, 19, 589–607. Crystal, J. D. (2009). Elements of episodic-like memory in ani- Bateson, M., Healy, S. D., & Hurly, T. A. (2003). Context- mal models. Behavioural Processes, 80, 269–277. dependent foraging decisions in rufous humming- Crystal, J. D. (2010). Episodic-like memory in animals. birds. Proceedings of the Royal Society of London B, 270, Behavioural Brain Research,215, 235–243. 1271–1276. Csibra, G. (2007). Teachers in the wild. Trends in Cognitive Bateson, M., Nettle, D., & Roberts, G. (2006). Cues of being Sciences, 11, 95–97. watched enhance cooperation in a real-world setting. Biology Cuthill, I. C. (2005). Th e study of function in behavioural ecol- Letters, 2, 412–414. ogy. Animal Biology, 55, 399–417. Beach, F. A. (1950). Th e snark was a boojum. American Danchin, E., Giraldeau, L. A., & Cezilly, F. (Eds.). (2008). Psychologist, 5, 115–124. Behavioural ecology. Oxford, England: Oxford University Birch, H. G. (1945). Th e relation of previous experience to Press. insightful problem-solving. Journal of Comparative Psychology, Darwin, C. (1859). On the origin of species. London: John 38, 367–383. Murray. Bird, C. D., & Emery, N. J. (2009a). Insightful problem solv- Darwin, C. (1871). Th e descent of man and selection in relation to ing and creative tool modifi cation by captive nontool-using sex. London: John Murray. rooks. Proceedings of the National Academy of Sciences, 106, Darwin, C. (1872/1965). Th e expression of the emotions in man 10370–10375. and animals. Chicago: Th e University of Chicago Press. Bird, C. D., & Emery, N. J. (2009b). Reply to Lind et al.: Insight Darwin, C. (1874). Th e descent of man and selection in relation to and learning. Proceedings of Th e National Academy of Sciences sex (2nd ed.). London: John Murray. (USA), 106, E77. Darwin, C. (1881). Th e formation of vegetable mould through the Bluff , L. A., Weir, A. A. S., Rutz, C., Wimpenny, J. H., & Kacelnik, action of worms with observations on their habits. London: A. (2007). Tool-related cognition in New Caledonian crows. John Murray. Comparative Cognition & Behavior Reviews, 2, 1–25. de Waal, F. B. M. (2008). Putting the altruism back into altru- Boakes, R. (1984). From Darwin to behaviourism. Cambridge, ism: Th e evolution of empathy. Annual Review of Psychology, England: Cambridge University Press. 59, 279–300. Bolhuis, J., & Verhulst, S. (Eds.). (2009). Tinbergen’s legacy: de Waal, F. B. M., & Ferrari, P. F. (2010). Towards a bottom-up Function and mechanism in behavioral biology. Cambridge, perspective on animal and human cognition. Trends in England: Cambridge University Press. Cognitive Sciences, 14, 201–207. Bolhuis, J. J., & Wynne, C. D. L. (2009). Can evolution explain de Waal, F. B. M., & Tyack, P. L. (Eds.). (2003). Animal Social how minds work? Nature, 458, 832–833. Complexity. Cambridge, MA: Harvard University Press. Bowden, E. M., Jung-Beeman, M., Fleck, J., & Kounios, J. Dennett, D. C. (1983). Intentional systems in cognitive ethol- (2005). New approaches to demystifying insight. Trends in ogy: Th e “Panglossian paradigm” defended. Th e Behavioral Cognitive Sciences, 9, 322–328. and Brain Sciences, 6, 343–390. Brannon, E. M. (2006). Th e representation of numerical magni- Dukas, R., & Ratcliff e, J. M. (Eds.). (2009). Cognitive ecology II. tude. Current Opinion in Neurobiology, 16, 222–229. Chicago: Th e University of Chicago Press. Brodin, A. (2010). Th e history of scatter hoarding studies. Dusenbery, D. B. (1992). Sensory ecology. New York: W.H. Philosophical Transactions of the Royal Society B, 869–881. Freeman. Burkhardt, R. W. (2005). Patterns of behavior. Chicago: Eacott, M. J., & Norman, G. (2004). Integrated memory for University of Chicago Press. object, place, and context in rats: A possible model of episod- Call, J., & Carpenter, M. (2001). Do apes and children know ic-like memory? Journal of Neuroscience, 24, 1948–1953. what they have seen? Animal Cognition, 4, 207–220. Eichenbaum, H., & Fortin, N. J. (2005). Bridging the gap Call, J., & Tomasello, M. (2008). Does the chimpanzee have a between brain and behavior: Cognitive and neural mecha- theory of mind? 30 years later. Trends in Cognitive Sciences, nisms of episodic memory. Journal of the Experimental 12, 187–192. Analysis of Behavior, 84, 619–629.

shettleworth 15

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 1515 99/22/2011/22/2011 12:23:2212:23:22 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

Eichenbaum, H., Fortin, N. J., Ergorul, C., Wright, S. P., & Hampton, R. R., Zivin, A., & Murray, E. A. (2004). Rhesus Agster, K. L. (2005). Episodic recollection in animals: “If it monkeys (Macaca mulatta) discriminate between knowing walks like a duck and quacks like a duck...” Learning and and not knowing and collect information as needed before Motivation, 36, 190–207. acting. Animal Cognition, 7, 239–246. Emery, N. J., & Clayton, N. S. (2004). Th e mentality of crows: Hasher, L., & Zacks, R. T. (1984). Automatic processing of funda- Convergent evolution of intelligence in corvids and apes. mental information. American Psychologist, 39, 1372–1388. Science, 306, 1903–1907. Hauser, M. D., Chomsky, N., & Fitch, W. T. (2002). Th e faculty Endler, J. A., & Basolo, A. L. (1998). Sensory ecology, receiver of language: What is it, who has it, and how did it evolve? biases and sexual selection. Trends in Ecology and Evolution, Science, 298, 1569–1579. 13, 415–420. Heinrich, B. (1995). An experimental investigation of insight in Epstein, R. (1985). Animal cognition as the praxist views it. common ravens (Corvus corax). Th e Auk, 112, 994–1003. Neuroscience & Biobehavioral Reviews, 9, 623–630. Herrmann, E., Call, J., Hernández-Lloreda, M. V., Hare, B., Epstein, R., Kirshnit, C. E., Lanza, R. P., & Rubin, L. C. (1984). & Tomasello, M. (2007). Humans have evolved specialized ‘Insight’ in the pigeon: Antecedents and determinants of an skills of social cognition: Th e cultural intelligence hypothesis. intelligent performance. Nature, 308, 61–62. Science, 317, 1360–1366. Ferkin, M. H., Combs, A., delBarco-Trillo, J., Pierce, A. A., & Heyes, C. (2008). Beast Machines? Questions of animal con- Franklin, S. (2008). Meadow voles, Microtus pennsylvanicus, sciousness. In M. Davies & L. Weiskrantz (Eds.), Frontiers have the capacity to recall the “what”, “where”, and “when” of consciousness (pp. 259–274). Oxford, England: Oxford of a single past event. Animal Cognition, 11, 147–159. University Press. Fitzpatrick, S. (2008). Doing away with Morgan’s Canon. Mind & Heyes, C. M., & Galef, B. G., Jr. (Eds.). (1996). Social learning Language, 23, 222–246. in animals: Th e roots of culture. San Diego, CA: Academic Fortin, N. J., Wright, S. P., & Eichenbaum, H. (2004). Press. Recollection-like memory retrieval in rats is dependent on Hodos, W., & Campbell, C. B. G. (1969). Scala naturae: Why the hippocampus. Nature, 431, 188–191. there is no theory in comparative psychology. Psychological Frank, M. J., Rudy, J. W., Levy, W. B., & O’Reily, R. C. (2005). Review, 76, 337–350. When logic fails: Implicit transitive inference in humans. Hoppitt, W. J. E., Brown, G. R., Kendal, R., Rendell, L., Memory & Cognition, 33, 742–750. Th ornton, A., Webster, M. M., & Laland, K. N. . (2008). Galef, B. G. (1998). Edward Th orndike: Revolutionary psycho- Lessons from animal teaching. TRENDS in Ecology and logist, ambiguous biologist. American Psychologist, 53, Evolution, 23, 486–493. 1128–1134. Hulse, S. H., Fowler, H., & Honig, W. K. (Eds.). (1978). Cognitive Galef, B. G., & Heyes, C. M. (2004). Introduction: Special issue Processes in Animal Behavior. Hillsdale, NJ: Erlbaum. on social learning. Learning & Behavior, 32, 1–3. Jeff ery, K. J. (Ed.). (2003). Th e neurobiology of spatial behaviour. Gallistel, C. R. (1990). Th e organization of learning. Cambridge, Oxford, England: Oxford University Press. MA: MIT Press. Jozefowiez, J., Staddon, J. E. R., & Cerutti, D. T. (2009). Gibbon, J., Church, R. M., Fairhurst, S., & Kacelnik, A. (1988). Metacognition in animals: How do we know that they know? Scalar expectancy theory and choice between delayed Comparative Cognition & Behavior Reviews, 4, 29–39. rewards. Psychological Review, 95, 102–114. Kacelnik, A. (2006). Meanings of rationality. In S. Hurley & Girndt, A., Meier, T., & Call, J. (2008). Task constraints mask great M. Nudds (Eds.), Rational animals? (pp. 87–106). Oxford, apes’ ability to solve the trap-table task. Journal of Experimental England: Oxford University Press. Psychology: Animal Behavior Processes, 34, 54–62. Kacelnik, A., & Bateson, M. (1996). Risky theories–the eff ects Goddard, M. J. (2009). Th e impact of human intuition in psy- of variance on foraging decisions. American Zoologist, 36, chology. Review of General Psychology, 13, 167–174. 402-–434. Griffi n, D. R. (1976). Th e question of animal awareness. Kacelnik, A., Brunner, D., & Gibbon, J. (1990). Timing New York: Rockefeller University Press. mechanisms in optimal foraging: Some applications of sca- Griffi n, D. R. (1978). Prospects for a cognitive ethology. Th e lar expectancy theory. In R. N. Hughes (Ed.), Behavioural Behavioral and Brain Sciences, 4, 527–538. mechanisms of food selection: Vol. G20 (pp. 61–82). Berlin: Griffi n, D. R., & Speck, G. B. (2004). New evidence of animal Springer-Verlag. consciousness. Animal Cognition, 7, 5–18. Kacelnik, A., & Marsh, B. (2002). Cost can increase preference Hampton, R. R. (2001). Rhesus monkeys know when they in starlings. Animal Behaviour, 63, 245–250. remember. Proceedings of the National Academy of Sciences Kamil, A. C. (1998). On the proper defi nition of cognitive (USA), 98, 5359–5362. ethology. In R. P. Balda, I. M. Pepperberg, & A. C. Kamil Hampton, R. R. (2005). Can rhesus monkeys discriminate (Eds.), Animal cognition in nature (pp. 1–28). San Diego, between remembering and forgetting? In H. S. Terrace & CA: Academic Press. J. Metcalfe (Eds.), Th e missing link in cognition: Origins of Kamil, A. C., & Mauldin, J. E. (1988). A comparative-ecolog- self-refl ective consciousness (pp. 272–295). New York: Oxford ical approach to the study of learning. In R. C. Bolles & University Press. M. D. Beecher (Eds.), Evolution and learning (pp. 117–133). Hampton, R. R. (2009). Multiple demonstrations of metacog- Hillsdale, NJ: Lawrence Erlbaum. nition in nonhumans: Converging evidence or multiple Kennedy, J. S. (1992). Th e new anthropomorphism. Cambridge, mechanisms? Comparative Cognition and Behavior Reviews, England: Cambridge University Press. 4, 17–28. Kohler, S., Moscovitch, M., & Melo, B. (2001). Episodic Hampton, R. R., & Schwartz, B. L. (2004). Episodic memory in memory for object location versus episodic memory for nonhumans: What and where, is when? Current Opinion in object identity: Do they rely on distinct encoding processes? Neurobiology, 14, 192–197. Memory & Cognition, 29, 948–959.

16 darwin, tinbergen, and the evolution of comparative cognition

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 1616 99/22/2011/22/2011 12:23:2212:23:22 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

Kohler, W. (1925/1959). Th e mentality of apes (E. Winter, Trans.). paradigm. Philosophical Transactions of the Royal Society B: New York: Vintage Books. Biological Sciences, 365, 859–867. Koriat, A., Hilit, M., & Nussison, R. (2006). Th e intricate rela- Premack, D. (2007). Human and animal cognition: Continuity tionships between monitoring and control in metacognition: and discontinuity. Proceedings of the National Academy of Lessons for the cause-and-eff ect relation between subjective Sciences (USA), 104, 13861–13867. experience and behavior. Journal of Experimental Psychology: Real, L. A. (1993). Toward a cognitive ecology. Trends in Ecology General, 135, 36–69. and Evolution, 8, 413–417. Kornell, N. (2009). Metacognition in humans and animals. Reid, P. J. (2009). Adapting to the human world: Dogs’ Current Directions in Psychological Science, 18, 11–15. responsiveness to our social cues. Behavioural Processes, 80, Kounios, J., & Beeman, M. (2009). Th e Aha! moment. Current 325–333. Directions in Psychological Science, 18, 210–216. Ristau, C. A. (Ed). (1991a). Cognitive ethology: Th e minds of other Krebs, J. R., & Davies, N. B. (1981). An introduction to animals. Hillsdale, NJ: Lawrence Erlbaum. behavioural ecology (1st ed.). Oxford, England: Blackwell Ristau, C. A. (Ed). (1991a). Cognitive ethology: Th e minds of other Scientifi c. animals. Hillsdale, NJ: Lawrence Erlbaum. Laland, K. N., & Galef, B. G. (Eds.). (2009). Th e question of ani- Ristau, C. A. (1991b). Aspects of the cognitive ethology of an mal culture. Cambridge, MA: Harvard University Press. injury-feigning bird, the piping plover. In C. A. Ristau (Ed.), Margoliash, D., & Nusbaum, H. C. (2009). Language: Th e Cognitive ethology: Th e minds of other animals (pp. 91–126). perspective from organismal biology. Trends in Cognitive Hillsdale, NJ: Lawrence Erlbaum. Sciences, 13, 505–510. Roberts, W. A. (2002). Are animals stuck in time? Psychological Marler, P., & Slabbekoorn, H. (Eds.). (2004). Nature’s music. San Bulletin, 128, 473–489. Diego, CA: Elsevier Academic Press. Roberts, W. A., & Feeney, M. C. (2009). Th e comparative Mayer, R. E. (1995). Th e search for insight: Grappling with study of mental time travel. Trends in Cognitive Sciences, 13, Gestalt psychology’s unanswered questions. In R. J. Sternberg 271–277. & J. E. Davidson (Eds.), Th e nature of insight (pp. 3–32). Roberts, W. A., Feeney, M. C., MacPherson, K., Petter, M., Cambridge, MA: MIT Press. McMillan, N., & Musolino, E. (2008). Episodic-like mem- Mitchell, S. D. (2005). Anthropomorphism and cross-species ory in rats: Is it based on when or how long ago? Science, modeling. In L. Daston & G. Mitman (Eds.), Th inking with 320, 113–115. animals: New perspectives on anthropomorphism (pp. 100– Romanes, G. J. (1892). Animal intelligence. New York: D. 117). New York: Columbia University Press. Appleton. Morgan, C. L. (1894). An introduction to comparative psychology Rosati, A. G., Stevens, J. R., Hare, B., & Hauser, M. D. (2007). (1st ed.). London: Walter Scott. Th e evolutionary origins of human patience: Temporal pref- Mulcahy, N. J., & Call, J. (2006). Apes save tools for future use. erences in chimpanzees, bonobos, and human adults. Current Science, 312, 1038–1040. Biology, 17, 1663–1668. Nieder, A. (2005). Counting on neurons: Th e neurobiology Schiller, P. H. (1957). Innate motor action as a basis of learning of numerical competence. Nature Reviews Neuroscience, 6, (C. H. Schiller, Trans.). In C. H. Schiller (Ed.), Instinctive 177–190. behavior (pp. 264–287). New York: International Universities Osvath, M. (2009). Spontaneous planning for future stone- Press. throwing by a male chimpanzee. Current Biology, 19, Schuck-Paim, C., Pompilio, L., & Kacelnik, A. (2004). State- R190–R191. dependent decisions cause apparent violations of rationality Papini, M. R. (2008). Comparative Psychology (2nd ed.). in animal choice. PLoS Biology, 2, —e402. New York: Psychology Press. Seed, A. M., Call, J., Emery, N. J., & Clayton, N. S. (2009). Penn, D. C., Holyoak, K. J., & Povinelli, D. J. (2008). Darwin’s Chimpanzees solve the trap problem when the confound mistake: Explaining the discontinuity between human and non- of tool-use is removed. Journal of Experimental Psychology: human minds. Behavioral and Brain Sciences, 31, 109–178. Animal Behavior Processes, 35, 23–34. Penn, D. C., & Povinelli, D. J. (2007). On the lack of evidence Sherry, D. F. (2005). Do ideas about function help in the study that non-human animals possess anything remotely resem- of causation? Animal Biology, 55, 441–456. bling a ‘theory of mind’. Philosophical Transactions of the Sherry, D. F. (2006). Neuroecology. Annual Review of Psychology, Royal Society B, 362, 731–744. 57, 167–197. Penn, D. C., & Povinelli, D. J. (2009). On becoming approxi- Shettleworth, S. J. (1993). Where is the comparison in compara- mately rational: Th e relational reinterpretation hypothesis. tive cognition? Alternative research programs. Psychological In S. Watanabe, A. P. Blaisdell, L. Huber, & A. Young (Eds.), Science, 4, 179–184. Rational animals, irrational humans (pp.23–44). Tokyo: Keio Shettleworth, S. J. (2010). Cognition, evolution, and behavior University Press. (2nd ed.). New York: Oxford University Press. Povinelli, D. J. (1994). Comparative studies of animal mental Shettleworth, S. J., Krebs, J. R., Stephens, D. W., & Gibbon, state attribution: A reply to Heyes. Animal Behaviour, 48, J. (1988). Tracking a fl uctuating environment: A study of 239–241. sampling. Animal Behaviour, 36, 87–105. Povinelli, D. J. (2000). Folk physics for apes. New York: Oxford Shields, W. E., Smith, J. D., Guttmanova, K., & Washburn, D. A. University Press. (2005). Confi dence judgments by humans and rhesus mon- Povinelli, D. J., Nelson, K. E., & Boysen, S. T. (1990). Inferences keys. Journal of General Psychology, 132, 165–186. about guessing and knowing by chimpanzees (Pan troglo- Silva, F. J., Page, D. M., & Silva, K. M. (2005). Methodological- dytes). Journal of Comparative Psychology, 104, 203–210. conceptual problems in the study of chimpanzees’ folk Pravosudov, V. V., & Smulders, T. V. (2010). Integrating ecol- physics: How studies of adult humans can help. Learning & ogy, psychology and neurobiology within a food-hoarding Behavior, 33, 47–57.

shettleworth 17

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 1717 99/22/2011/22/2011 12:23:2312:23:23 AMAM OUP UNCORRECTED PROOF – FIRST-PROOF, 09/22/11, NEWGEN

Silva, F. J., Silva, K. M., Cover, K. M., Leslie, A. L., & Rubalcaba, Tolman, E. C. (1948). Cognitive maps in rats and men. M. A. (2008). Humans’ folk physics is sensitive to physi- Psychological Review, 55, 189–208. cal connection and contact between a tool and reward. Tomasello, M., Carpenter, M., Call, J., Behne, T., & Moll, H. Behavioural Processes, 77, 327–333. (2005). Understanding and sharing intentions: Th e origins Smith, J. D., Beran, M. J., Couchman, J. J., & Coutinho, M. V. of cultural cognition. Th e Behavioral and Brain Sciences, 28, C. (2008). Th e comparative study of metacognition: Sharper 675–735. paradigms, safer inferences. Psychonomic Bulletin & Review, Tulving, E. (1972). Episodic and semantic memory. In 15, 679–691. E. Tulving & W. Donaldson (Eds.), Organisation of memory Smith, J. D., Beran, M. J., Redford, J. S., & Washburn, D. A. (pp. 381–403). New York: Academic Press. (2006). Dissociating uncertainty responses and reinforce- Tulving, E. (2005). Episodic memory and autonoesis: Uniquely ment signals in the comparative study of uncertainty moni- human? In H. S. Terrace & J. Metcalfe (Eds.), Th e miss- toring. Journal of Experimental Psychology: General, 135, ing link in cognition: Origins of self-refl ective consciousness 282–297. (pp. 3–56). New York: Oxford University Press. Smith, J. D., Schull, J., Strote, J., McGee, K., Egnor, R., & Udell, M. A. R., Dorey, N. R., & Wynne, C. D. L. (2010). What Erb, L. (1995). Th e uncertain response in the bottle- did domestication do to dogs? A new account of dogs’ sensi- nosed dolphin (Tursiops truncatus). Journal of Experimental tivity to human actions. Biological Reviews, 85, 327–345. Psychology: General, 124, 391–408. Urcuioli, P. J. (2006). Responses and acquired equivalence classes. Smith, J. D., Shields, W. E., Allendoerfer, K. R., & Washburn, In E. A. Wasserman & T. R. Zentall (Eds.), Comparative D. A. (1998). Memory monitoring by animals and humans. cognition: Experimental explorations of animal intelligence Journal of Experimental Psychology: General, 127, 227–250. (pp. 405–421). New York: Oxford University Press. Smith, J. D., Shields, W. E., Schull, J., & Washburn, D. A. von Bayern, A. M. P., Heathcote, R. J. P., Rutz, C., & Kacelnik, A. (1997). Th e uncertain response in humans and animals. (2009). Th e role of experience in problem solving and inno- Cognition, 62, 75–97. vative tool use in crows. Current Biology, 19, 1965–1968. Sober, E. (2005). Comparative psychology meets evolution- Wang, R. F., & Spelke, E. S. (2002). Human spatial representa- ary biology. Morgan’s canon and cladistic parsimony. In tion: Insights from animals. Trends in Cognitive Sciences, 6, L. Daston & G. Mitman (Eds.), Th inking with animals: New 376–382. perspectives on anthropomorphism (pp. 85–99). New York: Warneken, F., & Tomasello, M. (2009). Varieties of altruism in Columbia University Press. children and chimpanzees. Trends in Cognitive Sciences, 13, Sole, L. M., Shettleworth, S. J., & Bennett, P. J. (2003). 397–402. Uncertainty in pigeons. Psychonomic Bulletin and Review, 10, Wasserman, E. A., & Young, M. E. (2010). Same-diff erent 738–745. discrimination: Th e keel and backbone of thought and rea- Stephens, D. W. (1987). On economically tracking a variable soning. Journal of Experimental Psychology: Animal Behavior environment. Th eoretical Population Biology, 32, 15–25. Processes, 36, 3–22. Sternberg, R. J., & Davidson, J. E. (Eds.). (1995). Th e nature of Wasserman, E. A., & Zentall, T. R. (2006). Comparative cog- insight. Cambridge, MA: MIT Press. nition: A natural science approach to the study of animal Suddendorf, T., & Busby, J. (2005). Making decisions with the intelligence. In E. A. Wasserman & T. R. Zentall (Eds.), future in mind: Developmental and comparative identifi cation Comparative cognition: Experimental explorations of animal of mental time travel. Learning and Motivation, 36, 110–125. intelligence (pp. 3–11). New York: Oxford University Press. Suddendorf, T., & Corballis, M. C. (2008). New evidence for Weisberg, R. W. (2006). Creativity. Hoboken, NJ: John Wiley animal foresight? Animal Behaviour, 25, e1–e3. and Sons. Suddendorf, T., & Corballis, M. C. (2010). Behavioural evidence Wimpenny, J. H., Weir, A. A. S., Clayton, L., Rutz, C., & for mental time travel in nonhuman animals. Behavioural Kacelnik, A. (2009). Cognitive processes associated with Brain Research, 215, 292–298. sequential tool use in New Caledonian crows. PLoS One, 4, Susskind, J. M., Lee, D. M., Cusi, A., Feiman, R., Grabski, W., & e6471. Anderson, A. K. (2008). Expressing fear enhances sensory Wright, A. A. (2006). Memory processing. In E. A. Wasserman & acquisition. Nature Neuroscience, 11, 843–850. T. R. Zentall (Eds.), Comparative cognition: Experimental Taylor, A. H., Elliff e, D., Hunt, G. R., & Gray, R. D. (2010). explorations of animal intelligence (pp. 164–185). New York: Complex cognition and behavioural innovation in New Oxford University Press. Caledonian crows. Proceedings of the Royal Society of London Yoerg, S. I., & Kamil, A. C. (1991). Integrating cognitive B. 277, 2637–2643. ethology with cognitive psychology. In C. A. Ristau (Ed.), Taylor, A. H., Hunt, G. R., Holzhaider, J. C., & Gray, R. D. Cognitive ethology: Th e minds of other animals (pp. 273–289). (2007). Spontaneous metatool use by New Caledonian Hillsdale, NJ: Lawrence Erlbaum Associates. crows. Current Biology, 17, 1504–1507. Zeigler, H. P., & Marler, P. (Eds.). (2004). Behavioral neurobiol- Th orndike, E. L. (1911/1970). Animal intelligence. Darien, CT: ogy of birdsong: Vol. 1016, Annals of the New York Academy of Hafner Publishing. Sciences. New York: Wiley-Blackwell. Th ornton, A., & McAuliff e, K. (2006). Teaching in wild meer- Zentall, T. R., Clement, T. S., Bhatt, R. S., & Allen, J. (2001). kats. Science, 313, 227–229. Episodic-like memory in pigeons. Psychonomic Bulletin and Th orpe, W. H. (1956). Learning and instinct in animals. London: Review, 8, 685–690. Methuen. Zhou, W. Y., & Crystal, J. D. (2009). Evidence for remember- Tinbergen, N. (1951). Th e study of instinct. London & New York: ing when events occurred in a rodent model of episodic Oxford University Press. memory. Proceedings of the National Academy of Sciences, 106, Tinbergen, N. (1963). On aims and methods of ethology. 9525–9529. Zeitschrift fur Tierpsychologie, 20, 410–433.

18 darwin, tinbergen, and the evolution of comparative cognition

228_Vonk_Ch28.indd8_Vonk_Ch28.indd 1818 99/22/2011/22/2011 12:23:2312:23:23 AMAM