The Cognitive Ethology of an "Injury-Feigning" Plover: a Beginning

Etolog{a, 3:57-68 (1993)

The cognitive ethology of an "injury-feigning" plover: a beginning

C.A. Ristau

Barnard College of Columbia University, Dept. of Psychology, 3009Broadway, New York City, N.Y., USA 10027 and Hunter College of the City University of New York, Dept. of Psychology, 695 Park Avenue, New York City, NY, USA 10021

ABSTRACT. The cognitive ethology ofan "injury-feigning" plover: a beginning.- Interest in the cognitive capacities of animals has led me to study antipredator behavior of birds, specifically plovers, through observations andexperiments in natural environments. I asked: 1) Do injury-feigning parent plovers want to lead an intruder/ predatoraway from the nest? What data could provide pertinent evidence? 2) Are plovers responsive to an intruder's attention to their nest? 3) Can plovers learnto discriminate between a potentially "dangerous" intruderand a "safe" one? Results support positive answers to these questions. To progress in the difficult endeavor of experimentally investigating purposeful and other cognitive behavior of animals, we need the efforts of many related disciplines including philosophy, biology and psychology. KEY WORDS. Purposeful behavior, Intentionality, Injury-feigning, Antipredator behavior, Plovers, Cognitive ethology

Introduction and some philosophical based fieldstudies. problems Cognitive ethology differs from most previous studies of animal cognition, interpreted broadly to include animal learning and discrimination studies, Donald Griffin (1976, 1992) has been in its emphasis on possible animal mental states particularly influential in rekindling interest in the and interest in matters such as deception and possibility that animals may have mental communication. experiences, including awareness, purposes, and The major philosophical problem in cognitive consciousness and that such experiences are ethology is, as Bennett (seminar, 1985) has phrased amenable to scientific investigation. "Cognitive it, drawing conclusions about the minds of animals ethology" (Griffin, 1976) is a beginning exploration from premises about behavior in the circumstances of the mental experiencesof animals, particularly as in which the animals are behaving. We, and any they behave in their natural environment in the study of mental processes, need a conceptual theory course of theirnormal lives. Many, but not all, that relates mentalistic terms to patterns of studies in cognitive ethology have, to date, behavior. And then there is the problem of emphasized the importance of observations of awareness. It is a hotly disputed matter whether naturally occurring behavior and of experimentally animals, to whom one might apply concepts of

55 Ristau

intentions, of belief and desire are aware of those should the path to the goal be obstructed. Griffin beliefs, or say some, whether they even experience (1985, p.37) also suggests that one criterion of beliefs or desires, as opposed to the terms being a conscious awareness in animals is "versatile useful, functionalist stance for an experimenter to adaptability of behavior to changing circumstances assume. and challenges." This facet of behavior is also Though solid theories do not exist, useful stressed by Boden (1983) from the vantage point of approaches do. One is the intentional stance, ArtificialIntelligence. proferredby philosophers such as Bennett (1976, I will draw upon these ideas in proposing an Dennett (1978, 1983, 1987), and Searle (1980) ruxl interpretation of distraction behavior exhibited by discussed by Beer (1983, 1991, 1992). "Intentional" various species of ground nesting birds when is a philosophical term meaning "aboutness," intruders approach the eggs or young as purposive reference.It does not mean "on purpose," though (Ristau, 1983a). I particularly chose to study birds, "wants it to be the case that" is an intentional because we are less likely to empathize with their phrase. Some other intentional terms are "thinks mental states than we do with our pets or fellow that," "believes that," "wants it to be the case that." primates. This objectivity may help us to carefully Conventional scientific explanationis phrased at the specify the evidence forsuch an interpretation and to zero order of intentional analysis without recourse to suggest possible levels in the transition from any mental state. Among possible first order rudimentary to more full fledged knowledge ruxl intentional analyses is a description in terms of purposes( or beliefs and desires as termed by many purpose, e.g. "An organism wants it to be the case philosophers). I have chosen to study a behavior that x", where x could be a response such as which, like many behaviors, human and otherwise, "organism B to follow." A second order analysis is a mixture of some fixed, genetically transmitted might be phrased, "Organism A wants Organism B elements and more flexible behaviors. to believe that z"; ie organism A is concernedabout the mental state of B, as opposed merely to B's behavior. Given the difficultiesin findingpersuasive The "injury-feigning" plover evidence forsecond order intentional statements, my research focusseson first orderintentional analysis, specificallyon an organism's purposes. I do not, A. The plovers' behaviors toward however, mean to imply that a non-human animal's intruders purposive behavior is like the very flexible, fully cognitive, full conscious, purposeful beahavior we humans sometimes have. Indeed, the philosopher I shall concentrate on the Piping Plover, Bennett ( 1976) and othershave noted the importance Charadriusmelodus and shall include data from of a transition among various species from Wilson's plover, C. Wilsonia, although I am also rudimentary "registrations" and "goals" to full doing comparative studies of the Semipalmated fledged "beliefs" and "desires" characteristics of at Plover (C. semipalmatus), the Lesser Golden Plover least some human activity. (also known as the American Golden Plover) What might be, if not criteria, then descriptive (Pluvialis dominica) and the Black-Bellied Plover propertiesof purposefulbehavior?The psychologist (Pluvialis apricaria). In all these species, both Tolman (1932) has stressed persistence to the goal, parents incubate the eggs forabout 3 1/2 - 4 weeks. especially that requiring variations in behavior, At this point precocial young hatch which can run

56 Etolog(a, Vol. 3, 1993

freelyand feedthemselves on their first day. The squawks as well. To a human the sense is strong young flyin about another 3 weeks. that the bird is injured, and one finds oneself The nest, eggs, young and adult are all very well trudging even hundredsof meters after the bird, only camouflaged. The nest, like that of many birds to see it suddenly fly away with agility. By that which perform distraction displays, is simply a point one is far fromthe nest or young. scrape on the ground. Since the nest is easily Note that the plover does not always make a accessible to predators, protection of the eggs broken wing display when its offspring are depends on camouflage, preventing potential approached by a ground moving object. For predators' knowledge of the nest's location, and instance, during one season's experiments, parents keeping them out of the nest's vicinity. gave broken wing displays during approximately For a plover to be conspicuous requires special 40% of the close approaches to the nest. behaviors or vocalizations. During incubation and Sometimes, instead, the plover leaves the nest before the young can fly, both parents of both cryptically with a silent, low run. It may hide in species perform distraction displays to intruders hollows, tail towards the intruder, thus being very which move along the ground. (See review of difficultto detect. various species' behavior in Gochfeld,198 4.) Yet another kind of variability occurs in a related The Piping and Wilson's Plover are two species, the KilldeerC. vociferous. Only rarely does shorebirds which typically nest on beaches or sand the killdeerperform broken wing displays to cattle, dunes of Eastern United States. In these two species, which do not eat the eggs, but may accidentally there are several different kinds of distraction trample the eggs. Sitting tight on its nest, the behaviors. The bird, especially a Piping Plover, killdeer is reported to lunge at the last moment in may peep loudly while walking and keeping apace the face of cow, causing it to veer(Armstrong,1947; or ahead of the intruder. Or the bird may fly Walker, 1955; Graul, 1975). A somewhat similar conspicuously and slowly in a large circle, exposing set of reactions to mammals occurs among southern its underside and bright wing stripes. As it flies or lapwings in Africa(Walters, 1980). In short, at least as it walks at a distance fromits young, it can be some species which perform broken wing displays heard to vocalize a "peep." This sound is often what exhibit flexibility in their use of the behavior. first attracts the human's attention to the cryptically But precisely what is it that the bird is doing? Is colored bird against the sandy beach. Sometimes the this a Fixed Action Pattern, or possibly a plover may engage in false brooding, sitting down disorganized "hysterical" behavior as some have with feathers slightly fluffed and wriggling as termed it? (Skutch, 1976, p. 403). Does the bird though it were on a nest - but doing so in the wrong have to do it? Can it control initiation or stopping location, where there are no eggs. Or it may merely of the BWD? Can the behavior be construed as pace back and forth in the general vicinity of a intentional? Whatis evidence for the existence of an human, seeming to eye the presumed predator as it intention? does so. On some approaches of an intruder, the bird may B. What are some possi ble hypotheses do a gradation of broken-wing displays (BWD), about the plover's behavior? perhaps beginning with a fanning tail and gradually increasing the awkwardnessof walk until it has one, then both wings widely arched, fluttering and Note that the following hypotheses are not dragging. It may then vocalize loud, raucous mutually exclusive; it is quite possible that some

57 Ristau

combination may finally prove to be the most thought to be "convulsive", "deliriously excited," satisfactory. and "its behavior patterns were more or less disorganized" (Skutch, 1976, p. 403). If the bird's 1. Fixed action pattern (F AP): behavior were indeed so disorganized, one would predict random directions of display or at least not The bird's behavior is an FAP which occurs consistent "leading away fromthe nest or young." when the parent bird is in a certain hormonal condition and in the presence of an intruderand the 3. Approach/withdrawal tendencies: plover's nest or young. A Fixed Action Pattern, a concept developed by the ethologists Lorenz aro This point of view, espoused by Tinbergen Tin bergen (1951), is describedin a recent textbook (1952), students of Schneirla (1959), and others is as follows: similar to the conflict hypothesis, but emphasizes "The distinguishing characteristics of the more orderly behaviors by the bird than those behavior arethe innate and stereotypedcoordination predicted by a simple conflict hypothesis. It is and patterningof several muscle movements which, predictedthat the bird would make a BWD at the when released, proceed to completion without point of conflict of approach motivations requiring further sensory input. In terms of its (aggression and brooding)and withdrawal tendencies almost total independence of feedback, the fixed (escape). None of the possible predicted behaviors action pattern represents an extreme class of suffices to account for the complexity of the prewired behavioral performanceswhich have come observed behaviors. Test of this model, in which a to be known as 'motor programs."' (Gould, 1982, p. human directly approached incubating Golden 37) Plovers, did not support the hypothesis (Bjerkjedal, If the behavior is an FAP, there are several 1991). possible predictionsabout the direction in which the plover makes a broken-wing display. a) The BWD should be made in random directions; thus the bird 4. Pre-programmmed sequence of should be just as likely to display towardthe nest or behavior: young as away. b) The displaying bird merely goes away from the nest/young, c) merely goes away By this hypothesis, thebird behaves accordingto fromthe intruder, d) moves away fromboth the nest a programmed sequence of behavior, in which or young and the intruder. This hypothesis requires stimuli such as direction of movement of the that the plover must know the location aro intruder, size of intruder, nearness to nest,and so movements or trajectories of the young and the forth determine the response of the parent bird. At intruders in order to respond appropriately. That is least for the piping and Wilson's plovers, the no small feat. (And difficult to conceive of as variability observed in their behavior does not lend simply an FAP.) itself to an interpretation of a rigidly programmed sequence of behavior. If we allow for great flexibility in that programming, we are including 2. Conflict behavior: the possibility of learning (Hypothesis #5), and if we allow reprogramming, we might well be talking Earlier investigators often interpreted thebroken wing display to be the result of conflicting about purposefulbehaviors. Recognize, however, no program yet exists that adequately accounts for the motivations. The displaying bird's behavior was behavior of a whole animal in the real world, so the

58 Etolog{a, Vol. 3, 1993

kind of "super" program that could include problem for cognitive ethology. Determining descriptions of intentional behavior is not plausibly methods to investigate an animal's purpose is part included as part of the hypothesis of "pre of that formidable task. Based on the previous programmed behavior. discussion concerning descriptive characteristics of intentional behavior, I propose the following 5. Learning: observable behaviors as providing suggestive evidence in support of the hypothesis. I can make Plovers might be able to learn about various no claim that these are necessary and sufficient aspects of the situation, including which intruders conditions for intentional behavior. I have not arepotential predators and should be dealt with more succeededwhe recenturies of philosophical thought warily than others which are not. We have have failed. investigated that aspect in work I shall discuss The outline of my argument, an attempt at a (section IV; Ristau, 1991). The possibility of fairly general one forinferring purposeful behavior learning is not precluded by any other hypothesis. in an interactive situation is as follows: 1. The behavior exhibited should accomplish the 6. Intentional or purposeful behavior: presumed goal, though not neccessarily in every case. Thus the direction in which a parent bird The plover wants to leadthe intruderaway from moves while making a BWD should usually be the nest or young. It behaves so as to achieve this adequate to lead the intruder away from the objective, including a broken-wing display. This nest/young. There is, however, no requirement that hypothesis requires that the plover know the the behavior be optimal. Furthermore completely location and movements of young and accurate performance might well be suspect. In the intruders. I do not mean to imply that every interaction, the communicator should act so as to plover has independently thought of or learnedto maximize attention to it. Thus the bird should make a BWD, since the BWD is undoubtedly an position itself where it can readily be observed. evolved, genetically transmittedbehavior. However, 2. The organism should exhibit monitoring of strategies progress towards the presumed goal. Thus the for its effective use may well be learned, both displaying plover should ascertain the intruder's directly and by observation. The fact that a behavior attention, location and behavior. It may, however, or some aspect of it is learned or genetically pre be very difficult to obtain good evidence that an wired does not preclude the possibility of conscious organism is doing this. thinking associated with it (see Griffin, 1984, 3. If progress towards the presumed goal is 1985). In the next section, this hypothesis is disrupted, the organism should be able to modify its discussed more fully. behavior so as to again proceedtowards thegoal. 4. Flexibility achieved in one domain suggests C. Evidence needed to evaluate hypotheses that the organism may have such capacities in other about purposeful behavior, specifically domains as well. re: "the plover wants to lead the intruder away from nesUyoung" D. Specific methods

As already noted, drawing conclusions about mind from behavior is the major philosophical To gather the kinds of eveidencejust noted, my

59 Ristau

assistants and I conducted experiments on a barrier There areactually several possible definitionsof island off the coast of Virginia, southeast USA, "appropriate direction", but whatever the defintion, interacting with Piping and Wilson's Plovers which the data support the hypothesis that the plover is had nests or young. We initally attempted to use a attempting to get an intruderfurther away from the stuffed natural predator, a raccoon, mounted on a nest/young. For example, we can ask if the intruder radio-controlledminiature dune buggy. Encountering went to the location of the displaying bird, would it too many logistical problems, we switched to a be further fromthe nest at the end of the period of human intruderwhose actions were directedby me dsiplaying than at the beginning. The answer is yes, over a walkie talkie. Essentially, the humans' in 98% of the cases. Using the most stringent behavior was designed to be unpredictable to the definition, we ask, would an intruder who may birds, to that extent simulating the behaviorof a eventually get farfrom the nest, nevertheless, in the natural predator. Thus directions of the intruder's course of following the displaying adult, pass closer initial approach and changes in movements were to the offspring?This could happen if the displaying varied. The human intruderapproached the nest or adult and the intrtuder were on appoiste sides of the young, walked in the area of offspring, stopping at nest. Even by that stringent criterion, in 87% of the the nest and at other locations and either followedor instances, a path that followed the parent plvoer did not followthe displaying adult. would never bring the intruder closer to the nest or We used audio and videotape to record our young. In short, yes, the birds' direction of display observations of the birds' and human behavior, is adequate to get an intruder further fromoffspring. including locations and directionsof movement (i.e. Where in the intruder'svisual field does the bird compass points such as north northeast). There were make broken-wing displays?If the bird is displaying two sets of simultaneous observations, that of the in order to attract the intruder'sattention, one would intruderand the observer, synchronized by reporting expect the bird to be selective about where it time to the nearest second. Occasionally both displays; it should display where the intruderwill observers were theintruders. see it. In fact, 44 of the 45 BWDs were made to the front of the intruder rather than behind, that is, within a 180 degreearc of the intruder's visual field. The one possible exception occurred when an intruderwas searchingfor young, was actually very E. Results (further details in Ristau, near them, and moving in a zig-zag fashion, facing 1991) one way, then another. The parent made a BWD to the side of the intruder, directed away from the The reported data are drawn only from young, and headedopposite to the general trend of interactions in which the locations and directions of the intruder'smovement toward the chicks. intruder,displaying birds and nest or all chicks could Note that these data do not determine which be determined. Data have been combined for stages intruder characteristics the bird was responding to, with eggs and unfledgedchicks. In 45 instances of because the intruder was moving in most cases of broken wing displays, the data were sufficiently BWDs. Thus, direction of movement as well as detailed foranalysis. orientation of the intruder'seyes, head or body could 1. Evidence that the plovers make broken wing be cues. displays in a direction "appropriate"to lead intruders Positioning by the bird beforemaking a BWD: away from thenest or young Another question examined in detail was the

60 Etolog(a, Vol. 3, 1993

location of the bird when it began its broken-wing intruder did not follow, the bird engaged in any of a displays. If this behavior is a reflex that is elicited variety of behaviors. In over half of the occurrences, whenever an intruderapproaches closely enough, the bird stopped what it was doing and reapproached one might expect the display to occur wherever the the intruder by either flying or walking closer. In bird is located. However, the bird always moves nine instances (29% ), the bird either continued to before displaying. Sometimes the bird moves by make a BWD, or increased the intensity of the flying which is an easily and accurately observable display, for example, by flapping its wings more form of locomotion. One can argue that by flying to vigorously or vocalizing raucously while get to a location, rather than walking, a slower form displaying. Of the remaining five cases, after of locomotion, it is probably important to the displaying, the bird flew to the location of the plover to get to that location rapidly. In all the cases young (three cases), flew away (one case), and in of flying, namely 13, the bird's new position was one other case, did not reapproach or fly. closer to the intruder than was its position before flight. One would not expect such positioning if, as F. Summary some have suggested, the bird were attempting to get away from the intruder. Furthermore, in 11 of those 13 cases, not only The use of intense distraction displays, at least was the bird closer to the intruder, but it was closer by the plovers in this study, indicates they usually to the front of the intruder than it had been, i.e. perform the displays in a direction that would cause more directly in the center of the intruder's visual an intruder following them to get further away from field and/or the path of the moving intruder. the threatenednest or young. Furthermore, the birds 2. Evidence that birds making a broken-wing monitor the intruder's approach and modify their display monitor the intruder's behavior: To engage behavior in response to changes in intruder in these various behaviors strongly suggests the locomotion. I have interpretedthe data as providing birds aremonitoring the intruders. But how can one at least suggestive evidence for the purposive nature determine what a plover is monitoring? Plovers (or first order intentional analysis) of the birds' have laterally placed eyes, so they see over a wide behavior. As noted earlier, I do not claim that it is field. It would be difficult to specify exactly what comparable to the fully cognitive, fully conscious they areattending to within that field.They cannot, plans we humans sometimes make. however, see behind them. Observations, To those that will have none of attempts to photographs and videotapesshow that as a plover is study "consciousness" in animals, recognize that making a BWD, it often turnsits headsharply back even taking the stance of purposeful, or even over its shoulder, its eye toward the intruder. This intentional behavior, implying nothing of orientation strongly suggests monitoring of the consciousness, is a fruitfulenterprise. The stance led intruder. me to design experiments which I had not otherwise 3. Modification of displays in response to thought to do, which no one else had done, and changing intruder behavior: Suppose an intruder which revealed complexities in the behavior of the does not follow a displaying bird? What does the Piping Plover predator distraction behavior not parent do then? Such occurrences were part of the heretofore appreciated.I invite readers to adopt the experimental protocol. In short, if the intruder did stance of intentional behavior and to help delineate follow the displaying bird, the bird did not the levels and kinds of knowledge and purposiveness reapproach (come closer to) the intruder. If the an organism might have.

61 Ristau

The gaze experiments constraints. There is informationfrom the laboratory and the In other experiments I asked whether piping field that suggests a significant role for learning. plovers might be responsive to the attention of an Such work includes birds , namely stilts, intruder. Specifically, did the plover differentiate (Himanoptus himanoptus) mobbing egg-eating between an intruder who looked towards its nest Laughing gulls (Larus atricilla) in one geographic region, scanning the dunes where its nest was region, while stilts in another region do not mob located vs. one who looked in the opposite the largely insectivorous, but very similar looking, direction, scanning towardsthe ocean. As he or she Franklin's gulls (L. pipixcan).(Dinsmoor, 1977) In was scanning, each intruder walked past the nest the lab, by observing blackbirds apparently along the same paths at a considerabledistance from mobbing an innocuoushoney-eater, other blackbirds the nest (15-25) meters. The birds did discriminate learnedto do the same (Curio et al, 1978). In other between the two intruders, remaining off the nest lab experiments, pigeons (Columba livia) learned longer when an intruder looking twards it nest many distinctions between betweenvarious classes walked by than when one looking in the opposite of objects (both natural and man-made) and living direction passed. organisms which were displayed on photographic Other work has also indicated that animals can slides (review in Watanabe, in press), including a differentiate between direction of gaze, hognosed particular person in a variety of poses and attire snakes (Bumardt 1991), chickens (Gallup, 1972) aoo fromother people (Herrnsteinet al., 1976) anoles (Hennig, 1977) engaging in "death feigning" We conductedexperiments to determineif piping for longer periods of time when confrontedwith a plovers could learn to distinguish between two nearby human intruderstaring at them as compared different intruders, one of whom had acted "safely" to an intruderaverting his eyes. (review by Arduino towardsthe eggs and the other more "dangerously." and Gould, 1984 ). Important differencesbetween my The essence of the experiment: Two distinctively work and these other studies is both the greater dressed humans each walked "safely" by a nest distance of the intruder from the organism and, containing eggs, at a considerabledistance from the instead of staring directly at the bird, the intruder in nest (12-32 meters away), not paying attention to my studies is scanning the bird's nest region, which, the nest or to the parent bird. ThesePre-Tests were depending on the bird's movement, may or may not conductedto verifythat the plovers did not initially include the bird. reactdifferen tially to the two intruders.One intruder then closely approached the nest approximately twice, acting in a way towards the eggs that we Can birds discriminate between hoped would appear to the parent bird to be "safe" and "dangerous" intruders? dangerous or threatening; ie pausing and hovering over thenest (these arethe "dangerous" approaches). In another series of experiments, I asked how Then in the Post-Test trials, each intruder once plovers might come to know which intruders are again walked"safely" by the nest several times, at a potentially dangerous predators and which are more considerable distance from the nest and not paying likely to be benign? They could be innately attention to the nest or parent. Would the parent equipped to respond to sign stimuli of predators or birds now react differently to the previously could be learningto discriminate potential predators "dangerous"intruder as comparedto the intruderwho from benign intruders, possibly guided by innate had acted only "safely?"If the plovers can learn to

62 Etolog(a, Vol. 3, 1993

discriminate between human intruders, it is most temporary lack of enough different persons to probable that they normally are able to learn an even function as intruders. Under these unusual simpler discrimination, that between different circumstances, one bird made a reverse species of non-human intruders which have behaved discrimination, although its mate was correct. in differentially threatening ways towards the plovers' nest or young. In brief, the plovers did make that discrimination Discussion between the two humans, exhibiting higher levels of arousal to an intruder who had previously walked The results indicate that Piping Plovers can very close to the nest and acted dangerously as compared rapidly learn to discriminate between a previously to their reaction to the "safe"intruder. Remember, "dangerous" human intruder and a "safe" one, the differentreactions of the plovers aretested during typically after only two close ("dangerous") Post-Test trials when each intruder is walking approaches to the nest. Yet one must consider "safely" by the nest, looking straight aheadand not whether it is appropriate to label intruders as safe" paying attention to the nest or the parent. Some of and "dangerous".lt is not clear whether a plover can the differencesin behavior are the following: in form such concepts. However, in their natural world slightly over half the trials when the bird made a the plovers must react to some set of animals (and correct discrimination, the bird got off its nest possibly humans) as potential predators and others during the tangential walk of the "dangerous" as non threatening. intruder" while remaining on the nest for the What mechanism might underlie the plovers' tangential walk of the "safeintruder". Some of the discrimination? Classical, that is, Pavlovian other discriminating behaviors included the distance conditioning might be occuring, such that the of the intruderfrom the nest when the bird opted to plover becomes more highly aroused when an leave (leaving when the "dangerous"intruder was far intruder is close to the nest. That "dangerous" away but only when the "safe"intruder was close), intruder, even at a distance, becomes a conditioned the distance and duration the bird got offthe nest, stimulus for arousal. Simultaneously, the plover and its posture, i.e. craning its neck up high in an could be habituating to the "safe" intruderwho does alert neck posture vs. not. not come close to the nest on repeated walks past Of the 13 birds with sufficient data to use in the the nest. analysis, 11 birds made correct discriminations, ie For future work, it would be interesting to during Post-Test trials, they were more aroused by consider which aspects of the intruders' behavior are the previously "dangerous"intruder than they were most critical to the plover's reacting in a more by the "safe" intruder. For one bird there was no aroused way to the "dangerous" intruder. That particulardifference in response and one bird made a intruder not only approached the nest area much reversediscrim ination. more closely than did the "safe" intruder, it could The single incident of a reverse discrimination is also be consideredas "knowing" the location of the particularly interesting. In that experiment, the nest and possibly of having an "intention" to cause intruder's roles had been reversed such that the harm, while the "safe"intruder did not. Alternatively person now in the role of the "safe" intruder boo one could argue that the "dangerous" intruderwas been the "dangerous" intruder in two previous actually not very dangerous at all; although right at aborted attempts to conduct the experiment and vice the nest, it did not smash or remove any eggs. Other versa. This atypical procedurewas necessitated by a more straightforward manipulations to examine

63 Ristau

perceived "dangerousness" of an intruder, include of events can be more persuasive. varying its size, speed of approach, and direction of 3. The"quadrangle" approach, this last being of greater theoretical This leads me to a third point, how one significance. determines an organism's presumed goal, or in philosophical terms, its "belief/desire." To place constraints on the number of possible belief/desire General discussion pairs that could be held by an organism, Bennett (1991) suggests determining a quadrangle, the "sensory-input/belief/desire/behavior quadrangle." Finally I would like to make some general Rather than a single event, one finds a class of statements about possibly useful methods to explore events which constitute the range of sensory inputs. purposeful behavior or intentionality in general. To make the strongest case that explanation in I. Flexibility of behavior: An important criteria terms of mental states, specificallybeliefs/desires, is in evaluatingpurposeful behavior is the ability of required (rather than a stimulus trigger for the an organism to modify its behavior when progress observed behavior), one tries to determine a wide towards the presumed goal is obstructed in some range of events/conditions quite differentfrom one way. Note that some of the organism's another, which are made sense of, characterizedby, modifications may be innate or hard-wired the animal's beliefs. The class of events could adaptations, while others may be flexible, even consist of typical occurrences as well as infrequent novel. Comparative studies of other species might events, the latter being either naturally occurring or help elucidate evolutionary origins of apparently staged in an experiment. The animal's associated "novel" behaviors. We might also detect species in belief might be, for instance, that doing any of a different stages of evolution of anti-predator variety of behaviors in these contexts will lead to behavior, or subject to differentevolutionary or the same goal, such as access to food, or getting an learning history with differentecological conditions intruderaway fromthe nest/young. or predation pressures. Predators' intelligence or But these suggestions are only a beginning in specificsensory/ motoric capacities could be such a the formidable task of delineating the kinds of pressure.Likewise comparativework can help reveal knowledge and purposesan organism might have. I the limits of prey species capacities and flexibility. hope many scientists and philosophers will join 2. The use of one or few time occurrencesof togetherin this fruitful,but difficultenterprise. behavior: Extensive observations of naturally occurring events, in this case of predator-prey interactions, can permit one to observe rare behavioral sequences and strategies. This is a most Acknowledgements important point. If we consider the possibility that the species we are studying are intelligent, a markof that intelligence will be a novel response to an For more complete description of the work and unusual or unsuccessful interaction. Thus part of ideas presented here, please refer to Cognitive our data set with intelligent species is likely to Ethology: The Minds of Other Animals, 1991, include one or few time occurrences.They should edited by myself and published by Lawrence not be ignored. They can suggest fruitful Erlbaum Associates. experimentation. Furthermore, though any one I would like to thank the following fortheir help event by itself is likely to be uninterpretable,a class conducting the researchreported here: For the first

64 Etologfa, Vol. 3, 1993

season: The Harry Frank Guggenheim Foundation of plovers is a development fromthat point of view. for financial support; The Virginia Coast Reserve of In my research, I asked the following questions The Nature Conservancy for permission to use their about the birds: land; The Marine Science Consortium of Wallops l. Do parent plovers (Charadriusmelodus and C. Island, Virginia forliving and office accomodations; wilsonia) want to lead an intruder/predator away David Thompson, for initial preparations; Laura from the nest/young? What kinds of data could Payne, for observations. For subsequent field provide evidence for a purposive interpretation of seasons, I thank the Mashomack Preserve of the behavior? Some of the data gathered from Nature Conservancy on Shelter Island, New York interactions of the plovers with human intruders and its Director, Michael Laspia forliving and office were: Is the direction of a broken-wing display accomodations and supportive services, and the adequate to cause an intruderfollowing them to get Morton Wildlife Refuge and the town of further away from the offspring? Where in the Southampton, Long Island, N.Y. for permission to intruder's visual fielddoes the parent plover usually use their land for research. I am grateful to the make a broken-wing display (BWD)? (Answer: Whitehall Foundation for financial support of the frontal plane) When the plover flies to a different research and the Edna Bailey Sussman Fund and position before displaying, where does it go? BarnardCollege for support of research assistants. I (Answer: closer to the intruder and typically closer am also indebted toMargaret McVey who was my to the center of the intruder'svisual field/directionof colleague during a season and to her and Elsia locomotion) Does the displaying bird monitor the Hellen, William Langbauer, and Terry Metwijn for intruder's behavior?Does the displaying bird modify help in fieldwork and preliminary data analysis for its behavior in response to changing intruder that season. For subsequent seasons, I thank my interactions? Specifically, if the intruder does not research assistants Crista Diels, Angela Greer, following the injury-feigning bird, the bird engaged Samuel Hellings, Angelica Landrigans, Margaret in any of a variety of behaviors. In over half the Mosteller, Kim Pietrzak, Karen Seidler, Peter occurrences,the bird stopped what it was doing and Sherman, Salila Shivde, and Michelle Sullivan. I reapproachedthe intruderby either flying or walking thank Daniel C. Dennett forphilosophical insights, closer. In less than a third of the instances the bird Stevan Hamad for his helpful discussions, and in either continued to make a BWD or increased the particular, Donald R. Griffin for constructive intensity of the display. comments on a longer ms. fromwhich this paperis 2. Are plovers responsive to an intruder's drawn, and for his continued encouragements. attention to their nest, interpretedas direction of a human intruder'seye gaze? 3. Can plovers learn to discriminate between a potentially "dangerous" intruder and a "safe" one, Summary defined respectively as a human which has previously approached closely to the nest and another which has not. Interest in the cognitive capacitiesof animals has Results support positive answers to the been rekindled,perhaps especially since Donald R. questionsposed. Griffin's book in 1974, The Question of Animal These findings do not support a number of Awareness. My fieldobservations and experiments alternative hypotheses, for example that injury about injury-feigningand other antipredatorbehavior feigning is proximately governed by conflicts

65 Ristau

between escape, aggressive and brooding drives, piano frontal) Cuando vuelan a una pos1c1on hypotheses termed "conflict" and "approach/ diferenteantes de hacer el comportamiento, l,donde withdrawal." They do begin to support a purposive van? (respuesta: hacia el intruso y normalmente interpretationof the use of displays by these species hacia el centro del campo visual o de la direcci6n � of plovers. desplazamiento de este) l Vigil a el actor el To identify purposive behavior experimentally comportamiento del intruso? ;,Modifica su and to investigate other possible mental states of comportamiento en respuesta a los cambios de las animals is a most difficult task. To progress in this interacciones del intruso? Especificamente, si el endeavor, we need and invite the efforts of many intruso no sigue al ave que tinge el dafio realiza related disciplines including philosophy, biology, varios tipos de comportamiento. Aproximadamente comparative, experimental, developmental andsocial la mitad de las veces se para y se aproxima al psychology and, of course, ethology. intruso para o bien volar o caminar cerca. En menos de un tercio de las veces continua haciendo el comportamiento BWD o bien incrementa la intensidad del mismo. Resumen 2. l,Responden los Charadrius a la atenci6n del intruso a su nido interpretadocomo mirar fijamente un intrusohumano? La etolog(a cognitiva de "simulacion delesiones" d! 3. ;,Pueden los Charadrius aprendera discrirniar un Charadridae: un comienzo. entre un intruso potencialmente "peligroso" y uno El interes en las capacidadescognitivas de los "seguro", definidos respectivamente como un animales ha vuelto a suscitarse, quizas humano que previamente se habfa aproximado especialmente desdela aparici6n del libro de Donald mucho al nido y otroque no? R. Griffin en 1974, The question of animal Los resultados apoyan positivamente las awareness. Mis observaciones de campo y preguntasplan teadas. experimentos sobre la simulaci6n de lesiones y Estos resultados no apoyan varias hip6tesis otros comportamientos antidepredadores en alternativas, por ejemplo que fingir el dafio es Charadrius es un desarrollo desdeeste pun to de vista. gobernado pr6ximamente por conflictos entre En mi investigaci6n me plantee las siguientes motivaciones de escape, agresi6n y cuidado de las cuestiones sobre aves: crfas, hip6tesis denominadas "de conflicto" y 1. ;,Quierenlos padres de Charadriusmelodus y "aproximaci6n/alejamiento". Estas comienzan a Ch. wilsonia alejar al intruso/depredador del apoyaruna interpretaci6n de intencionalidad del uso nido/joven? ;,Que clase de datos podrfan poner � de los comportamientos por estas especies � manifiesto una interpretaci6na prop6sito de este Charodrius. comportamiento? Algunos de los datos obtenidos Identificar experimentalmente el prop6sito del de las interacciones de los Charadius con comportamiento e investigar otros posibles estados intrusos humanos fueron: i,ES la direcci6n del mentales de los animales es una tareamuy diffcil. comportamiento de distracci6n de ala rota adecuado Para progresar en este empefio, necesitamos y paracausar que el intruso le siga para alejarlode las solicitamos los esfuerzos de muchas disciplinas crfas? ;,Donde, con respecto al campo visual del relacionadas incluyendo filosoffa, biologfa, intruso, los padres de Charadrius realizan el psicologfa comparada, experimental, de desarrolloy comportamiento de ala rota (BWD)? (respuesta: social, y por supuesto, etologfa.

66 Etologfa, Vol. 3, 1993

References Dennett, D.C., 1987. The intentional stance. Cambridge, Mass.: BradfordBooks, M.I.T. Press Dinsmoor,J.J., 1977. Notes on avocets and stilts in Arduino, P.J. & Gould, J.L., 1984. Is tonic TampaBay, Fla. FloridaField National, 5:25-30 immobility adaptive? Animal Behaviour, Gallup, G.G., Jr., Cummings, W.H., & Nash, 32:921-923. R.F., 1972. The experimenteras an independent Armstrong, E.A., 1947. Birddisplay and behaviour. variable in studies of animal hypnosis in London: Lindsay Drummond. chickens. Animal Behaviour, 20:166-169. Beer, C.G., 1991. From folk psychology to Gochfeld, M., 1984. Antipredator behavior: cognitive ethology. In: Cognitive ethology: The Aggressive and distractiondisplays of shorebirds. minds of other animals-essays in honor of In: Shorebirds: Breeding behavior am Donald R. Griffin: 19-33 (C.A. Ristau, Ed.). populations: 289-377 (J. Burger & B.L. Olla, Hillsdale, N. J.: Lawrence Erlbaum Associates. Eds.). New York: Plenum. Beer, C.G., 1992. Conceptual issues in cognitive Gould, J.L., 1982. Ethology: The mechanisms am ethology. In: Advances in the Study of Behavior, evolution of behavior. New York: W.W. Norton. Vol. 21:69-109 (B.P.J. Slater, J.S. Rosenblatt, Graul, W.D., 1975. Breeding biology of the C. Beer & M. Milinski, Eds.). San Diego: mountain plover. Wilson Bulletin, 87:6-31. Academic Press. Griffin, D.R., 1976. The question of animal Bennett,J., 1976. Linguistic behaviour. Cambridge: awareness. 2nd ed. 1981. New York: Rockefeller Cambridge University Press. University Press. Bennett, J., 1991. How is cognitive ethology Griffin, D.R.,1992. Animal minds. Chicago: Univ. possible? In: Cognitive Ethology: TheMinds of of Chicago Press. Other Animals-Essays in Honor of Donald R. Hennig, C.W., 1977. Effectsof simulated predation Griffin: 35-49 (C.A. Ristau, Ed.). Hillsdale, N. on tonic immobility in Ano/is carolinensis: The J.: Lawrence Erlbaum Associates. role of eye contact. Bulletin of the Psychonomic Boden, M., 1983. Artificial intelligence and animal Society, 9:239-242. psychology. NewIdeas in Psychology, 1:11-33 Herrnstein, R.J., Loveland, D.H., & Cable, C., Byrkjedal,I.,1991. The role of drive conflicts as 1976. Natural concepts in pigeons. Journal of a Mechanism for nest-protection behaviour in Experimental Psychology: Animal Behavior the shorebird Pluvialis dominica. Ethology, Processes, 2:285-302. 87:149-159 Ristau, C.A., 1983. lntentionalist plovers or just Curio, E., Ernst, U., & Vieth, W., 1978. The dumb birds? Commentary on Dennett, D.C. adaptive significance of avian mobbing. II. "Intentional systems in cognitive ethology: The Cultural transmission of enemy recognition in 'Panglossian paradigm' defended."Behavioral am blackbirds: Effectiveness and some constraints. Brain Sciences, 6: 373-375. Zeitschriftfur Tierpsychologie, 48: 184-202. Ristau, C.A., 1988 Thinking, communicating, and Dennett, D.C., 1978. Brainstorms. Cambridge, deceiving: Means to master the social Mass.: Bradford Books,M.I.T. Pres environment. In: Evolution of social behavior Dennett, D.C., 1983. Intentional systems in and integrative levels, T.C. Schneirla Conference cognitive ethology: the "Panglossian paradigm" Series: 213-240 (G. Greenberg & E. Tobach defended.Behavioral and Brain Sciences, 6:343- Eds.). Hillsdale, N.J.: Lawrence Erlbaum 390. Associates.

67 Ristau

Ristau, C.A., 1991. Aspects of the cogrutive Skutch, A.F., 1976. Parent birds and their young. ethology of an injury-feigningbird, the Piping Austin, Texas: University of Texas Press. Plover, In: Cognitive ethology: The minds of Tinbergen, N., 1969. The study of instinct. New other animals-essays in honor of Donald R. York: Oxford University Press. (original Griffin: 91-126 (C.A. Ristau, Ed.). Hillsdale, publication 1951). N.J.: Lawrence ErlbaumAssociates. Tolman, E.C., 1932. Purposive Behavior in Schneirla, T.C., 1972. An evolutionary and Animals and Men. New York: Appleton developmental theory of biphasic processes Century. underlying approach and withdrawal I. In: Walker, J., 1955. Mountain plover. Audobon. Selected writings of T. C. Schneirla: 297-339 57:210-212. (L.R. Aronson, E. Tobach, J.S. Rosenblatt & Walters, J., 1980. The evolution of parental D.S. Lehrman, Eds.). San Francisco: W.H. behavior in lapwings. Ph.D. dissertation. Freeman & Co. (reprinted from Nebraska University of Chicago, Chicago, Illinois. symposium on motivation: 1-42 (M.R. Jones, Watanabe, S., Lea, F.E.G. & Dittrich, W.H. in Ed.) Lincoln, Neb.: University of Nebraska press. What can we learn from experiments on Press. pigeon concept discrimination? In: Avian vision Searle, J.R., 1980. Minds, brains and programs. andcognition: xxx-xxx (Bischof, H.J. & Zeigler, Behavioral andBrain Sciences. 3:417-457. H.P., Eds.). Cambridge, Mass.: MITPress.

(Recibido : agosto 1993)