ORIGINS OF SENSITIVITY TO PHYSICAL CAUSALITY AND ANIMACY:Evidence from newborn babies and newly-hatched chicks
Year: 2010
Type: dissertation
Abstract: Identifying animated subjects and causal relations between them ought to be fundamental for the survival of animals. Humans' striking ability to attribute “animacy” (i.e. the capability of an object to have self-produced motion and of being a causal-agent) and causal relations to moving objects solely on the basis of their pattern of motion is well attested (Heider and Simmel, 1944; Michotte, 1963; Schlottmann, Ray, Mitchell and Demetriou, 2006). Nothing is known, though, about the origin of this sensitivity, which could be either implicit or the outcome of some learning processes. A further source of debate concerns whether this capability is present both in human and non-human animals. For such reasons, the purpose of this thesis was to investigate the sensitivity to physical causality from a comparative perspective, employing both newborn babies and newly-hatched domestic chicks (Gallus gallus) as experimental subjects. Moreover, chicks’ sensitivity to self-propelled motion as a cue to animacy attributions was investigated. Sensitivity to physical causality in a comparative perspective Attribution of causality (i.e. appreciation of the interaction between any two events in terms of a cause-effect relationship) appears to be a powerful mechanism for perceiving every day experiences, in such a way we may make sense of almost any aspect of the physical world. Understanding causal representations has long been a challenge both for philosophers (Aristotele, Physics) and psychologists (see Sperber, Premack and Premack, 1995). Psychological research has both focused on exploring higher-level cognitive processing (i.e., the dynamics of causal inferences, Sperber et al., 1995), whilst at the same time much critical work on causality has concentrated on the role of perceptual processing (Michotte, 1963). As regards the latter perspective, the earliest ideas were put forward by Michotte (1963), who described the most well known example of perceptual causality, the “Launching Event”. In this event an object, “A”, starts moving towards a second object, “B”, in a straight line and at constant speed. When A comes into contact with B, A stops moving and B moves off in the same direction and at the same or slightly lesser speed in which object A had been moving. When this event is viewed, most people see object A “launching” B, thus causing its movement. According to Michotte, the “causal impression” triggered by the Launching Event is automatically produced by a “perceptual input analyzer”, that would provide an original idea of cause: a causal connection is experienced as a result of what is a perceptual process, rather than a conceptually mediated inference (Michotte, 1963). The fact that this perception is hardly affected by higher-level beliefs or intentions (whereas it seems to depend on a highly constrained collection of visual cues; e.g. Choi and Scholl, 2004, 2006; Schlottmann et al., 2006; White, 2006), would support a precocial ontogenetic (and phylogenetic) origin. Michotte himself, even though his work was based on verbal reports by adult observers, took a strong nativistic position on the origin of causal representations, claiming that the perceptual mechanism for causal impressions was innate, and furthermore that the output of this mechanism was the source of all subsequently developing causal representations (Michotte, 1963). Recently, a growing interest in the earliest causal representations has produced a range of experimental data in developmental studies (Leslie, 1984; Leslie and Keeble, 1987; Cohen and Amsel, 1998; Cohen and Oakes, 1993; Cohen, Rundell, Spellman and Cashon, 1999; Oakes, 1994; Oakes and Cohen, 1990), such data used in support of two very different models of the origin of causal perception. On the one hand, Leslie and colleagues (Leslie and Keeble, 1987) suggested that there would be an innate, domain-specific visual module that operates ‘‘automatically and incorrigibly upon the spatio-temporal properties of events yet producing abstract descriptions of their causal structure” (Leslie, 1988). In contrast, Cohen and colleagues suggested a constructivist account: the perception of causality would be demonstrably not innate and not domain specific (Cohen and Amsel, 1998; Cohen, Chaput and Cashon, 2002). The causality of events could be either perceived as a result of perceptual and cognitive development (Oakes and Cohen, 1990) or constructed during development from non-causal resources within core cognition (Carey, 2009). Until now, however, perception of physical causality has never been investigated in subjects under conditions in which previous experiences have been effectively controlled for (e.g., in newborn babies). Moreover, nothing is known specifically about the capability to perceive physical causality in animal species other than our own. For these reasons this thesis was aimed at investigating spontaneous preferences for physical causality both in newborn babies and in newly-hatched domestic chicks. The domestic chick seemed to be an ideal subject to address this issue, since it is a highly visual and precocial species, and has shown capable of perceiving shapes defined by motion information alone, much as humans can (e.g. Mascalzoni, Regolin and Vallortigara, 2009; Vallortigara, Regolin and Marconato, 2005). Two experiments were conducted in which chicks’ spontaneous preferences for a Causal Stimulus (i.e., Launching Effect) vs a Non-Causal Stimulus were assessed. All subjects came from eggs incubated and hatched in total darkness under controlled conditions; this prevented any prior visual experience before testing. On their second day of life, chicks underwent a spontaneous preference test, in which they had to choose between two video animations (a Causal Animation and a Non-Causal Animation), simultaneously presented at the two opposite ends of a runway. At the beginning of the test chicks were taken individually from the dark hatchery and placed in the central sector of the runway, which was virtually subdivided into three areas. The behaviour of the chick was observed for a total of six consecutive minutes. In order to approach one of the testing animations, the subject had to leave the central area and enter one of the side areas. If the chick remained in the mid compartment, this was deemed to indicate no preference, whereas if it went into an end-side compartment this was regarded as a preference for the stimulus placed at that end of the runway. All the stimuli were video animations involving two square-shaped objects of the same dimension and colour. For the Launching Stimulus one of the 2 objects (“A”) started to move towards a second, stationary object (“B”). Immediately after contact, A stopped and B moved off in the same direction and at the same speed that A previously showed. Thus, when seen together this stimulus reproduced the typical Michotte’s Launching Effect, in which adult humans report as seeing a physical-causality relationship between the two movements. In Exp. 1, a “Delay Stimulus” was used as the Non-Causal Stimulus. It was identical to the Launching Stimulus except for a 2-sec-delay between the moment of contact and the initiation of motion of object B: such a delay is known to abolish any impression of physical causality in adult humans (Michotte, 1963). The Non-Causal Stimulus employed in Exp. 2, instead, was a “Passing” Stimulus in which the initial moving object (A) moved towards the stationary object B, passed in front of it and reached the final position occupied by B at the end of the Launching sequence. In this animation no physical causality was implied but the spatio-temporal continuity typical of the Launching Effect was preserved. In Exp. 1, thus, chicks preference between a Launching and a Delay Stimulus was tested, whether in Exp. 2 chicks were asked to choose between a Launching and a Passing Stimulus. The results did not reveal any spontaneous preference for either stimuli in chicks shortly after hatching. Therefore, it seems that newly-hatched domestic chicks have no preferences for approaching a causal event compared to a non-causal event. The other branch of the thesis dealt with newborn babies’ sensitivity to physical causality. Three experiments were conducted, all employing an infant-control preferential looking technique which allowed for the testing of newborns’ spontaneous preferences for a Causal vs a Non-Causal stimulus by measuring the length of time of their fixations on the stimuli. In this procedure, a baby was sat on the experimenter’s lap in front of a monitor. Two testing stimuli were presented simultaneously, on the left and the right side of the monitor. Each newborn was given two paired presentations of the test stimuli in which the position of the stimuli was reversed (the initial left-right order of presentation was counterbalanced across subjects). The duration of the infant’s fixations on each stimulus was recorded on-line and subsequently re-coded off-line by a second experimenter. Each trial ended when the newborn did not fixate on the display for at least 10 s. In Exp. 1 newborns’ spontaneous preference for a Launching Stimulus vs a Non-Causal Control Stimulus was investigated. Two computer-presented animation sequences were created, each sequence featuring two identical gray discs (“A” and “B”). The Launching video animation had the usual spatio-temporal features of a Launching Effect (Michotte, 1963). Both discs, moreover, moved with identical speed and covered the same distance. As described previously, in this sort of display adult subjects perceive disc A as “launching” disc B hence causing its movement. The Non-Causal Control animation was identical to the Launching one except for the fact that, maintaining the same starting positions, the order of the displacement of the two discs was swapped temporally: thus B moved first and A started its movement only after B had stopped. In this sequence
Author: Elena Mascalzoni
Language: English
Subfield: Developmental and Educational Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W120415079