Early predispositions for social stimuli: the case of face perception
Year: 2010
Type: dissertation
Abstract: Converging evidence from several different species has led to the proposal that some newborn vertebrates, including humans and domestic chicks, have visual predispositions to attend to the head region of conspecifics (or of animals of other species that share similar characteristics) (Johnson and Horn 1988; Morton and Johnson, 1991; Sugita 2008). More specifically it has been claimed that these newborn preferences are (i) domain-relevant to the extent that other naturally occurring stimuli do not draw attention in the same way, (ii) are not based on rapid early learning, but are present from birth/hatching, and (iii) may be mediated by phylogenetically ancient brain routes common to many vertebrates. One of the most common criticisms of the work supporting domain-relevant face biases in human newborns is that the majority of the studies conducted (with some notable exceptions) regard newborns of more than a few hours old. Thus, it remains possible that very rapid early learning contributes to the specificity of some of the effects observed (Nelson, 2001). This criticisms of the data from human newborns can be addressed by testing newly hatched visually-deprived chicks whose preference for visual stimuli can be assessed prior to any other visual experience with faces, a procedure obviously not viable with human newborns for ethical reasons. The main aim of present study is thus to investigate from a new perspective and employing an animal model (the domestic chick) a long-debated issue: the specificity vs. non-specificity of the perceptual factors involved in face preferences. One central issue in cognitive science is how brain processes knowledge of specific domains, as for example the visual information regarding faces. Some existing evidences seem to indicate that faces are processed by anatomically and/or functionally dedicated domain-specific brain circuits, in humans and possibly also in animals. This idea seems also to be consistent with the striking processing abilities that human beings have demonstrated in face perception. It seems plausible that, because of the relevance of faces for social life, natural selection led to the evolution of innate face-specific devices that are available prior to any postnatal experience. In fact, a number of studies have demonstrated that, even from a few hours or minutes after birth, a schematic pattern representing an upright face elicits greater visual attention in human newborns than do similar stimuli presenting the same internal face-features in scrambled positions or arranged upside-down (see Morton and Johnson, 1991 for a review). According to Morton & Johnson (1991) this evidence would indicate that human newborns respond to the unique structural configuration of the face (i.e. to facedness) due to a face-detecting mechanism (CONSPEC) that contains a configural representation of the structure of the face’s inner features (three darker areas in an upside-down triangular configuration). Similarly, domestic chicks have been shown to posses an unlearned representation for the appearance of a social object. Such a representation would guide the imprinting process ensuring that chicks would imprint on an animate object instead of on an inanimate feature of the environment. In particular, this unlearned representation of the appearance of a social object would include a very broad structural description of the features contained in the head region of a vertebrate creature (e.g. Johnson & Horn, 1988). It has thus been hypothesised that the same mechanism described to explain face preferences in newborn babies (CONSPEC) would underlie also spontaneous preferences of visually naïve chicks (Morton and Johnson, 1991). However, in the human developmental literature it has often been debated whether newborns’ preference for faces would not be simply the secondary effect of some more general and not domain-specific visual preference displayed by newborn vertebrates. This kind of domain-general preference would direct the organism’s attention toward any stimulus presenting a certain physical attribute, regardless of whether this stimulus is a face or a non-face object (Acerra, Burnod and de Schonen, 2002; Kleiner, 1987; Macchi Cassia et al., 2008; Simion et al., 2002; Turati et al., 2002). In my thesis I will thus assess the role of three perceptual properties (that are considered influential in the developmental literature) in determining face preferences in visually naïve domestic chicks. The role of the vertical asymmetry of inner elements Recently, in the developmental literature, it has been argued that newborn infants’ preferences for faces could be a secondary effect determined by non-specific biases due to constraints imposed by the immature visual system of the child. In particular, Turati et al. (2002) provided evidence that the preference for face-like stimuli could be determined by an “up-down bias” (Simion et al., 2002) that directs the babies’ attention toward any configuration presenting more elements in its upper part (a “top-heavy configuration”, as opposed to a “bottom-heavy configuration”, presenting more elements in its lower part). Thus, we decided to investigate whether visually naive domestic chicks spontaneously prefer schematic faces or other top-heavy stimuli and whether such a preference is determined by an up-down bias, as suggested by some literature on newborn babies (Turati et al., 2002). Eggs were incubated and hatched in the darkness and after hatching, for about 24 h, chicks were placed in an uniform white cage whose walls and floor were made of opaque white paper. The only stimulus available to chicks during this rearing period was an artificial orange cardboard imprinting object that presented an identical outline with respect to stimuli that will be employed at test (the outline of a schematic face and neck), but no inner feature. Thus, the imprinting object could not provide chicks with any bias for one of the test stimuli or with any information regarding faces’ inner features. During the second day of life chicks underwent a spontaneous preference test, in which each chick had to choose between one of two stimuli that were simultaneously presented at the two opposite ends of a longitudinal runway. The runway was divided into three virtual sectors. At the beginning of the test each chick was placed in the central sector of the apparatus. In order to approach one of the two stimuli the chick had to leave the central sector of the apparatus and enter one of the two lateral sectors adjacent to the stimuli. Thus, entrance and residence of the chick in one of the side compartments indicated a preference for the adjacent object. The test lasted for 6 minutes. Stimuli employed at test were identical to the imprinting object in size, shape and colour, but presented three inner features (three black squares) that were absent in the imprinting object. The two stimuli in each test-pair differed from one another only in the configuration created by the three black squares (in line with standard stimuli used in the developmental literature). Changing the disposition of the three inner elements it was thus possible to test chicks preferences for face-like vs. non-face like objects controlling also for the role of the up-down bias. Each chick was tested only once and was thus exposed to only one pair of stimuli. With this procedure we demonstrated that chicks prefer to approach a schematic face-like configuration similar to those used in newborn studies with respect to another equally top-heavy, but non-face-like, stimulus (Exp. 1). In line with this result, in a second experiment chicks did not show any evidence of a preference for a top-heavy stimulus over a bottom-heavy stimulus, if none of them represents a face (Exp. 2). Moreover, chicks preferred to stay near a bottom-heavy and face-like stimulus over a top-heavy non-face-like one (Exp. 4). Thus, results obtained in this first cycle of experiments demonstrated that domestic chicks, that are visually naïve with respect to faces’ inner structure, show a spontaneous preference for schematic face-like stimuli that resemble the representation of a face as theorised by Morton and Johnson (1991). Moreover, the vertical asymmetry of inner elements (i.e. the up-down bias, Simion et al., 2002; Turati et al., 2002) does not seem to have a crucial role in the expression of this preference in the animal species that we tested. The role of the spatial frequencies composing stimuli During the past decades another alternative interpretation, based on the so called linear system model (LSM, e.g. Kleiner, 1987), has been proposed to explain face preferences in newborn babies, as opposed to the CONSPEC model proposed by Morton and Johnson (1991). According to this interpretation (confirmed also by the results of a recent neural network model by Acerra, Burnod and de Schonen, 2002) faces would elicit preferential attention in newborn babies simply because they happen to be composed of the range of spatial frequencies more visible to newborns. In previous studies existing in the literature, stimuli were controlled for properties such as vertical symmetry or for the presence of structure (Morton and Johnson, 1991; Farroni et al., 2005; Rosa Salva, Regolin and Vallortigara, 2009). The research conducted in similar studies already generated a good amount of evidence for the role of the above mentioned properties. We thus decided to concentrate our efforts on the role of another potentially relevant perceptual property, namely, spatial frequency composing stimuli. The use of stimuli that exactly match spatial frequencies between face-stimuli and control-stimuli, comparing faces to frequency matched visual noise, is already a common standard in works investigating neural correlates of face perception (Csibra et al., 2004; Blasi et al., 2007). However, to the best of our knowledge, this approach has not yet been s
Author: Orsola Rosa‐Salva
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W54940221