Experimental evidence of structural representation of hands in early infancy

Abstract

Hands convey important social information, such as an individual’s emotions, goals, and desires, are used to direct attention through pointing, and are a major organ for haptic perception. However, very little is known about infants’ representation of human hands. In Experiment 1, infants tested in a familiarization/novelty preference task discriminated between images of intact hands and images that contained first-order structure distortions (i.e., with locations of fingers altered to result in an unnatural configuration). In Experiment 2, infants tested in a spontaneous preference task exhibited a preference for scrambled hand images over intact images, indicating that 3.5-month-olds have gained sufficient sensitivity to the configural properties of hands to discriminate between intact versus scrambled images without any training in the laboratory. In both procedures, infants’ performance was disrupted by the inversion of images, suggesting that infants’ performance in upright conditions was not based on low-level features. These results indicate that sensitivity to the structure of hands develops early in life. This may lay the foundation for the development of the functional use of hand information for social communication.

Keywords

Social perception body-part perception representation of hands infant hand identification

Sensitivity to Hand Structure Information in Infancy

Like faces and bodies, hands convey a great deal of social information, such as an individual’s goals, emotions, and desires (Morrisey & Rutherford, 2013). Hands also play a major part in directing attention through pointing (Daum, Ulber, & Gredebäck, 2013). Moreover, starting early in life, hands are used to explore the world haptically (Molina et al., 2015; Rochat, 1989). However, very little is known about infants’ representation of hands. This is in contrast to our understanding of the development of knowledge about faces. For example, it is known that infants exhibit sensitivity to first-order structural information in faces very soon after birth. First-order information refers to the basic spatial configuration of social stimuli (Diamond & Carey, 1986). For example, all faces share the same structure, with major features such as the eyes, nose, and mouth arranged in a particular order. This structural information allows us to efficiently identify and deal with socially important stimuli such as faces in our environment. Goren, Sarty, and Wu (1975) reported that as soon as 9 min after birth, infants prefer to look at structurally intact face-like stimuli over distorted stimuli (i.e., scrambled faces; also see Johnson, Dziurawiec, Ellis, & Morton, 1991). Thus, infants exhibit sensitivity to the first-order structure of faces soon after birth.

Bodies and body parts, such as hands, are also defined by particular first-order configurations (Reed, Stone, Bozova, & Tanaka, 2003; Slaughter & Heron, 2004; Zieber, Kangas, Hock, & Bhatt, 2015). Zieber, Kangas, Hock, and Bhatt (2015) found that 3.5-month-olds are sensitive to first-order relations in whole human bodies. Infants in that study discriminated between images of intact bodies and those in which the parts were scrambled to form new configurations. It is unknown, however, whether infants of this age are sensitive to structural properties of hands. We addressed this issue in the current study by examining whether 3.5-month-olds distinguish between images of intact hands and those in which the fingers are unnaturally displaced. If infants exhibit sensitivity to the images depicting scrambled hands, it would indicate that the structural properties of hands are part of the representation of humans early in life. In the following paragraphs, we briefly review studies that have examined the early development of sensitivity to the structure of faces and bodies followed by studies that have explored infants’ processing of hands.

Development of Face and Body Structure Processing in Infancy

There is reason to believe that infants are born with a template of the human face that directs early attention to and engenders a preference for facial (or at least face-like) stimuli early in life (Morton & Johnson, 1991; Simion & Di Giorgio, 2015). For example, newborns look longer at images that contain more information in the top half (resembling a normal face configuration) than at images that contain more information in the bottom half (resembling an inverted face), thereby responding to one of the defining first-order configural characteristics of a face (Simion, Valenza, Macchi Cassia, Turati, & Umiltà, 2002; Turati, Simion, Milani, & Umiltà, 2002). Presumably, faces are so recognizable even immediately following birth because of their first-order configurations. This possibility is supported by the studies described above in which newborns discriminated between faces with intact featural arrangements versus those with scrambled features (e.g., Goren, Sarty, & Wu, 1975; Johnson et al., 1991).

Body processing is similar to face processing in many respects. For example, Reed, Stone, Grubb, and McGoldrick (2006) found that changing first-order spatial relations of bodies by scrambling part positions interferes with adults’ discrimination of posture changes. Thus, as in the case of faces, the specific spatial relations between body parts are critical. There is some uncertainty, however, when it comes to pinpointing the developmental trajectories of body representation in infancy. Slaughter and colleagues (2004, 2012) suggest that visuo-spatial knowledge of human bodies is slow to develop and substantially lags the development of facial knowledge. However, other research suggests that at least some aspects of body knowledge are evident early in life. For instance, infants as young as 3 months old are sensitive to the overall organization of body parts (Gliga & Dehaene-Lambertz, 2005; Zieber et al., 2015), indicating knowledge of the first-order structure of bodies. Infants as young as 5 months old also process body information holistically, demonstrated in their detection of changes in the orientation of limbs in the context of the whole body but not of parts presented in isolation or in the context of scrambled images (Hock, White, Jubran, & Bhatt, 2016). Based on these and other such findings, Bhatt, Hock, White, Jubran, and Galati (2016) suggest that body knowledge develops fairly quickly, based either on a general social cognition system that prepares infants to process critical information from a variety of sources of social information (bodies, faces, and voices) or a body-specific knowledge system that benefits from correlated information from the rapidly developing face-processing system.

While the development of knowledge about the structure of human bodies as a whole may not necessarily proceed in tandem with knowledge about particular parts, such as the hands, it is likely that there is a correlation. If this is the case then it is possible that 3.5-month-olds would exhibit sensitivity to the structure of hands given that infants of this age respond systematically to changes in body structure (Gliga & Dehaene-Lambertz, 2005; Zieber et al., 2015).

Hand Processing in Infancy

Many studies have examined infants’ systematic attention to hands (e.g., Aslin, 2009; Deák, Krasno, Triesch, Lewis, & Sepeta, 2014; Frank, Vul, & Saxe, 2012; Slaughter & Neary, 2011; Yoshida & Smith, 2008; Yu & Smith, 2013). Infants attend to hands from birth (Van der Meer, 1997; White, Castle, & Held, 1964), suggesting early exposure to this socially significant body part. Slaughter and Neary (2011) found no statistical difference in 4- and 6-month-olds’ attention to hands versus faces, although infants behaviorally responded more to faces than to hands. These researchers also reported that infants in their study looked at locations above hands as if they expected hands to be associated with faces. Aslin (2009) found that 4- and 8-month-olds playing with blocks looked at hands more than 20% of the time, which was more than they looked at faces.

Infants also exhibit sensitivity to hand movement and functions early in life. Using point-light displays, Fox and McDaniel (1982) found that 6-month-olds prefer to look at the biological motion pattern of a hand grasping an object rather than a foil stimulus, thus showing that infants are sensitive to biological motion patterns of hands starting early in life. Also, newborns look longer at an impossible hand movement over a possible one (Longhi et al., 2015). In addition, young infants appear to be sensitive to the goal-oriented movement of hands. When 6-month-old infants were habituated to a hand reaching for and grasping one of two toys, infants looked longer during test trials when the actor’s arm grasped a different toy after following the same path compared to when the arm followed a different path to the same toy (Guajardo & Woodward, 2004). However, this effect was not found when the procedure was performed with a claw or a hand wearing a glove. This finding indicates that when the surface features of the hand are disrupted, it affects the way infants interpret grasping motions. This suggests that even young infants have a fairly veridical representation of hands and their functions.

Infants’ own actions may also contribute to their representation of hands. Systematic hand actions are evident very early in life (Molina et al., 2015). For instance, Rochat (1987) found that newborns explore soft and hard objects differently with their hands, suggesting sensitivity to the feedback they receive from their hand actions. Moreover, by 3 months of age, infants begin to engage in exploratory behaviors such as scratching, which vary with the size and texture of objects (Rochat, 1989). Thus, infants begin to systematically explore the environment with their hands quite early in life. The feedback that young infants receive during the process of exploration might reveal characteristics of hands that contribute to infants’ representation of the structure of hands.

Research on neural structures such as mirror neurons that respond to hands is also relevant to our understanding of hand representation in infancy. Of particular interest is the fact that certain areas of Japanese monkeys’ brains are directly responsive to hand movements (Borra, Gerbella, Rozzi, & Luppino, 2017; Taira, Mine, Georgopoulos, Murata, & Sakata, 1990) because it suggests there may be structures of the brain dedicated to processing hand signals. There is also evidence of human analogues of monkey brain structures that respond to actions and observations of actions of body parts such as the hands and feet (Fox et al., 2016; Molenberghs, Cunnington, & Mattingley, 2012). Meltzoff et al. (2018) report that touches to 7-month-olds’ hands produce activation in the contralateral hand areas of the somatosensory cortex, suggesting neural structures that represent hands early in life. Meltzoff et al. also found that 7-month-olds’ response to viewing other people’s hands being touched is systematically reflected in brain structures associated with self-other processing. It is possible such structures that represent hands are innately functional or rapidly developing early in life.

In summary, the studies described in this section indicate that hands are salient stimuli early in life, are used to systematically explore objects, and may be neurally represented in a specialized manner in young infants. These findings beg the question of the exact nature of infants’ hand representation. As noted earlier, the first-order structure (i.e., the canonical arrangement of parts) is critical for the identification of social stimuli such as hands. Thus, one important question about the development of hand representation is whether young infants are sensitive to the first-order structure of hands. We addressed this issue in the current study.

A Theory of Body-Part Knowledge Development

Although many theorists have addressed the development of knowledge about faces and bodies and social perception in general early in life, to our knowledge, only one has directly addressed the development of knowledge about body parts. Meltzoff and his colleagues have theorized that infants have innate knowledge about body parts or rapidly acquire body knowledge through observation and imitation very early in life (Marshall & Meltzoff, 2015; Meltzoff, 2011). Such knowledge could include that of the structure and function of specific parts (e.g., hands). In particular, Meltzoff and Moore’s (1997) “active intermodal mapping” hypothesis and Meltzoff’s (2011) “like-me” model of social cognitive development assume that “organ identification” occurs in newborns. Meltzoff and Moore (1997) specifically suggest that hands might be an organ that is identified early in life. As noted earlier, identification of social stimuli occurs through the recognition of features and the structural arrangement of features. Thus, if young infants recognize hands as suggested by Meltzoff’s model, then it is possible they would be sensitive to the structural configuration of hands. We tested this possibility in the current study.

Experiment 1

The first step in understanding infants’ sensitivity to the first-order structure of hands is to determine whether they discriminate between two hand images based on first-order distortions. We used a familiarization/novelty preference procedure to examine this possibility in Experiment 1. Infants were familiarized with an image of an intact hand and tested for their preference between this image and another in which the first-order structure of a hand was scrambled. We chose to study 3.5-month-olds because hands have been shown to be particularly salient during play by 4 months of age (Aslin, 2009). Furthermore, 3.5-month-olds exhibit sensitivity to first-order relational information in bodies (Zieber et al., 2015). It is possible that such knowledge of first-order relational information in bodies extends to hands. Finding that infants can discriminate an intact hand from its first-order distorted counterpart would be an initial indication that infants are sensitive to the normal configuration of hands. In contrast, if infants are unable to discriminate between intact and scrambled hand images, then it would suggest that they do not have a clear representation of the structure of human hands by 3.5 months.

However, discrimination between intact and scrambled images by itself does not necessarily indicate that infants are sensitive to hand structure because their performance may solely reflect the discrimination of a learned arbitrary image from a novel image. In prior studies, evidence of specific representation of social stimuli has been obtained by examining whether infants exhibit an inversion effect (e.g., Bhatt, Bertin, Hayden, & Reed, 2005; Zieber et al., 2015). For example, Zieber et al. (2015) concluded that 3.5-month-olds are sensitive to relative sizes of body parts by finding that infants discriminate between intact images and those with changes in part sizes, but only when images are presented upright and not when they are inverted. Such inversion effects indicate that infants’ performance with upright stimuli is not solely discrimination between two arbitrary images that may differ in terms of some low-level image features such as symmetry, but is based on the specific orientations of social stimuli.

Thus, in Experiment 1 we tested a separate group of infants with inverted versions of the images used in the upright condition. Morrisey and Rutherford (2013) found that adults exhibit an inversion effect when tested on hands, suggesting their representation of hands includes a specific orientation. However, as noted by Morrisey and Rutherford (2013), unlike faces and bodies, hands do not necessarily have a canonical top-bottom orientation because they are seen and function at various angles. Thus, it is not possible to predict a priori whether infants would exhibit an inversion effect with hands. Nevertheless, testing infants on inverted images would be a good control because if they exhibit an inversion effect, it would suggest their performance on upright images is not based on an arbitrary discrimination between two kinds of stimuli that differ in terms of low-level image features. If, however, infants fail to exhibit an inversion effect, then one cannot conclude that infants’ performance was based on hand representation.

Method

Participant recruitment and test procedure used in this and the following experiment were approved by the Institutional Review Board of the University of Kentucky.

Participants

The participants in this experiment and in Experiment 2 were full-term infants (gestation > 38 weeks) with a minimum birth weight of 2500 grams. In total, 32 3.5-month-old infants (mean age = 106 days, SD = 9.03; 18 female) successfully completed this experiment. Data from five additional infants were excluded due to side bias (i.e. looking greater than 95% to one side; n = 2), not looking on one test trial (n = 1) or fussiness (n = 2). Infants were recruited through birth announcements in the local newspaper and from a local hospital. The majority of the participants were from middle-class Caucasian families.

Stimuli

The stimuli were images of hands positioned in intact and scrambled configurations (Figure 1). The intact images were created using Poser 2.0 software (Curious Labs, Santa Cruz, CA) and were modified using Adobe Photoshop CS5 to generate the scrambled images. Four unique hand gestures were constructed using four different hands. A scrambled version of each hand was constructed by altering the location of one or more fingers (Figure 1). The reorganization was such that it altered the first-order configuration (i.e., overall gestalt) of the hands. On average, the scrambled hands subtended horizontal and vertical visual angles of 13.56º and 16.81º, and the intact hands subtended horizontal and vertical visual angles of 13.02º and 17.00º, respectively. Infants in the upright condition were familiarized and tested with images depicting hands in upright orientations, whereas those in the inverted condition were familiarized and tested with inverted stimuli that were similar to those used by Morrisey and Rutherford (2013) in which adults exhibited an inversion effect.

Figure 1.

Examples of the upright (A) and inverted (B) stimuli used in Experiment 1. During familiarization, infants saw two identical images of an intact hand (top panel). During the test, infants saw the familiar intact hand paired with a scrambled image (bottom panel).

Apparatus and Procedure

Infants were seated approximately 45 cm in front of a 50 cm computer monitor in a darkened chamber. They were seated on the lap of a parent, who was wearing opaque glasses that prevented them from seeing the images on the screen and potentially biasing infants’ looking patterns. Infants were familiarized during two 30 s trials in which they were exposed to two identical intact hand images, one on the left and one on the right side of the screen. They were then tested on two 10 s paired-comparison trials to determine whether they discriminate between the intact familiarization hand image and a novel hand image that had been scrambled. Given previous work documenting 3.5-month-olds’ preferences for scrambled over intact bodies (Zieber et al., 2015), we expected that infants in the current study would have a greater chance of displaying their ability to discriminate when familiarized to intact hands. That is, given an a priori preference for scrambled images, infants familiarized to scrambled images and tested with intact images as novel images may not exhibit discrimination as indexed by novelty preference because preference for scrambled images could compete with novelty preference for the intact images during the test. Thus, infants in Experiment 1 were only familiarized to intact hand images and tested with novel scrambled images.

At the beginning of each familiarization and test trial, infants saw alternating purple and green shapes to direct their attention to the center of the screen. Once the infant’s attention was on the center of the screen, as determined by the experimenter via a live video feed, the first familiarization trial began following a key press by the experimenter. After two familiarization trials, each lasting until the infant accumulated 30 s of looking, infants were tested on two 10 s trials in which an intact hand image and its corresponding scrambled image appeared on the screen. One image appeared on the left side of the screen and the other on the right. The left/right location of the scrambled hand on the first test trial was counterbalanced across infants and then switched on the second trial. Each infant was tested on one of four pairs of hand stimuli. Also, half of the infants were familiarized and tested with upright images (upright condition) whereas the others were tested with inverted images (inverted condition; see Figure 1).

A video camera located on top of the monitor recorded the infant’s looks. A naïve coder unaware of the left-right location of the stimuli then coded the infant’s test performance offline using a DVD player slowed to 25% of the normal speed. A second coder verified coding reliability for 25% of the infants. The Pearson correlation between coders was .97.

Results and Discussion

The dependent measure was the percent preference for the novel hand (see Table 1). This was calculated by dividing the total looking time to the novel hand across both test trials by the total looking time to both the intact and scrambled hand across both trials; this number was then multiplied by 100. An outlier analysis (Tukey, 1977; using SPSS version 22.0) did not reveal any outliers. In the following analyses, we contrasted performance in each condition to the chance level of 50% and directly compared novelty preference scores in the upright and inverted conditions.

Table 1.

Mean time to accumulate 30 sec of familiarization (seconds), mean look duration (seconds) to the novel and familiar hand images, and percent preference for the novel stimulus in Experiment 1 (upright and inverted stimuli).

	N	Time to accumulate 30 s of familiarization mean	Mean look duration to intact stimulus	Mean look duration to scrambled stimulus	Mean scrambled percent preference	t	95% confidence interval for scrambled percent preference	Effect size Cohen’s d
Upright	16	Familiarization 1: 47.41	6.43	8.50	57.28	2.45*	50.94–63.62	0.61
		Familiarization 2: 43.73
Inverted	16	Familiarization 1: 44.88	6.92	7.01	49.61	-0.13	44.69–55.90	0.03
		Familiarization 2: 41.41

Note. *p < .05, two-tailed; significantly different from the chance level of 50%.

One-sample t tests comparing each group’s score against the chance level of 50% revealed that infants in the upright condition discriminated between the intact and scrambled hand images by exhibiting a novelty preference score (M = 57.28%, SE = 2.98; see Table 1) that was significantly different from chance, t(15) = 2.45, p = .027, two-tailed, d = 0.61. In contrast, infants in the inverted condition failed to exhibit discrimination: their novelty preference score (M = 49.61%, SE = 2.96; see Table 1) was not significantly different from chance, t(15) = -0.134, p = .895, two-tailed, d = 0.03.

Moreover, an independent-samples t test revealed that the novelty preference score of infants tested on upright images was significantly greater than the score of infants tested on inverted images, t(30) = 1.83, p = .038, one-tailed, d = 0.65. We used a one-tailed test when comparing performance in the upright and inverted conditions because there is no expectation that infants in the inverted condition would perform better than infants in the upright condition. Moreover, although 13 out of the 16 infants in the upright condition had a mean preference for the novel hand that was above chance (binomial p = .01), only eight out of 16 infants in the inverted condition had a mean preference for the novel hand that was above chance (binomial p = .60). Thus, similar to findings with adults (Morrisey & Rutherford, 2013), infants exhibited an inversion effect, with superior performance on upright than on inverted images.

Thus, in Experiment 1 infants in the upright condition discriminated between intact and scrambled hand images but those tested on inverted images failed to discriminate. These results indicate that 3.5-month-olds are sensitive to the structure and orientation of hands. Moreover, their discrimination performance is not based on arbitrary low-level image features such as symmetry.

Experiment 2

The results of Experiment 1 indicate that infants at 3.5 months of age are sensitive to the first-order structure of hands when tested in a familiarization/novelty preference procedure. It is still unknown, however, whether infants come into the laboratory with knowledge about the typical first-order structure of a hand. One way to examine this issue is to test infants for a preference between intact and scrambled hand images without any familiarization. A spontaneous preference between intact versus scrambled hand images would indicate that infants’ representation of the first-order structure of hands is not dependent upon training in the laboratory.

Thus, in Experiment 2, infants were tested for their spontaneous preference between intact versus scrambled hand images. Zieber et al. (2015) found that 3.5-month-olds have a spontaneous preference for scrambled whole-body images over intact images. If this preference extends to body parts such as hands, then infants would be expected to have a similar preference for scrambled hand images over intact images. We tested this possibility in Experiment 2.

Half of the 3.5-month-old infants in Experiment 2 were tested with upright hands, whereas the other half were tested with inverted hands. If infants exhibit a preference between intact and scrambled hands when they are upright but not when they are inverted, it would suggest that 3.5-month-old infants possess sufficient knowledge of the configural properties of hands to be spontaneously sensitive to disruptions of first-order structure. Such a finding would additionally reinforce the evidence of first-order relational knowledge of hands found in Experiment 1.

Method

Participants

Thirty-six 3.5-month-old infants (mean age = 108 days, SD = 15.56; 18 female) successfully completed this study and were recruited in the same manner as those in Experiment 1. An additional 3.5-month-old’s data were excluded from analyses due to experimenter error. The majority of the participants were from middle-class Caucasian families.

Stimuli, Apparatus, and Procedure

In this experiment, infants were tested on six 10 s test trials in a spontaneous preference paired-comparison procedure. A spontaneous preference procedure was chosen because we sought to examine infants’ sensitivity to the structure of hands without any training or exposure in the laboratory. The stimuli used in this study included the four sets of intact and scrambled images used in Experiment 1 (Figure 2). In addition, two new pairs of intact or scrambled hand images were added to increase the generalizability of the findings and to allow infants to be tested on more trials. The inverted stimuli were created by flipping the upright images about the horizontal axis (Figure 2). The six pairs of intact or scrambled stimuli were divided into two sets, each containing three pairs. Half of the infants were tested on one of these sets whereas the others were tested on the second set. The left/right location of the intact hand on the first trial was counterbalanced across infants, and this location switched across trials in such a manner that the location of the intact or scrambled hand image was not repeated on more than two consecutive trials. Moreover, across infants, on any given trial the scrambled hand was presented equally often on the left as on the right. The same equipment utilized in Experiment 1 was used here. Coding of the infants’ performance was conducted in the same manner as in Experiment 1. A second coder verified the coding of 25% of the infants to document reliability. There was a Pearson correlation of .96 between coders.

Figure 2.

Examples of the scrambled and intact hand images used in Experiment 2. Infants in the upright condition were tested with upright images; infants in the inverted condition were tested with inverted images.

Results and Discussion

The dependent measure was the percent preference for the scrambled hands (see Table 2). This was calculated by dividing the total looking time for the scrambled hands across all six trials by the total looking time for both the intact and scrambled hands across all six trials; this ratio was then multiplied by 100. An outlier analysis (Tukey, 1977) revealed that the preference score of one infant in the upright condition was an outlier: this infant’s score was greater than 3 times the interquartile range below the lower quartile edge. Thus, this score was not included in the following analyses.

Table 2.

Mean look duration (seconds) at the intact and scrambled hand images, and percent preference for the scrambled stimulus in Experiment 2.

	N	Mean look duration to intact stimulus	Mean look duration to scrambled stimulus	Mean scrambled percent preference	t	95% confidence interval for scrambled percent preference	Effect size Cohen’s d
Upright	17	22.70	26.46	53.71	2.94*	51.03–56.39	0.70
Inverted	18	22.28	21.26	48.50	-0.70	44.00–53.00	0.17

Note. *p < .05, two-tailed; significantly different from the chance level of 50%.

One-sample t tests indicated that 3.5-month-olds’ percent preference for the scrambled hand (M = 53.71%, SE = 1.26; see Table 2) was significantly different from chance (50%) in the upright condition, t(16) = 2.94, p = .010, two-tailed, d = 0.7, but did not differ from chance in the inverted condition, (M = 48.50%, SE = 2.13), t(17) = -0.70, p = .492, two-tailed, d = 0.17 (see Table 2). Moreover, the percent preference score in the upright condition was significantly greater than the score in the inverted condition, t(33) = 2.07, p = .023, one-tailed, d = 0.71. Overall, 13 of the 17 infants in the upright condition had a mean preference for the scrambled hand that was above chance (binomial p = .025) whereas only 8 out of 18 infants in the inverted condition had a mean preference for the scrambled hand that was above chance (binomial p = .759).

Thus, like the adult participants in Morrisey and Rutherford (2013), infants exhibited an inversion effect by displaying a preference in the upright condition but not in the inverted condition. This inversion effect suggests that infants did not rely on low-level features to exhibit a preference in the upright condition. These results indicate that infants as young as 3.5 months old are sensitive to the typical first-order structure of hands without any training in the laboratory.

The results of both Experiments 1 and 2 are consistent in indicating that infants discriminate between intact and scrambled hand images when they are presented upright but not when they are inverted. However, it could be argued that the number of participants in each experiment (32 and 36 participants in Experiments 1 and 2, respectively) was not large enough to arrive at strong conclusions even though the effect size in each experiment was moderate to large (d around 0.6–0.7). Thus, we conducted a joint analysis of data from both experiments (resulting in 33 infants in the upright condition and 34 in the inverted condition), with a procedure (familiarization, spontaneous preference) X orientation (upright, inverted) ANOVA. This analysis revealed a significant main effect of orientation, F(1, 63) = 7.22, p = .009, η_P² = .103; neither the procedure main effect nor the interaction was significant. Thus, the combined results of both experiments once again indicated that performance was superior on upright images than on inverted images. These results clearly suggest that 3.5-month-olds are sensitive to the organization of hand images.

General Discussion

The current study found that infants are sensitive to the typical structure of hands early in life. In Experiment 1, 3.5-month-olds tested on a familiarization/novelty preference procedure discriminated between intact hand and scrambled hand images. In Experiment 2, they exhibited a spontaneous preference for scrambled over intact hand images. In both procedures, infants’ performance was disrupted by inversion of images. The inversion effects suggest that infants were not relying solely on low-level features to process first-order relational information in hands. These findings indicate that, as in the case of faces and bodies, structural information about hands is part of infants’ human representation quite early in life. Given that hands aid in the expression of emotions, goals, and desires (Morrisey & Rutherford, 2013), play a major role in joint attention (Daum et al., 2013; Deák, Flom, & Pick, 2000), and permit haptic exploration of the environment (Molina et al., 2015; Rochat, 1989), early sensitivity to hand information might facilitate the development of social and non-social cognition in infancy and adulthood.

The current findings are consistent with Meltzoff’s (2011) model of social cognition. As noted earlier, Meltzoff and Moore’s (1997) “active intermodal mapping” hypothesis and Meltzoff’s (2011) “like-me” model of social cognitive development assume that “organ identification” occurs in newborns. Meltzoff and Moore (1997) specifically identified the hand as an organ that even young infants would be able to identify (p.183; see also, Meltzoff, 2011). These researchers based their model on findings indicating that newborns treat body parts as discrete entities and their imitation is specific to organs. They assume that knowledge of body parts is either innate or rapidly acquired early in life. The current finding of early knowledge of the structure of hands by 3.5 months of age is thus consistent with Meltzoff’s model.

As noted earlier, there is evidence of specialized neural structures for the processing of hand information in monkeys and humans (Borra et al., 2017; Fox et al., 2016; Molenberghs et al., 2012; Taira et al., 1990). In particular, Meltzoff et al. (2018) report that touches to the hands of 7-month-olds are reflected in activity in specific locations of the somatosensory cortex. It is possible such brain structures that represent hands are innately functional or rapidly developing early in life and underlie young infants’ response to the configuration of hands.

Moreover, there is the potential for a large amount of information to be learned early on about the structure and function of hands because, starting early in life, infants are exposed to hands (Van der Meer, 1997; White et al., 1964) and use their hands to actively explore the environment (Rochat, 1989). Thus, infants may learn about the structure of hands within 3.5 months of age based on experience. In particular, infants’ proprioceptive experience of their own hands might contribute to the development of representation of hands in general. Alternatively, infants might come into the world already able to identify certain body parts, such as hands (Meltzoff and Moore, 1997). In either case, the hand may be a “special” organ like the face because of the significant role it plays in social communication and action.

The finding that infants as young as 3.5 months old are sensitive to the structure of hands adds to the existing literature on bodies and their social importance in infancy (Bhatt, Hock, White, Jubran, & Galati, 2016; Reed et al., 2003; Zieber et al., 2015). It is consistent with studies that indicate that 3.5-month-olds, and perhaps younger infants, are sensitive to the first-order organization of bodies (Gliga & Dehaene-Lambertz, 2005; Zieber et al., 2015) and of faces (Bhatt et al., 2005; Johnson et al., 1991). Some researchers (e.g., Bhatt et al., 2016; Simion & Di Giorgio, 2015) have suggested that a general cognition system engenders the development of knowledge about various sources of social information such as faces, bodies, and voices. This system might facilitate the early processing of critical information about the structure of critical human body parts, such as hands, that play a significant role in social communication.

It would be beneficial to examine the exact nature of 3.5-month-olds’ sensitivity to the scrambling of hand images. The discrimination between intact and scrambled images in the upright condition may be based on differences in the overall gestalt of the images (i.e., the overall shape) or based on specific representation of the location of particular parts. One way to address these possibilities might be to use eye-tracking technology to examine whether infants attend to particular regions of hand images that have been scrambled. If so, it would suggest that infants are sensitive to the particular location of individual parts.

It will also be important for future studies to examine any developmental changes in hand representation. We chose to study 3.5-month-olds in the current experiment because prior research indicates that infants this age are sensitive to structural information in bodies (Gliga & Dehaene-Lambertz, 2005; Zieber et al., 2015) and we wished to examine whether such knowledge extended to a key body part such as hands. However, the fact that 3.5-month-olds are sensitive to the structure of hands does not necessarily imply that their representation of hands is fully developed. As noted earlier, hands have a variety of functions and it is not clear that young infants are sensitive to all of the functions. Moreover, young infants may not be sensitive to other structural aspects of hands. For instance, hands are not only characterized by the relative location of parts but are also defined by the relative size of parts (e.g., the fingers are of a certain length compared to the size of the palm). It is possible that sensitivity to other such characteristics of hands develop later in life. It is thus important to examine infants of different ages in future research to obtain a more comprehensive understanding of hand representation and its development.

Future research should also examine whether young infants’ access to structural information extends to other body parts besides hands and faces. As noted earlier, Meltzoff and Moore (1997) suggested that young infants should be able to identify a variety of body parts, including arms, trunk, legs, and feet. If social stimuli are identified by their features as well as by the structural arrangements of these features (Reed et al., 2003; Slaughter & Heron, 2004; Zieber et al., 2015), then the Meltzoff and Moore (1997) proposal predicts that young infants would be sensitive to the structure of many other body parts in addition to hands. If, however, there is something special about hands, then young infants may not be sensitive to the structure of other body parts.

In summary, the findings of the current study demonstrate that infants’ representation of humans at 3.5 months of age includes the spatial arrangement of parts. Given the richness of the information provided by hands both in terms of social cues from others and haptic cues from objects, early representation of the structure of hands might set the stage for the development of efficient social and non-social information processing later in life.

Footnotes

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: this research was supported by grants from the National Science Foundation (BCS-1121096) and the National Institute of Child Health and Human Development (HD075829). The authors would like to thank the infants and parents who participated in this study.

References

Aslin

R. N.

(2009). How infants view natural scenes gathered from a head-mounted camera. Optometry and Vision Science, 86, 561–565.

Bhatt

R. S.

Bertin

Hayden

Reed

(2005). Face processing in infancy: Developmental changes in the use of different kinds of relational information. Child Development, 76, 169–181.

Bhatt

R. S.

Hock

White

Jubran

Galati

(2016). The development of body structure knowledge in infancy. Child Development Perspectives, 10, 45–52.

Borra

Gerbella

Rozzi

Luppino

(2017). The macaque lateral grasping network: A neural substrate for generating purposeful hand actions. Neuroscience & Biobehavioral Reviews, 75, 65–90.

Daum

M. M.

Ulber

Gredebäck

(2013). The development of pointing perception in infancy: Effects of communicative signals on covert shifts of attention. Developmental Psychology, 49, 1898–1908.

Deák

G. O.

Flom

R. A.

Pick

A. D.

(2000). Effects of gesture and target on 12- and 18-month-olds’ joint visual attention to objects in front of or behind them. Developmental Psychology, 36, 511–523.

Deák

G. O.

Krasno

A. M.

Triesch

Lewis

Sepeta

(2014). Watch the hands: Infants can learn to follow gaze by seeing adults manipulate objects. Developmental Science, 17, 270–281.

Diamond

Carey

(1986). Why faces are and are not special: An effect of expertise. Journal of Experimental Psychology: General, 115, 107–117.

Fox

N. A.

Bakermans-Kranenburg

M. J.

Yoo

K. H.

Bowman

L. C.

Cannon

E. N.

Vanderwert

R. E.

Ferrari

P. F.

van IJzendoorn

M. H.

(2016). Assessing human mirror activity with EEG mu rhythm: A meta-analysis. Psychological Bulletin, 142, 291–313.

10.

Fox

McDaniel

(1982). The perception of biological motion by human infants. Science, 218, 486–487.

11.

Frank

M. C.

Vul

Saxe

(2012). Measuring the development of social attention using free-viewing. Infancy, 17, 355–375.

12.

Gliga

Dehaene-Lambertz

(2005). Structural encoding of body and face in human infants and adults. Journal of Cognitive Neuroscience, 17, 1328–1340.

13.

Goren

C. C.

Sarty

P. Y. K.

(1975). Visual following and pattern discrimination in face-like stimuli by newborn infants. Pediatrics, 56, 544–549.

14.

Guajardo

J. J.

Woodward

A. L.

(2004). Is agency skin deep? Surface attributes influence infants’ sensitivity to goal-directed action. Infancy, 6, 361–384.

15.

Hock

White

Jubran

Bhatt

R. S.

(2016). The whole picture: Holistic body posture recognition in infancy. Psychonomic Bulletin & Review, 23, 426–431.

16.

Johnson

M. H.

Dziurawiec

Ellis

Morton

(1991). Newborns’ preferential tracking of face-like stimuli and its subsequent decline. Cognition, 40, 1–19.

17.

Longhi

Senna

Bolognini

Bulf

Tagliabue

Macchi Cassia

Turati

(2015). Discrimination of biomechanically possible and impossible hand movements at birth. Child Development, 86, 632–641.

18.

Marshall

P. J.

Meltzoff

A. N.

(2015). Body maps in the infant brain. Trends in Cognitive Sciences, 19, 499–505.

19.

Meltzoff

A. N.

(2011). Social cognition and the origins of imitation, empathy, and theory of mind. In Goswami

(Ed.), Childhood cognitive development (2nd ed., pp. 49–75). Malden, MA: Wiley Blackwell.

20.

Meltzoff

A. N.

Moore

M. K.

(1997). Explaining facial imitation: A theoretical model. Early Development & Parenting, 6, 179–192.

21.

Meltzoff

A. N.

Ramírez

R. R.

Saby

J. N.

Larson

Taulu

Marshall

P. J.

(2018). Infant brain responses to felt and observed touch of hands and feet: An MEG study. Developmental Science.

22.

Molenberghs

Cunnington

Mattingley

J. B.

(2012). Brain regions with mirror properties: A meta-analysis of 125 human fMRI studies. Neuroscience & Biobehavioral Reviews, 36, 341–349.

23.

Molina

Sann

David

Touré

Guillois

Jouen

(2015). Active touch in late-preterm and early-term neonates. Developmental Psychobiology, 57, 322–335.

24.

Morrisey

M. N.

Rutherford

M. D.

(2013). Do hands attract attention? Visual Cognition, 21, 647–672.

25.

Morton

Johnson

M. H.

(1991). CONSPEC and CONLERN: A two-process theory of infant face recognition. Psychological Review, 98, 164–181.

26.

Reed

C. L.

Stone

V. E.

Bozova

Tanaka

(2003). The body inversion effect. Psychological Science, 14, 302–308.

27.

Reed

C. L.

Stone

V. E.

Grubb

J. D.

McGoldrick

J. E.

(2006). Turning configural processing upside down: Part and whole-body postures. Journal of Experimental Psychology, 32, 73–87.

28.

Rochat

(1987). Mouthing and grasping in neonates: Evidence for the early detection of what hard or soft substances afford for action. Infant Behavior and Development, 10, 435–449.

29.

Rochat

(1989). Object manipulation and exploration in 2- to 5-month-old infants. Developmental Psychology, 25, 871–884.

30.

Simion

Di Giorgio

(2015). Face perception and processing in early infancy: Inborn predispositions and developmental changes. Frontiers in Psychology, 6, 1–11.

31.

Simion

Valenza

Macchi Cassia

Turati

Umiltà

(2002). Newborns’ preference for up–down asymmetrical configurations. Developmental Science, 5, 427–434.

32.

Slaughter

Heron

(2004). Origins and early development of human body knowledge. Monographs of the Society for Research in Child Development, 69, 1–102.

33.

Slaughter

Heron-Delaney

Christie

(2012). Developing expertise in human body perception. In Slaughter

Brownell

(Eds.), Early development of body representations (pp. 81–100). Cambridge, England: Cambridge University Press.

34.

Slaughter

Neary

(2011). Do young infants respond socially to human hands? Infant Behavior and Development, 34, 374–377.

35.

Taira

Mine

Georgopoulos

A. P.

Murata

Sakata

(1990). Parietal cortex neurons of the monkey related to the visual guidance of hand movement. Experimental Brain Research, 83, 29–36.

36.

Tukey

J. W.

(1977). Exploratory data analysis. Reading, MA: Addison-Wesley.

37.

Turati

Simion

Milani

Umiltà

(2002). Newborns’ preference for faces: What is crucial? Developmental Psychology, 38, 875–881.

38.

Van der Meer

A. L. H.

(1997). Keeping the arm in the limelight: Advanced visual control of arm movements in neonates. European Journal of Paediatric Neurology, 1, 103–108.

39.

White

B. L.

Castle

Held

(1964). Observations on the development of visually-directed reaching. Child Development, 35, 349–364.

40.

Yoshida

Smith

L. B.

(2008). What’s in view for toddlers? Using a head camera to study visual experience. Infancy, 13, 229–248.

41.

Smith

L. B.

(2013). Joint attention without gaze following: Human infants and their parents coordinate visual attention to objects through eye-hand coordination. PLoS One, 8, e79659.

42.

Zieber

Kangas

Hock

Bhatt

R. S.

(2015). Body structure perception in infancy. Infancy, 20, 1–17.