Budgerigars (Melopsittacus undulatus) copy virtual demonstrators in a two-action test

Juvenile budgerigars (Melopsittacus undulatus) observed thin film transistor video playback of a virtual conspecific demonstrator using its beak to remove a stopper from a food box, either by pulling the object upward or by pushing it downward. Simultaneously (Experiment 1) or subsequently (Experiment 2), the observers were offered a similar stopper box and rewarded with access to food for each removal response, regardless of its direction. Observers of upward movement made a greater proportion of up responses in total and showed a stronger tendency to increase the proportion of up responses over test trials than observers of downward movement. These findings provide the first demonstration of which the authors are aware that birds are able not only to detect and respond to a moving video image but also to copy its movements.     

Becoming a high‐fidelity – super – imitator: what are the contributions of social and individual learning?

In contrast to other primates, human children's imitation performance goes from low to high fidelity soon after infancy. Are such changes associated with the development of other forms of learning? We addressed this question by testing 215 children (26–59 months) on two social conditions (imitation, emulation) – involving a demonstration – and two asocial conditions (trial‐and‐error, recall) – involving individual learning – using two touchscreen tasks. The tasks required responding to either three different pictures in a specific picture order (Cognitive: Airplane→Ball→Cow) or three identical pictures in a specific spatial order (Motor‐Spatial: Up→Down→Right). There were age‐related improvements across all conditions and imitation, emulation and recall performance were significantly better than trial‐and‐error learning. Generalized linear models demonstrated that motor‐spatial imitation fidelity was associated with age and motor‐spatial emulation performance, but cognitive imitation fidelity was only associated with age. While this study provides evidence for multiple imitation mechanisms, the development of one of those mechanisms – motor‐spatial imitation – may be bootstrapped by the development of another social learning skill – motor‐spatial emulation. Together, these findings provide important clues about the development of imitation, which is arguably a distinctive feature of the human species.     

Defining Elemental Imitation Mechanisms: A Comparison of Cognitive and Motor-Spatial Imitation Learning Across Object- and Computer-Based Tasks

During the first 5 years of life, the versatility, breadth, and fidelity with which children imitate change dramatically. Currently, there is no model to explain what underlies such significant changes. To that end, the present study examined whether task-independent but domain-specific—elemental—imitation mechanism explains performance across imitation tasks or domains. Preschool-age children (n = 156) were tested on 4 imitation tasks, 2 object-based (animal, puzzle box) and 2 computer-based (cognitive, motor-spatial). All tasks involved 3 serial actions. The animal task involved making an animal face, and the puzzle box task involved manipulating a box to retrieve a reward. The cognitive task involved responding to 3 different pictures in a specific picture order, and the motor-spatial task involved responding to 3 identical pictures in a specific spatial order. A principal component analysis including performance on all 4 tasks produced 2 components: “cognitive imitation” (cognitive and animal tasks) and “motor-spatial imitation” (motor-spatial and puzzle box tasks). Regression analyses replicated these results. These findings provide preliminary support for the hypothesis that underlying performance across these different tasks involves multiple—elemental—imitation mechanisms for learning and copying domain-specific information across tasks.     

Automatic imitation and spatial compatibility in a key-pressing task

Automatic imitation has been often confounded with spatial compatibility effects. Heyes (2011) called attention to this confound, and proposed some criteria which must be satisfied before these effects could be unequivocally taken to be an index of the functioning of the human mirror system. Evidence satisfying such criteria has been reported by Catmur and Heyes (2011), using a relatively unfamiliar finger abduction movement. However, because many previous studies relied on more familiar actions, we aimed at testing whether analogous effects could be obtained with a more practiced key-pressing task. In Experiment 1, we used anatomical controls (i.e., views of right vs. left hands) under conditions affording mirror imitation, and showed that spatial compatibility masked the effects of automatic imitation. Experiment 2 used rotated conditions to control for this spatial-anatomical confound, and it showed unequivocal effects of automatic imitation, which were obtained regardless of its relation to the spatial stimulus-response mapping. These results cast some doubts on the interpretation of previous reports relying exclusively on scenes presented from a mirror perspective, and suggest the convenience of using both rotated scenes and anatomical controls in order to assess automatic imitation.     

Bio-inspired cognitive model of motor learning by imitation

Learning, and even more so by imitation, is an essential Cognitive Functions because it is carried out throughout life and allows us to adapt our behaviors from other beings through observation. In this work, we propose a model, and implementation of the cognitive function of imitation motor learning (IML), based on psychological and neuroscientific evidence. According to the evidence, learning by imitation includes imitation of action and imitation of action over an object sub-processes. The imitation of action consists of the movement of the limbs. The imitation of action over an object consists of the interaction with an object within the environment. We achieve an implementation of the proposed model for IML and endow a virtual entity with it. In order to validate the proposal, we use a case study to analyze the sub-processes performance. From results, we conclude that both imitation of action and imitation of action over an object sub-processes play an essential role in getting the agent to interact with stimuli within the environment.     

Imitative behavior

This article reviews the results of experimental studies on imitative behavior reported by various investigators, and then discusses the possible brain mechanisms responsible for this behavior. It was found that human infants in their first hours of life were already capable of spontaneous imitation of simple motor acts demonstrated by an adult, without previous training or reward; these observations suggest that imitative behavior is an innate process that can be considered anunconditional reflex of imitation. It was also found that satiated animals resumed eating when they saw their companions eating. In the latter case, the imitative reflex triggered the previously acquired feeding behavior. Similar mechanisms could be responsible for the phenomenon of eating more in the presence of companions than in their absence, as well as that of preferring the food chosen by companions. When followed by a reward, the imitative act can be learned—that is, transformed into aninstrumental conditional response; learning by imitation of simple motor acts was observed in animals, and that of complex motor acts was observed in children who had already achieved a certain developmental stage. In animals, learning complex motor tasks was facilitated by previous observation of a companion performing this task. In this case, the presence of the observer during the session could lead to habituation of the experimental situation and production of associations between this situation and stimuli or emotions related to the reward or punishment, and might result in more efficient learning later. The imitative behavior can be inhibited by stimuli producing responses antagonistic to the act of imitation.     

Are Automatic Imitation and Spatial Compatibility Mediated by Different Processes?

Automatic imitation or “imitative compatibility” is thought to be mediated by the mirror neuron system and to be a laboratory model of the motor mimicry that occurs spontaneously in naturalistic social interaction. Imitative compatibility and spatial compatibility effects are known to depend on different stimulus dimensions—body movement topography and relative spatial position. However, it is not yet clear whether these two types of stimulus–response compatibility effect are mediated by the same or different cognitive processes. We present an interactive activation model of imitative and spatial compatibility, based on a dual‐route architecture, which substantiates the view they are mediated by processes of the same kind. The model, which is in many ways a standard application of the interactive activation approach, simulates all key results of a recent study by Catmur and Heyes (2011) . Specifically, it captures the difference in the relative size of imitative and spatial compatibility effects; the lack of interaction when the imperative and irrelevant stimuli are presented simultaneously; the relative speed of responses in a quintile analysis when the imperative and irrelevant stimuli are presented simultaneously; and the different time courses of the compatibility effects when the imperative and irrelevant stimuli are presented asynchronously.     

Is imitation learning the route to humanoid robots?

This review investigates two recent developments in artificial intelligence and neural computation: learning from imitation and the development of humanoid robots. It is postulated that the study of imitation learning offers a promising route to gain new insights into mechanisms of perceptual motor control that could ultimately lead to the creation of autonomous humanoid robots. Imitation learning focuses on three important issues: efficient motor learning, the connection between action and perception, and modular motor control in the form of movement primitives. It is reviewed here how research on representations of, and functional connections between, action and perception have contributed to our understanding of motor acts of other beings. The recent discovery that some areas in the primate brain are active during both movement perception and execution has provided a hypothetical neural basis of imitation. Computational approaches to imitation learning are also described, initially from the perspective of traditional AI and robotics, but also from the perspective of neural network models and statistical-learning research. Parallels and differences between biological and computational approaches to imitation are highlighted and an overview of current projects that actually employ imitation learning for humanoid robots is given.     

Role of Constant,Random and Blocked Practice in an Electromyography-Based Oral Motor Learning Task

Purpose: The role of principles of motor learning (PMLs) in speech has received much attention in the past decade. Oral motor learning, however, has not received similar consideration. This study evaluated the role of three practice conditions in an oral motor tracking task.

Method: Forty-five healthy adult participants were randomly and equally assigned to one of three practice conditions (constant, blocked, and random) and participated in an electromyography-based task. The study consisted of four sessions, at one session a day for four consecutive days. The first three days sessions included a practice phase, with immediate visual feedback, and an immediate retention phase, without visual feedback. The fourth session did not include practice, but only delayed retention testing, lasting 10–15 minutes, without visual feedback.

Results: Random group participants performed better than participants in constant and blocked practice conditions on all the four days. Constant group participants demonstrated superior learning over blocked group participants only on day 4.

Conclusion: Findings indicate that random practice facilitates oral motor learning, which is in line with limb/speech motor learning literature. Future research should systematically investigate the outcomes of random practice as a function of different oral and speech-based tasks.     

Learning associations between action and perception: effects of incompatible training on body part and spatial priming

Observation of another person executing an action primes the same action in the observer's motor system. Recent evidence has shown that these priming effects are flexible, where training of new associations, such as making a foot response when viewing a moving hand, can reduce standard action priming effects (Gillmeister, Catmur, Liepelt, Brass, & Heyes, 2008). Previously, these effects were obtained after explicit learning tasks in which the trained action was cued by the content of a visual stimulus. Here we report similar learning processes in an implicit task in which the participant's action is self-selected, and subsequent visual effects are determined by the nature of that action. Importantly, we show that these learning processes are specific to associations between actions and viewed body parts, in that incompatible spatial training did not influence body part or spatial priming effects. Our results are consistent with models of visuomotor learning that place particular emphasis on the repeated experience of watching oneself perform an action.     

Seeing How It Sounds: Observation,Imitation, and Improved Learning in Piano Playing

This study centers upon a piano learning and teaching environment in which beginners and intermediate piano students (N = 48) learning to perform a specific type of staccato were submitted to three different (group-exclusive) teaching conditions: audio-only demonstration of the musical task; observation of the teacher's action demonstration followed by student imitation (blocked observation); and observation of the teacher's action demonstration while alternating imitation of the task with the teacher's performance (interleaved observation). Students submitted to interleaved observation were more proficient at the learned task with no significant differences for students of different proficiency levels.     

THE DEVELOPMENT AND GENERALIZATION OF DELAYED IMITATION1

An individual's behavior can be identified as imitative if it temporally follows the behavior of another individual and if its topography is controlled by the demonstrated behavior [Baer, Peterson, and Sherman (Journal of the Experimental Analysis of Behavior, 1967, 10 , 405–416)]. This definition takes into account both temporal and topographical characteristics of the behavior in question. More recent research in the area of imitation has interpreted the temporal component of the above definition differentially by limiting imitation to those topographically similar responses occurring within 3, 5, or 10 sec after a model's demonstration. Yet, Gewirtz and Stingle (Psychological Review, 1968, 75 , 375–397) pointed out that much of the imitation seen in young children is not of this immediate nature, but instead occurs sometime after a model's response. They further suggest that this type of imitative behavior can be characterized as a response class and is susceptible to development and modification as a function of consequences delivered to subjects contingent on this type of delayed responding. Four retarded children, three initially imitative and one nonimitative, were individually trained to imitate a number of motor responses in an immediate and a delayed fashion. Immediate imitation was defined as a response similar to a model's demonstration occurring within 5 sec after the model's demonstration; delayed imitation was defined as a response similar to a model's demonstration occurring more than 5 sec, but not more than 25 sec, after the model's demonstration. A reversal (ABAB) design was employed to examine the experimental development of a generalized delayed imitative repertoire. Untrained probe responses were demonstrated to subjects systematically through the ongoing training. Generalized immediate and delayed imitation were observed in each subject; this generalization was restricted to the type of imitation currently undergoing training. This development of a generalized imitation repertoire was observed in each subject. That is, these subjects imitated some responses that had never been specifically trained. More importantly, a training package consisting of prompting, fading, and consequences for delayed imitation functioned to develop generalized delayed imitation. These data exemplify a special case of generalization that was a function of the most recent training history of immediate or delayed imitation. The reversal design demonstrated that imitations of nontrained models were either delayed or immediate, depending upon which form of imitation was currently receiving training. Therefore, for each form of imitation trained, delayed or immediate, a corresponding response class was demonstrated. These data relate to data reported by Garcia, Baer, and Firestone (Journal of Applied Behavior Analysis, 1971, 4 , 101–112). The association lies in the proposition that there are identifiable boundaries of generalized imitation and that these boundaries are functionally related to previous training histories.     

Imitation as a Learning Strategy during Sibling Teaching

ABSTRACT

The sibling relationship represents a unique bond characterized by a high degree of closeness and intimacy, which fosters teaching and learning. Two studies investigated associations between sibling-directed teaching, imitation as a learning strategy, and learner involvement during a semi-structured, video-taped construction task. Study 1 also examined associations with the teacher’s Theory of Mind (ToM) abilities; Study 2 focused on associations with birth order and sibling relationship quality. In both studies, siblings ranged from preschool-age to the cusp of middle childhood (Study 1 n = 61; Study 2 n = 72). Findings across both studies indicated that learners engaged in significantly more nonverbal than verbal imitation and imitation was predominately immediate and not deferred. Teachers responded to both verbal and nonverbal learner imitation positively, corrected, or did not respond. In Study 1, learner task involvement was predicted by learner age and nonverbal imitation, while teaching strategies that were positively related to the teacher’s ToM abilities were associated with learner imitation. In Study 2, younger siblings’ reports of a positive sibling relationship were significantly associated with learner imitation. Birth order differences were only evident for younger (but not older) sibling learner imitation and task involvement. Findings are discussed in light of relationships and social constructivist theories of development.     

Cognitive mechanisms of visuomotor transformation in movement imitation: Examining predictions based on models of apraxia and motor control

When we observe a movement and then reproduce it, how is this visual input transformed into motor output? Studies on stroke patients with apraxia suggest that there may be two distinct routes used for gesture imitation; an indirect route that recruits stored movement memories (motor programs) and a direct route that bypasses them. The present study examined 30 healthy adults ages 18–80 (mean age = 44.0 years, SD = 19.5) to learn how motor programs are recruited or bypassed in movement imitation depending upon task conditions (whether familiar letters or novel shapes are imitated) and perceptual factors (whether shapes or letters are perceived). Subjects were asked to imitate the movements of a model who formed shapes and letters on a sheer mesh screen, and to report whether they perceived the task as a shape or a letter. Movements were recorded using a Vicon motion analysis system, and subsequently analyzed to determine the degree of difference between the demonstrated and produced movements. As predicted, letter perception on the letter tasks resulted in increased temporal error when the demonstrated stroke order conflicted with subjects’ habitual pattern of letter formation. No such interference effects were observed when the letter tasks were perceived as shapes. These findings are discussed in the context of current theories on imitation, and implications for rehabilitation and motor re-learning are presented.     

Index to Volume 163

This study investigated the effects of the motor and verbal aspects of modeling on imitation. The subjects were 2- and 3-year-old children (N = 96). The child's imitation responses were recorded during the play period that followed each modeled act. Each child observed the model in one of four modeling conditions. In Condition 1, the model “flew” a telephone while saying that he or she was flying an airplane. Imitation was recorded as motor if the child flew the telephone but was recorded as verbal and realistic if the child flew an airplane. In Condition 2, the model flew an airplane while saying that he or she was flying a telephone. Imitation was recorded as verbal if the child flew the telephone. In Condition 3, the model flew an airplane and said that he or she was flying an airplane. If the child flew an airplane, imitation was scored as motor, verbal, and. realistic. In Condition 4, the model flew a telephone and said that he or she was flying a telephone. Imitation was scored as motor and verbal if the child flew the telephone but was scored as realistic if the child flew the airplane. In Condition 1, 2-year-olds displayed more motor imitation than 3-year-olds, and 3-year-olds displayed more verbal-reality imitation than 2-year-olds. Boys displayed more motor imitation than girls. There were no age or sex differences in Condition 2. In Condition 3, 2-year-olds imitated more than 3-year-olds, with 3-year-old girls imitating the least. In Condition 4, reality imitation was largely due to 2-year-old boys' imitation of masculine-type acts.     

Programmed to learn? The ontogeny of mirror neurons

Mirror neurons are increasingly recognized as a crucial substrate for many developmental processes, including imitation and social learning. Although there has been considerable progress in describing their function and localization in the primate and adult human brain, we still know little about their ontogeny. The idea that mirror neurons result from Hebbian learning while the child observes/hears his/her own actions has received remarkable empirical support in recent years. Here we add a new element to this proposal, by suggesting that the infant's perceptual‐motor system is optimized to provide the brain with the correct input for Hebbian learning, thus facilitating the association between the perception of actions and their corresponding motor programs. We review evidence that infants (1) have a marked visual preference for hands, (2) show cyclic movement patterns with a frequency that could be in the optimal range for enhanced Hebbian learning, and (3) show synchronized theta EEG (also known to favour synaptic Hebbian learning) in mirror cortical areas during self‐observation of grasping. These conditions, taken together, would allow mirror neurons for manual actions to develop quickly and reliably through experiential canalization. Our hypothesis provides a plausible pathway for the emergence of mirror neurons that integrates learning with genetic pre‐programming, suggesting new avenues for research on the link between synaptic processes and behaviour in ontogeny.     

Sequence learning in pianists and nonpianists: An fMRI study of motor expertise

Previous studies of motor learning have proposed a distinction between fast and slow learning, but these mechanisms have rarely been examined simultaneously. We examined the influence of longterm motor expertise (slow learning) while pianists and nonpianists performed alternating epochs of sequenced and random keypresses in response to visual cues (fast learning) during functional neuroimaging. All of the participants demonstrated learning of the sequence as demonstrated by decreasing reaction times (RTs) on sequence trials relative to random trials throughout the session. Pianists also demonstrated faster RTs and superior sequence acquisition in comparison with nonpianists. Withinsession decreases in bilateral sensorimotor and parietal activation were observed for both groups. Additionally, there was more extensive activation throughout the session for pianists in comparison with nonpianists across a network of primarily right-lateralized prefrontal, sensorimotor, and parietal regions. These findings provide evidence that different neural systems subserve slow and fast phases of learning.     

Dissociable perceptual-learning mechanisms revealed by diffusion-model analysis

Performance on perceptual tasks improves with practice. Most theories address only accuracy data and tacitly assume that perceptual learning is a monolithic phenomenon. The present study pioneers the use of response time distributions in perceptual learning research. The 27 observers practiced a visual motion-direction discrimination task with filtered-noise textures for four sessions with feedback. Session 5 tested whether the learning effects transferred to the orthogonal direction. The diffusion model (Ratcliff, Psychological Review, 85, 59–108, 1978) achieved good fits to the individual response time distributions from each session and identified two distinct learning mechanisms with markedly different specificities. A stimulus-specific increase in the drift-rate parameter indicated improved sensory input to the decision process, and a stimulus-general decrease in nondecision time variability suggested improved timing of the decision process onset relative to stimulus onset (which was preceded by a beep). A traditional d’ analysis would miss the latter effect, but the diffusion-model analysis identified it in the response time data.     

A one-trial inhibitory avoidance task to zebrafish: Rapid acquisition of an NMDA-dependent long-term memory

The behavioral tasks aiming to evaluate learning and memory mechanisms currently available to zebrafish (Danio rerio) involve long training sessions frequently along multiple days and are based on shuttle box or active-avoidance protocols, preventing a detailed analysis of cellular and molecular time-dependent processes involved in memory acquisition and consolidation. In order to explore zebrafish’s potential contribution to the characterization of the molecular machinery underlying learning and memory rapidly acquired and reliable paradigms are necessary. In this study we present a rapid and effective learning protocol in a single-trial inhibitory avoidance in zebrafish. In a simple apparatus, adult animals learned to refrain from swimming from a white into a dark compartment in order to avoid an electric shock during a single-trial training session that lasted less than 2 min. The resulting memory is robust, long-lasting and sensitive to NMDA-receptor antagonist MK-801 given in the tank water immediately after training. Experiments aiming to further characterize the events underlying memory formation, retrieval or extinction or those looking for cognitive profiling of mutants, neurotoxicological studies and disease models may benefit from this task, and together with complementary strategies available for zebrafish may significantly improve our current knowledge on learning and memory mechanisms.     

Understanding the flexibility of action–perception coupling

The idea that observing an action triggers an automatic and obligatory activation of an imitative action in the motor system of the observer has recently been questioned by studies examining complementary actions. Instead of a tendency for imitation, cooperative settings may facilitate the execution of dissimilar actions, resulting in a relative disadvantage for imitative actions. The present study aimed at clarifying the contribution of associative learning and interference of task representations to the reversal of congruency effects. To distinguish between the two, an experiment was designed, in which we increased the effects of associative learning and minimized the effects of task interference. Participants completed a series of imitation and complementary action runs, in which they continuously imitated or complemented the actions of a virtual co-actor. Each run was alternated with a test run showing the same actions but including color-cues, and the participants were instructed to respond to color instead of the actor’s posture. Reaction times to test runs showed no reversal of facilitation effects between the imitation and complementary action conditions. This result strongly argues that associative learning cannot adequately account for reversed facilitation effects. Our study provides additional support for action–perception models that allow flexible selection of action–perception coupling and challenges the existing models purely based on stimulus–response associations.

---