Portal de Periódicos da CAPES

From Observed Action Identity to Social Affordances

2021; Elsevier BV; Volume: 25; Issue: 6 Linguagem: Inglês

10.1016/j.tics.2021.02.012

1879-307X

Guy A. Orban, Marco Lanzilotto, Luca Bonini,

Psychology of Moral and Emotional Judgment

A substantial fraction of neurons in the monkey anterior intraparietal area (AIP) and its human homologue phAIP are selective for observed manipulative actions (OMAs).OMA selective neurons encode the identity of the observed actions, up to the level of semantic representation in phAIP.OMA identity may result from the combination of two visual signals originating in the superior temporal sulcus (STS) and concerning: (i) observed body movements: and (ii) the changes in the hand/object relationship (action effects).Others' observed actions, beyond grasping, may be specified in parietal territories, underpinning 'social affordance' processing and the selection of potential behavioral responses in parieto-premotor circuits. Others' observed actions cause continuously changing retinal images, making it challenging to build neural representations of action identity. The monkey anterior intraparietal area (AIP) and its putative human homologue (phAIP) host neurons selective for observed manipulative actions (OMAs). The neuronal activity of both AIP and phAIP allows a stable readout of OMA identity across visual formats, but human neurons exhibit greater invariance and generalize from observed actions to action verbs. These properties stem from the convergence in AIP of superior temporal signals concerning: (i) observed body movements; and (ii) the changes in the body–object relationship. We propose that evolutionarily preserved mechanisms underlie the specification of observed-actions identity and the selection of motor responses afforded by them, thereby promoting social behavior. Others' observed actions cause continuously changing retinal images, making it challenging to build neural representations of action identity. The monkey anterior intraparietal area (AIP) and its putative human homologue (phAIP) host neurons selective for observed manipulative actions (OMAs). The neuronal activity of both AIP and phAIP allows a stable readout of OMA identity across visual formats, but human neurons exhibit greater invariance and generalize from observed actions to action verbs. These properties stem from the convergence in AIP of superior temporal signals concerning: (i) observed body movements; and (ii) the changes in the body–object relationship. We propose that evolutionarily preserved mechanisms underlie the specification of observed-actions identity and the selection of motor responses afforded by them, thereby promoting social behavior. Manual skills are a hallmark of primates, particularly humans. They have made possible most of our transformational impact on the world, which was driven by an evolutionarily preserved but expanding network of cortical areas in the primate lineage that subserves the neural control of manipulative actions [1.Padberg J. et al.Parallel evolution of cortical areas involved in skilled hand use.J. Neurosci. 2007; 27: 10106-10115Crossref PubMed Scopus (111) Google Scholar, 2.Kaas J.H. Stepniewska I. Evolution of posterior parietal cortex and parietal-frontal networks for specific actions in primates.J. Comp. Neurol. 2016; 524: 595-608Crossref PubMed Scopus (53) Google Scholar, 3.Borra E. et al.The macaque lateral grasping network: a neural substrate for generating purposeful hand actions.Neurosci. Biobehav. Rev. 2017; 75: 65-90Crossref PubMed Scopus (54) Google Scholar, 4.Goldring A.B. Krubitzer L.A. Chapter 26 - Evolution of parietal cortex in mammals: from manipulation to tool use.in: Kaas J.H. Evolutionary Neuroscience. Second Edition. Academic Press, 2020: 627-656Crossref Google Scholar]. Interestingly, an equally well-articulated neural machinery is required to resolve the visual complexity of observed manipulative actions (OMAs) (see Glossary) performed by other individuals, because this ability is of critical importance for action planning during social interaction and interindividual coordination [5.Lanzilotto M. et al.Neuronal encoding of self and others' head rotation in the macaque dorsal prefrontal cortex.Sci. Rep. 2017; 7: 8571Crossref PubMed Scopus (7) Google Scholar, 6.Sacheli L.M. et al.How task interactivity shapes action observation.Cereb. Cortex. 2019; 29: 5302-5314Crossref PubMed Scopus (5) Google Scholar, 7.Ninomiya T. et al.A causal role for frontal cortico–cortical coordination in social action monitoring.Nat. Commun. 2020; 11: 5233Crossref PubMed Scopus (1) Google Scholar]. Indeed, as compared with other complex static visual stimuli, such as objects [8.Bao P. et al.A map of object space in primate inferotemporal cortex.Nature. 2020; 583: 103-108Crossref PubMed Scopus (19) Google Scholar], faces [9.Chang L. Tsao D.Y. The code for facial identity in the primate brain.Cell. 2017; 169: 1013-1028.e14Abstract Full Text Full Text PDF PubMed Scopus (190) Google Scholar,10.Freiwald W.A. The neural mechanisms of face processing: cells, areas, networks, and models.Curr. Opin. Neurobiol. 2020; 60: 184-191Crossref PubMed Scopus (4) Google Scholar], others' gaze direction [11.Shepherd S.V. et al.Mirroring of attention by neurons in macaque parietal cortex.Proc. Natl. Acad. Sci. U. S. A. 2009; 106: 9489-9494Crossref PubMed Scopus (93) Google Scholar], and body posture [12.Kumar S. et al.Transformation of visual representations across ventral stream body-selective patches.Cereb. Cortex. 2019; 29: 215-229Crossref PubMed Scopus (8) Google Scholar], observed actions of others are inherently dynamic stimuli, and their dynamics are essential for an observer's brain to compute their identity, despite the rapid changes in their retinal image. This is probably the reason why James Gibson claimed that 'animals are by far the most complex objects of perception that the environment presents to an observer' [13.Gibson J.J. Ecological Approach to Visual Perception. Houghton Mifflin, 1979Google Scholar]. Body movements are a fundamental component of an 'action'; nonetheless, they represent only one such component. In fact, an action is much more than a set of coordinated body movements, since it aims to produce a change in the environment in which the subject is immersed [14.Bonini L. et al.Neurophysiological bases underlying the organization of intentional actions and the understanding of others' intention.Conscious. Cogn. 2013; 22: 1095-1104Crossref PubMed Scopus (24) Google Scholar]. Thus, the changes an agent's action causes in the target object constitute an element that is almost as important as the body movement itself, because its dynamics make the OMA predictable in terms of its motor goal [15.Oram M.W. Perrett D.I. Integration of form and motion in the anterior superior temporal polysensory area (STPa) of the macaque monkey.J. Neurophysiol. 1996; 76: 109-129Crossref PubMed Scopus (283) Google Scholar, 16.Kilner J.M. et al.Motor activation prior to observation of a predicted movement.Nat. Neurosci. 2004; 7: 1299-1301Crossref PubMed Scopus (283) Google Scholar, 17.Maranesi M. et al.Mirror neuron activation prior to action observation in a predictable context.J. Neurosci. 2014; 34: 14827-14832Crossref PubMed Scopus (43) Google Scholar]. These two types of signal, specifying: (i) how the dynamics of body movement unfold; and (ii) how it will change the position or shape of an object, naturally coexist in everyday manipulative actions, and characterize the action identity. Both elements are crucial. For example, the same grasping act performed on a branch may serve to secure the body while climbing, to manipulate it for grabbing fruits, or to use it to hit something or someone else: in spite of the body-movement similarity, these clearly constitute different actions with different consequences. Similarly, the same effect of moving an object away from the body can be achieved by pushing it, throwing it, or kicking it, which clearly constitute different actions despite the similar consequence they produce in the outside world. Here, we first review evidence of neuronal signatures of OMA-identity coding in the primate brain, which point to area AIP as a critical node for this function. We then elucidate the connectional architecture that enables the convergence and integration in AIP of the two main sources of information needed to encode OMA identity: body movements and hand–object-interaction signals (i.e., attainment of the motor goal). Finally, we propose an extension of this model to a larger variety of action classes beyond the manipulative ones and of parietal areas in addition to AIP, which should drive future studies on the neural mechanisms underlying the computation of action identity in the non-human and human primate brain. Area AIP has long been considered a crucial node of the cortical motor system because of its role in routing visual information regarding 3D objects [18.Murata A. et al.Selectivity for the shape, size, and orientation of objects for grasping in neurons of monkey parietal area AIP.J. Neurophysiol. 2000; 83: 2580-2601Crossref PubMed Google Scholar, 19.Verhoef B.-E. et al.Effects of microstimulation in the anterior intraparietal area during three-dimensional shape Categorization.PLoS One. 2015; 10e0136543Crossref PubMed Scopus (11) Google Scholar, 20.Schaffelhofer S. Scherberger H. Object vision to hand action in macaque parietal, premotor, and motor cortices.eLife. 2016; 5e15278Crossref PubMed Scopus (47) Google Scholar] and observed actions [21.Pani P. et al.Grasping execution and grasping observation activity of single neurons in the macaque anterior intraparietal area.J. Cogn. Neurosci. 2014; 26: 2342-2355Crossref PubMed Scopus (45) Google Scholar,22.Maeda K. et al.Functional properties of parietal hand manipulation-related neurons and mirror neurons responding to vision of own hand action.J. Cogn. Neurosci. 2015; 27: 560-572Crossref PubMed Scopus (40) Google Scholar] from temporal [23.Perrett D.I. et al.Frameworks of analysis for the neural representation of animate objects and actions.J. Exp. Biol. 1989; 146: 87-113Crossref PubMed Google Scholar,24.Singer J.M. Sheinberg D.L. Temporal cortex neurons encode articulated actions as slow sequences of integrated poses.J. Neurosci. 2010; 30: 3133-3145Crossref PubMed Scopus (0) Google Scholar] and other parietal [25.Gamberini M. et al.Sensory properties of the caudal aspect of the macaque's superior parietal lobule.Brain Struct. Funct. 2018; 223: 1863-1879PubMed Google Scholar,26.Breveglieri R. et al.Neurons modulated by action execution and observation in the macaque medial parietal cortex.Curr. Biol. 2019; 29: 1218-1225.e3Abstract Full Text Full Text PDF PubMed Scopus (6) Google Scholar] regions to the premotor cortex [27.Maranesi M. et al.Cortical processing of object affordances for self and others' action.Front. Psychol. 2014; 5: 538Crossref PubMed Scopus (24) Google Scholar], whereas neighboring inferior parietal convexity areas were deemed to play a more important role in the processing of other's observed actions [28.Rozzi S. et al.Functional organization of inferior parietal lobule convexity in the macaque monkey: electrophysiological characterization of motor, sensory, and mirror responses and their correlation with cytoarchitectonic areas.Eur. J. Neurosci. 2008; 28: 1569-1588Crossref PubMed Scopus (0) Google Scholar,29.Rizzolatti G. et al.Cortical mechanisms underlying the organization of goal-directed actions and mirror neuron-based action understanding.Physiol. Rev. 2014; 94: 655-706Crossref PubMed Scopus (239) Google Scholar]. Extant studies have focused almost exclusively on the neural coding of graspable objects and grasping actions, with the exception of recent investigations that have recorded AIP neuronal activity while monkeys observed a larger set of OMA exemplars [30.Lanzilotto M. et al.Anterior intraparietal area: a hub in the observed manipulative action network.Cereb. Cortex. 2019; 29: 1816-1833Crossref PubMed Scopus (16) Google Scholar,31.Lanzilotto M. et al.Stable readout of observed actions from format-dependent activity of monkey's anterior intraparietal neurons.Proc. Natl. Acad. Sci. U. S. A. 2020; 117: 16596-16605Crossref PubMed Scopus (3) Google Scholar]. The findings of these latter studies demonstrate a crucial role of area AIP in routing visual information about OMAs to the other nodes of the cortical action observation network. What are the mechanisms through which the brain can achieve a stable readout of the identity of others' manipulative actions? In a recent study [31.Lanzilotto M. et al.Stable readout of observed actions from format-dependent activity of monkey's anterior intraparietal neurons.Proc. Natl. Acad. Sci. U. S. A. 2020; 117: 16596-16605Crossref PubMed Scopus (3) Google Scholar], AIP neurons displayed a marked selectivity for OMAs performed by another monkey (i.e., grasping and grooming) among a variety of stimuli, including emotional facial gestures (i.e., lip smacking and screaming), neutral facial gestures (i.e., yawning and chewing), and other dynamic stimuli (i.e., still monkey, a moving animal, and a landscape) presented on a screen. In that study, AIP neurons were also tested with a large set of OMA exemplars (i.e., dragging, dropping, grasping, pulling, pushing, rotating, and squeezing) previously used to reveal action-identity coding in monkey AIP [30.Lanzilotto M. et al.Anterior intraparietal area: a hub in the observed manipulative action network.Cereb. Cortex. 2019; 29: 1816-1833Crossref PubMed Scopus (16) Google Scholar]. In addition, OMA exemplars were presented in four visual formats, resulting from the combination of two body postures of an actor (standing and sitting) and two viewpoints (lateral and frontal) (Figure 1A ). The results showed that 38% of AIP neurons showed selectivity for OMAs in at least one format, with distinct sets of neurons exhibiting a preference for a specific exemplar (or set of exemplars), in addition to tuning for the visual presentation format (see example neuron in Figure 1A). However, no neuron exhibiting fully visual-invariant OMA selectivity was found. In fact, information about visual format and action identity was dynamically integrated according to a multiplicative mixing model [31.Lanzilotto M. et al.Stable readout of observed actions from format-dependent activity of monkey's anterior intraparietal neurons.Proc. Natl. Acad. Sci. U. S. A. 2020; 117: 16596-16605Crossref PubMed Scopus (3) Google Scholar], as previously described for static images in the inferior temporal cortex [32.Ratan Murty N.A. Arun S.P. Multiplicative mixing of object identity and image attributes in single inferior temporal neurons.Proc. Natl. Acad. Sci. U. S. A. 2018; 115: E3276-E3285Crossref PubMed Scopus (3) Google Scholar]. Such a multiplicative mixing of visual information enables the decoding of an early signal about the viewpoint (50 ms after stimulus onset) and the actor's body posture (at 100 ms) and, slightly later (150 ms), even the decoding of OMA identity in a format-independent manner. Crucially, the accuracy with which OMA identity is decoded depends upon the presence of a subset of units that maintain a relatively stable OMA selectivity across formats despite considerable rescaling of their firing rate according to the visual specificities of each format (as in the example neuron of Figure 1A). What is the relationship, if one exists, between neuronal representations of individual OMA exemplars in AIP? The clustering of individual exemplars in the neural space [31.Lanzilotto M. et al.Stable readout of observed actions from format-dependent activity of monkey's anterior intraparietal neurons.Proc. Natl. Acad. Sci. U. S. A. 2020; 117: 16596-16605Crossref PubMed Scopus (3) Google Scholar] indicated that actions characterized by the movement of the hand toward a target lying on a table (e.g., grasping or dragging) were more closely linked and, consequently, segregated from those in which the hand was already in contact with the manipulated object (e.g., rolling or squeezing, Figure 1B). This clustering of action exemplars was largely independent of the variety of combinations of viewpoints and body postures (Figure 1C), suggesting that the dynamic relationships between the actor's hand and the target object, which are relatively stable across formats, make a fundamental contribution to the neural representation of OMA identity. Interestingly, the same stimuli used to study monkey AIP neurons (Figure 1A) were recently presented to two human patients participating in a brain–machine interface clinical trial, allowing the researchers to record single-neuron activity from the rostral intraparietal sulcus [33.Aflalo T. et al.A shared neural substrate for action verbs and observed actions in human posterior parietal cortex.Sci. Adv. 2020; 6eabb3984Crossref PubMed Scopus (0) Google Scholar], a region deemed to include the phAIP [34.Orban G.A. Functional definitions of parietal areas in human and non-human primates.Proc. Biol. Sci. 2016; 283: 1828Google Scholar]. The findings revealed impressive similarities with those reported in monkeys. First, in each viewpoint, approximately 20% of phAIP neurons were OMA selective, as in the monkey: the majority of them showed facilitated response to OMAs (Figure 1D), whereas a smaller set (about 15%) were suppressed in both humans and monkeys. Second, phAIP neurons could be tuned to any of the exemplars tested, but coverage of OMA exemplars was more uniform in humans than in the monkeys. Third, OMA exemplars could be decoded from the phAIP population activity recorded in each of the two tested patients, providing significant information about the observed exemplar with the same latency reported in the monkey (150 ms from video onset). Finally, format-dependent coding was evident also among human neurons, and although it is difficult to reach a firm conclusion based on the available evidence, it is plausible that a multiplicative mixing of visual format and OMA-identity information has been preserved from the common ancestor of humans and monkeys. However, differently from the monkey AIP, a sizable fraction of human phAIP OMA-selective neurons exhibited format-invariant tuning (80% were posture invariant and 55% viewpoint invariant), which is consistent with the evidence of generalization across viewpoints during OMA-discrimination tasks in humans [35.Platonov A. Orban G.A. Action observation: the less-explored part of higher-order vision.Sci. Rep. 2016; 6: 36742Crossref PubMed Scopus (8) Google Scholar]. The greater invariance of human OMA-selective neurons may thus facilitate the recruitment of neural representations of observed actions, even by reading action verbs [33.Aflalo T. et al.A shared neural substrate for action verbs and observed actions in human posterior parietal cortex.Sci. Adv. 2020; 6eabb3984Crossref PubMed Scopus (0) Google Scholar], a uniquely human capacity. To summarize, human and non-human primates (i.e., macaques) have a remarkably similar neuronal machinery in homologue regions of the rostral intraparietal sulcus, which encode OMA identity at a variable degree of visual invariance and abstraction in order to access it, for example, via the human reading of written words [33.Aflalo T. et al.A shared neural substrate for action verbs and observed actions in human posterior parietal cortex.Sci. Adv. 2020; 6eabb3984Crossref PubMed Scopus (0) Google Scholar]. The functional similarities between basic properties of monkey and human OMA-selective neurons raise the fundamental question of what the underlying anatomical architecture might be. The tuning for OMAs is prevalent in the caudal portion of AIP, a region where the influence of own-hand visual feedback and overall visual responsiveness was found to be stronger than in the rostral sector [30.Lanzilotto M. et al.Anterior intraparietal area: a hub in the observed manipulative action network.Cereb. Cortex. 2019; 29: 1816-1833Crossref PubMed Scopus (16) Google Scholar]. In that study, neural tracers were injected at three distinct positions along the rostro-caudal extent of the physiologically investigated region. The results confirmed previous anatomical findings [36.Borra E. et al.Cortical connections of the macaque anterior intraparietal (AIP) area.Cereb. Cortex. 2008; 18: 1094-1111Crossref PubMed Scopus (302) Google Scholar] (Figure 2A ) and revealed quantitative differences in the connectivity patterns between the caudal and rostral AIP (Figure 2B). In particular, the caudal part of AIP with stronger OMA selectivity, exhibited stronger connections with rostral and caudal prefrontal regions, caudal parietal convexity and lateral intraparietal area, and a variety of occipito-temporal regions (Figure 2C). Although OMA-identity coding has yet to be investigated in brain regions other than AIP, previous neurophysiological studies [23.Perrett D.I. et al.Frameworks of analysis for the neural representation of animate objects and actions.J. Exp. Biol. 1989; 146: 87-113Crossref PubMed Google Scholar] reported that neurons in the lower bank of the rostral superior temporal sulcus (STS), known as hand–object-interaction neurons, signal the relationship between a moving hand and its target. Indeed, the discharge of such neurons was lower when the hand or the target was presented in isolation or at some distance one from each other. Furthermore, some STS neurons responded when the observed hand was that of the recorded monkey, similarly to many AIP neurons [22.Maeda K. et al.Functional properties of parietal hand manipulation-related neurons and mirror neurons responding to vision of own hand action.J. Cogn. Neurosci. 2015; 27: 560-572Crossref PubMed Scopus (40) Google Scholar,30.Lanzilotto M. et al.Anterior intraparietal area: a hub in the observed manipulative action network.Cereb. Cortex. 2019; 29: 1816-1833Crossref PubMed Scopus (16) Google Scholar]; such responses might enable monkeys to assess the consequences of their own hand–object interactions. Importantly, these responses were relatively unaffected by most properties of the object except its rigidity or food quality. Finally, these neurons responded also when tested with different body movements that resulted in similar effects on the object, suggesting that they essentially code the hand–object interaction rather than the observed action itself. The anatomical location of these STS neurons corresponds to area TEa [37.Seltzer B. Pandya D.N. Afferent cortical connections and architectonics of the superior temporal sulcus and surrounding cortex in the rhesus monkey.Brain Res. 1978; 149: 1-24Crossref PubMed Scopus (575) Google Scholar], one of the most prominent sources of temporal projection to AIP, targeting mainly its caudal part, where OMA-selective neurons prevail [30.Lanzilotto M. et al.Anterior intraparietal area: a hub in the observed manipulative action network.Cereb. Cortex. 2019; 29: 1816-1833Crossref PubMed Scopus (16) Google Scholar]. Thus, the TEa–AIP pathway (red arrow in Figure 2C) likely represents the source to AIP of visual information about the dynamics of hand–object interaction resulting from the observed manipulations. Areas IPa/PGa (Figure 2C) represent another potentially relevant source of visual information about OMA identity. A study that approximately targeted this middle-STS region [38.Vangeneugden J. et al.Functional differentiation of macaque visual temporal cortical neurons using a parametric action space.Cereb. Cortex. 2009; 19: 593-611Crossref PubMed Scopus (75) Google Scholar], reported neuronal selectivity for two features of observed forelimb actions, portrayed by stick figures: static posture and body-part deformation, encoded by 'snapshot' neurons and kinematic features, encoded by 'motion' neurons. These cells could provide a rich set of information about others' body-part movements, which are critical for extracting OMA identity. Importantly, another study recently showed that the middle-STS region is involved in the visual processing of social interactions [39.Sliwa J. Freiwald W.A. A dedicated network for social interaction processing in the primate brain.Science. 2017; 356: 745-749Crossref PubMed Scopus (85) Google Scholar,40.Ong W.S. et al.Neuronal correlates of strategic cooperation in monkeys.Nat. Neurosci. 2021; 24: 116-128Crossref PubMed Scopus (0) Google Scholar], constituting a key node of the recently proposed 'third visual pathway' [41.Pitcher D. Ungerleider L.G. Evidence for a third visual pathway specialized for social perception.Trends Cogn. Sci. 2020; 25: 100-110Abstract Full Text Full Text PDF PubMed Scopus (2) Google Scholar]. Thus far, there is little evidence for the view independence of middle-STS neurons. Indeed, middle-STS body-patch neurons display mostly view-dependent coding of body posture and identity [12.Kumar S. et al.Transformation of visual representations across ventral stream body-selective patches.Cereb. Cortex. 2019; 29: 215-229Crossref PubMed Scopus (8) Google Scholar], which is in line with previously reported properties of STS neurons encoding body movements (such as walking and bending the knee) [42.Jellema T. Perrett D.I. Neural representations of perceived bodily actions using a categorical frame of reference.Neuropsychologia. 2006; 44: 1535-1546Crossref PubMed Scopus (0) Google Scholar]. Thus, IPa/PGa may provide view-dependent information regarding body movements to AIP (blue arrow in Figure 2C), coherently with the strong tuning for visual formats reported in monkey's AIP [31.Lanzilotto M. et al.Stable readout of observed actions from format-dependent activity of monkey's anterior intraparietal neurons.Proc. Natl. Acad. Sci. U. S. A. 2020; 117: 16596-16605Crossref PubMed Scopus (3) Google Scholar]. The anatomo-functional evidence reviewed in the preceding text suggests that the monkey's caudal area AIP receives from the STS two convergent sources of visual information relevant to OMA-identity processing (Figure 3, Key Figure): body-movement signals from IPa/PGa and hand–object-interaction signals from TEa. Considering the homology of STS regions [43.Jastorff J. et al.Integration of shape and motion cues in biological motion processing in the monkey STS.Neuroimage. 2012; 60: 911-921Crossref PubMed Scopus (0) Google Scholar], this scheme can be extended to humans. Indeed, the phAIP of the monkey TEa is located in the posterior occipitotemporal sulcus and extends into the fusiform gyrus [43.Jastorff J. et al.Integration of shape and motion cues in biological motion processing in the monkey STS.Neuroimage. 2012; 60: 911-921Crossref PubMed Scopus (0) Google Scholar]: this region may contribute to processing object changes caused by others' actions [44.Wurm M.F. Caramazza A. Lateral occipitotemporal cortex encodes perceptual components of social actions rather than abstract representations of sociality.Neuroimage. 2019; 202116153Crossref PubMed Scopus (3) Google Scholar]. By contrast, the human homologue of the monkey IPa/PGa regions may be split between the posterior occipitotemporal sulcus and the posterior middle temporal gyrus, and extend into the posterior STS [43.Jastorff J. et al.Integration of shape and motion cues in biological motion processing in the monkey STS.Neuroimage. 2012; 60: 911-921Crossref PubMed Scopus (0) Google Scholar]; these regions correspond to the activations in the lateral and ventral temporal cortex during observation of body movements [45.Beauchamp M.S. et al.Parallel visual motion processing streams for manipulable objects and human movements.Neuron. 2002; 34: 149-159Abstract Full Text Full Text PDF PubMed Scopus (409) Google Scholar,46.Jastorff J. Orban G.A. Human functional magnetic resonance imaging reveals separation and integration of shape and motion cues in biological motion processing.J. Neurosci. 2009; 29: 7315-7329Crossref PubMed Scopus (0) Google Scholar]. Why is OMA identity represented in primates' intraparietal cortex? As mentioned previously, in both humans and monkeys, OMA-selective neurons can show either facilitated or suppressed visual responses; however, when monkeys are tested during active execution of reaching–grasping actions in the dark, only facilitated neurons (not suppressed ones) also show a genuine motor response. On this basis, we proposed [30.Lanzilotto M. et al.Anterior intraparietal area: a hub in the observed manipulative action network.Cereb. Cortex. 2019; 29: 1816-1833Crossref PubMed Scopus (16) Google Scholar] that OMA-selective AIP visuomotor neurons provide signals for action planning based on the monkey's processing of what another is doing. This mechanism would work alongside the one previously described for object affordances. Indeed, the physical features of observed objects are represented in both parietal [18.Murata A. et al.Selectivity for the shape, size, and orientation of objects for grasping in neurons of monkey parietal area AIP.J. Neurophysiol. 2000; 83: 2580-2601Crossref PubMed Google Scholar,21.Pani P. et al.Grasping execution and grasping observation activity of single neurons in the macaque anterior intraparietal area.J. Cogn. Neurosci. 2014; 26: 2342-2355Crossref PubMed Scopus (45) Google Scholar] and premotor [20.Schaffelhofer S. Scherberger H. Object vision to hand action in macaque parietal, premotor, and motor cortices.eLife. 2016; 5e15278Crossref PubMed Scopus (47) Google Scholar,47.Jeannerod M. et al.Grasping objects: the cortical mechanisms of visuomotor transformation.Trends Neurosci. 1995; 18: 314-320Abstract Full Text PDF PubMed Scopus (1029) Google Scholar,48.Bonini L. et al.Space-dependent representation of objects and other's action in monkey ventral premotor grasping neurons.J. Neurosci. 2014; 34: 4108-4119Crossref PubMed Scopus (67) Google Scholar] neurons, forming a parieto-frontal circuit with an experimentally established causal role in visually guided reaching–grasping actions [49.Gallese V. et al.Deficit of hand preshaping after muscimol injection in monkey parietal cortex.Neuroreport.