Action preparation in grasping reveals generalization of precision between implicit and explicit motor processes
Introduction
It is well known that the human higher-level visual system exhibits two main cortical processing streams. The so-called ventral stream projects from V1 (the primary visual cortical area) to IT (the inferotemporal cortex). The so-called dorsal stream, conversely, projects from V1 to PPC (the posterior parietal cortex). What are these streams for? That is, do these distinct neural structures subserve different functions? Ungerleider and Mishkin (1982) proposed that the ventral stream processes object identity whereas the dorsal stream processes space. Underpinning this “what” vs. “where” distinction were earlier observations on the behavioral consequences of brain lesions in humans (Holmes, 1918; Newcombe, 1969; for reviews see Cooper, 2012; Cooper and O'Sullivan, 2016) and on cortical vs. non-cortical visual pathways in non-human animals (Schneider, 1969). Goodale and Milner (1992; Milner and Goodale, 1995, 2006; 2008) refined this proposal, suggesting that the difference may lie not in what visual information is processed, but on the purpose of this processing. In the ventral stream, visual processing aims at achieving representations that remain invariant under changes in the conditions of observation, enabling object recognition and identification ("vision-for-perception”). In the dorsal stream, conversely, processing aims at monitoring parameters relevant to limb movement, enabling the online control of goal-directed movements ("vision-for-action”). Milner and Goodale's proposed division of labor is generally referred to as the “two-visual-systems hypothesis” (TVSH).
The TVSH has provided a framework for interpreting data from diverse approaches such as neuropsychological patient studies (e.g. Goodale et al., 1991), brain imaging studies (e.g. Faillenot et al., 1997), and combination psychophysics-kinematics studies (e.g., Aglioti et al., 1995). In addition, the TVSH has proved heuristic in stimulating research on the functional interpretation of the dorsal-ventral distinction. One important issue here has been the degree of encapsulation within vision-for-action. Is this subsystem completely independent, or does it interact with other visual processes and if so, when and how? Although the original version of the TVSH stressed independence, later revisions of the model have explicitly recognized the need to study how vision-for-action may interact with other visual processes (e.g. Goodale, 2011; Milner, 2017; Whitwell et al., 2014). Here we focus on one potential form of such interactions, namely, whether seeing contextual objects before acting on a different test object affects how the action is prepared and executed. The theoretical motivation of such question is developed in the paragraphs that follow.
According to the TVSH, visual information is coded in relation to different spatial frameworks depending on the processing goal: object recognition or online motor control. These two processing goals are subject to different computational constraints. Object recognition requires to represent features such as size, shape, and orientation in relation to the location of other objects. Such object-relative representations promote constancy under changes in viewing conditions, a key requirement for recognition (Biederman, 1987; Marr, 1982; Milner and Goodale, 2006). Online motor control, conversely, requires that features be coded with respect to effectors. Effector-relative spatial representations provide the motor system with the online information required to configure limb trajectories and hand shaping for the intended actions. As a consequence of these differing constraints, vision-for-perception emphasizes relatively long-lasting spatial and temporal allocentric relations. Vision-for-action, conversely, focuses on the here-and-now, fast-decaying relations between objects and effectors to produce continuously updated egocentric representations. Thus, in the TVSH vision-for-action is mainly conceived as a system for real-time control (Westwood and Goodale, 2003) and should be unaffected by the temporal context preceding a movement. In a comprehensive recent review, Schenk and Hesse (2018) examined the empirical support for such a strong prediction, which they called the “dorsal amnesia” hypothesis, and concluded that dorsal amnesia is not generally supported. We take this as a call for further research on the nature of potential temporal interactions (or lack thereof) within specific motor tasks that may relate to specialized cortical networks (see Schenk et al., 2011). This line of research promises to further qualify functional subdivisions within the primate visual systems beyond useful first-order approximations suggested by dichotomous models such as the TVSH.
Our starting point is the hypothesis that seeing an object elicits object-related motor representations even if no action is executed and even if no other processing of the object is required by the task (see for instance Chao and Martin, 2000; Ellis and Tucker, 2000; Grafton et al., 1997; Grèzes and Decety, 2002; Tucker and Ellis, 1998, 2001; 2004). We call this the automatic visuomotor coding hypothesis. Based on this hypothesis, one might expect that seeing an object having size and shape that make it graspable relative to a typical human hand should automatically evoke a motor process for grasping. The so-called “canonical” neurons observed in monkey area F5 are generally believed to constitute a candidate neural mechanism underpinning the automatic elicitation of motor representations from visual information (Rizzolatti and Gentilucci, 1988; Murata et al., 1997). In humans, evidence from visuomotor priming studies (Craighero et al., 1996; Craighero et al., 1998) also supports the idea that seen visual features elicit motor representations. However, it is currently not clear whether motor representations elicited from seeing objects will affect the actual planning or actual execution of a later object-directed action, and how this might take place (but see, Hesse et al., 2008; Seegelke et al., 2016). A more detailed account of visuomotor priming studies is provided in the following sections along with corresponding predictions with regard to our study.
Here we capitalized on a recent study by our laboratory (Uccelli et al., 2019) which exploited a visual version of the Uznadze size-contrast illusion. In the Uznadze illusion, the same object is perceived as smaller when preceded by a larger object and as larger when preceded by a smaller object (Uznadze, 1966; Bruno et al., 2018). To test whether vision-for-action remains unaffected by this form of size contrast, as predicted by the TVSH, Uccelli and collaborators compared the maximum in-flight finger aperture (maximum grip aperture, MGA) for grasping and the indicated finger aperture in cross-modal size matches as a function of the size (smaller or larger) of a preceding distractor. In contrast with the TVSH prediction, they found that participants opened their fingers more when grasping targets preceded by smaller inducers, and less when grasping targets preceded by larger inducers (in comparison with a baseline condition with inducers having the same size as the test). This result runs counter the predictions of the TVSH. However, it does not specifically address automatic visuomotor coding. In this paper, we modified our earlier paradigm to perform such a test.
To test whether motor processes associated with grasping an object are modulated by motor processes elicited by previously viewing another object, we varied the size of distractor and test stimuli and explicitly associated different types of precision grips (Napier, 1956) to each level of size. Specifically, small stimuli were associated with pincer grips (two fingers), medium stimuli with tripod grips (three fingers), and large stimuli with pentapod grips (all five fingers). In an open-loop paradigm, we asked participants to grasp test objects using the associated grip types. Presentation of test objects was preceded by that of distractors that putatively elicited congruent or incongruent grips. Our specific objective was therefore to explore potential interactions between a hypothesized first, implicit motor representation and a second, explicit motor process. Critically, we focused on the planning phase of the movement, i.e., the sensorimotor computations which are carried out before movement initiation. To evaluate the preparatory phase of the action, we assessed the brief temporal interval spent by the participant observing the target before grasp initiation (preview reaction times, PRT). As preview reaction times were completely under the control of participants, we expected that their duration may reveal benefits (shorter PRTs) or costs (longer PRTs) as a consequence of interactions with implicit motor representations elicited by the distractors. That is, as the action goal and stimulus position were kept the same, we attributed differences in PRTs to interactions between implicit and explicit motor processes in the planning phase of the movement. Different patterns of benefits and costs are predicted by different hypotheses about the nature of such interactions, as detailed in the next paragraph. To evaluate the actual execution of the grips, we recorded grip kinematics and measured the MGA. Predictions for the pattern of this kinematic marker are developed in the paragraph following the next.
Section snippets
Preview reaction times predictions
Real-time motor programming. According to the real-time view of motor control, actions are guided on the basis of fast-decaying, continuously updated egocentric representations (Westwood and Goodale, 2003). This view predicts that each new grip will be programmed anew without taking into account visual information that may have been available before the presentation of the grip target. Thus, preview reaction times should be unaffected by the congruence of the target with a previously presented
General methods
We performed two experiments. Here we present their common methodological features. Details specific to each experiment are presented in subsequent sections along with the corresponding results and discussion.
Participants
We recruited 16 participants (8 females and 8 males; age range 23–32) from the University of Parma community. All were right-handed, had normal or corrected-to-normal eyesight, no history of neurological diseases, and were unaware of the purpose of the study.
Design
The design of the experiment resulted from crossing distractor and target stimuli as within-participant independent variables. We used three disk sizes (small, medium, and large) for both the distractor and the target stimuli. This yielded
Experiment 2
We introduced two changes in the experimental design of the first experiment. One was merely a simplification, aimed at better revealing potential differences. To achieve this goal, we dropped all conditions involving tripod grasp and focused on the comparison between the pincer and the pentapod. The other, and more critical, change was the addition of a no-distractor baseline. Preview reaction times associated with such baselines are informative about the preparatory phase of each type of
General discussion and conclusions
Our findings support the automatic visuomotor coding hypothesis. Specifically, our two experiments provide support for the hypothesis that a distractor stimulus can activate an implicit motor representation for grasping it and that this process is obligatory as distractors were irrelevant to our task. Thus, motor representations elicited by an implicit preparatory process can interact with motor representations involved in planning a subsequent actual grasp on a different object. These findings
Funding
This research was not funded by any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
CRediT authorship contribution statement
Veronica Pisu: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Visualization, Writing - original draft, Writing - review & editing. Stefano Uccelli: Data curation, Investigation. Lucia Riggio: Conceptualization, Writing - review & editing. Nicola Bruno: Conceptualization, Formal analysis, Investigation, Methodology, Project administration, Visualization, Writing - review & editing.
Declaration of competing interest
None.
References (70)
- et al.
Visual coding for action violates fundamental psychophysical principles
Curr. Biol.
(2008) - et al.
Size-contrast illusions deceive the eye but not the hand
Curr. Biol.
(1995) - et al.
Mixed effects modeling with crossed random effects for subjects and items
J. Mem. Lang.
(2008) - et al.
Both vision-for-perception and vision-for-action follow Weber's law at small object sizes, but violate it at larger sizes
Neuropsychologia
(2016) - et al.
Visual similarity modulates visual size contrast
Acta Psychol.
(2018) - et al.
No evidence for visuomotor priming in a visually guided action task
Neuropsychologia
(2005) - et al.
Human neuroimaging reveals the subcomponents of grasping, reaching and pointing actions
Cortex
(2018) - et al.
Representation of manipulable man-made objects in the dorsal stream
Neuroimage
(2000) Transforming vision into action
Vis. Res.
(2011)- et al.
Separate visual pathways for perception and action
Trends Neurosci.
(1992)
Premotor cortex activation during observation and naming of familiar tools
Neuroimage
Does visual perception of object afford action? Evidence from a neuroimaging study
Neuropsychologia
Visual information about object size and object position are retained differently in the visual brain: evidence from grasping studies
Neuropsychologia
Two visual systems reviewed
Neuropsychologia
Do we have distinct systems for immediate and delayed actions? A selective review on the role of visual memory in action
Cortex
Vision-for-perception and vision-for-action: which model is compatible with the available psychophysical and neuropsychological data?
Vis. Res.
Visuomotor priming of action preparation and motor programming is similar in visually guided and memory-guided actions
Neuropsychologia
Action priming by briefly presented objects
Acta Psychol.
Biomechanical factors may explain why grasping violates Weber's law
Vis. Res.
Coordinated control programs for movements of the hand
Surround inhibition in the motor system
Exp. Brain Res.
Recognition-by-components: a theory of human image understanding
Psychol. Rev.
The speed-accuracy trade-off in manual prehension: effects of movement amplitude, object size and object width on kinematic characteristics
Exp. Brain Res.
An analysis of transformations
J. Roy. Stat. Soc.
Higher visual function: hats, wives and disconnections
Practical Neurol.
Here, there and everywhere: higher visual function and the dorsal visual stream
Practical Neurol.
Evidence for visuomotor priming effect
Neuroreport
Visuomotor priming
Vis. Cognit.
The new statistics: why and how
Psychol. Sci.
Micro-affordance: the potentiation of components of action by seen objects
Br. J. Psychol.
Visual pathways for object-oriented action and object recognition: functional anatomy with PET
Cerebr. Cortex
Elements of Psychophysics
The GRASP taxonomy of human grasp types
IEEE Transactions on Human-Machine Systems
The information capacity of the human motor system in controlling the amplitude of movement
J. Exp. Psychol.
Naming and grasping common objects: a priming study
Exp. Brain Res.
Cited by (6)
Influence of colour on object motor representation
2022, NeuropsychologiaCitation Excerpt :Although this notion is still matter of debate (see for example, Chong and Proctor, 2020; de Wit et al., 2017), it is well accepted that perception and action are interrelated. Empirical evidence suggests that a motor representation might be automatically elicited when seeing a graspable object, even when actions are not directed upon target object (e.g., Bub and Masson, 2010; Buccino et al., 2009; Craighero et al., 1996; De Stefani et al., 2012; Ellis et al., 2007; Garofalo et al., 2020; Grèzes et al., 2003; Pisu et al., 2020; Rounis et al., 2018; Tucker and Ellis, 2001, 2004; Vainio et al., 2014). So far, two object features are mainly considered to study the action possibilities emerging from the interaction between the agent and the object, that are size/shape and orientation (Ellis and Tucker, 2000; Tucker and Ellis, 1998).
Precision in grasping: Consistent with Weber's law, but constrained by “safety margins”
2021, NeuropsychologiaAsymmetric effects of graspable distractor disks on motor preparation of successive grasps: A behavioural and event-related potential (ERP) study
2020, International Journal of PsychophysiologyCitation Excerpt :Specifically, we aimed to assess how an implicit motor representation might affect the preparation of an explicit successive grasp. If PRTs for grasping a target are modified by the presentation of incongruent distractors (as reported by Pisu et al., 2020), and if this effect occurs at the level of motor preparation (as hypothesized again by Pisu et al.), then corresponding ERP modulations should be observed in R-LRP's, but not in S-LRP or P300 components. To determine the adequate sample size, we conducted an a-priori power analysis.
The effect of the Uznadze illusion is temporally dynamic in closed-loop but temporally constant in open-loop grasping
2023, Quarterly Journal of Experimental PsychologyPriming of the Sander Parallelogram illusion separates perception from action
2021, Experimental Brain Research