THEORETICAL REVIEW

Embodied memories: Reviewing the role of the bodyin memory processes

Francesco Ianì1

# The Psychonomic Society, Inc. 2019

AbstractThis review aims at exploring the role of the body and its sensorimotor processes in memory. Recent theories have suggested thatmemories can profitably be seen as mental simulations consisting in the reactivation of sensorimotor patterns originally associ-ated with events at encoding, rather than amodal mental representations. The sensorimotor model of memory (SMM) claims thatthe body is the medium where (and through which) sensorimotor modalities actually simulate the somatosensory components ofremembered events, and predicts that memory processes can be manipulated through manipulation of the body. The reviewanalyzes experimental evidence in favor of the SMMand the claim that the body is at stake inmemory processes. The review thenhighlights how, at the current state of research, the majority of this evidence concerns the effect of body manipulations onmemory processes rather thanmemory representations. It considers the need for a more circumstantial outline of the actual extentto which the body is capable of affecting memory, specifically on some important areas still unexplored, such as the sense ofrecollection. Resulting strengths and limitations of the SMM are discussed in relation to the more general debate on the embodiedcognition.

Keywords Memory . Body . Embodied cognition . Sensorimotor reactivation . Recollection

Over the past decades, scholars have changed the way sciencelooks at human memory because of a paradigm shift that hasset aside the multistore model of memory (e.g., Atkinson &Shiffrin, 1968). The first theoretical repositioning occurredwhen the concept of network made it possible to supersedethe idea that memory was a deposit of discrete immutableabstract items neatly stored in our brains, and proposed thesemantic network model instead. The latter model depictedmemory as a set of interrelated functions implemented in anetwork of interconnected nodes (e.g., Collins & Loftus,1975). In both the semantic network and the multistore model,however, memory traces had nothing to do with action andperception, but were simply combinations of abstract symbolsmanipulated according to syntactical rules (Fodor, 1983;Pylyshyn, 1984). The most significant change, so to speak,was still to come.

This occurred about 30 years ago and changed the veryontology of human cognition and memory. The shift wasprompted by the so-called grounding problem (Harnad,1990): abstract symbol manipulation models cannot explainhow symbolic systems connect with the world. Mental rep-resentations need to be grounded in perception and action,as they cannot be a Bfree-floating system of symbols^(Dijkstra & Zwaan, 2014, p. 296). Memories are not abstractitems neatly stored in our brains simply because they emergefrom proximal sensory projections that include sensorimotorelements in their representations. The new somatic nature ofmemory, therefore, appears to be strictly linked to the sym-bol grounding hypothesis; that is, the idea that symbols onlyget their meaning through sensorimotor experience (Harnad,1990). Since the seminal paper by Harnad (1990), an ever-increasing number of different theoretical approaches haveendorsed the view that the body is key to shaping higherlevel cognitive functions such as memory (see, e.g., Wilson,2002). Robust evidence has substantiated the claim that be-cause mental representations are indeed grounded in bothaction and perception, no theory of cognition can bypassthe grounding problem and be true to the facts. Inparticular, the one maintained by Damasio (1989) has

* Francesco Ianìfrancesco.iani@unito.it

1 Dipartimento di Psicologia, Università di Torino, Via Po, 14,10123 Turin, Italy

https://doi.org/10.3758/s13423-019-01674-xPsychonomic Bulletin & Review (2019) 26:1747–1766

Published online: 25 October 2019

http://crossmark.crossref.org/dialog/?doi=10.3758/s13423-019-01674-x&domain=pdfmailto:francesco.iani@unito.it

become clear: information recognition as well as informationrecall require activity in multiple brain areas to take placenear the sensory and motor regions. Consequently, scientistsand researchers have come to conceptualize mental statesand processes also in terms of embodied cognition and havecommitted to the view that memory trace activation, at leastpartly, enacts neural systems typically associated with per-ceptual and motor mechanisms engaged in encoding inputinformation (e.g., Barsalou, 1999; Dijkstra & Zwaan, 2014;Glenberg, 1997). As argued by Glenberg (1997), previoustheories simply presupposed that memory is a storage sys-tem and assumed that storing could be studied independentlyof how our own body and actions affect memory function-ing and are in turn affected by that in real life situations. Inpoint of fact, memory and (in general) cognitive functionshave evolved to serve human agency and facilitate the inter-actions between us and our environment (Varela, Thompson,& Rosch, 1991; Barkow, Cosmides, & Tooby, 1992; Buss,2005; Samson & Brandon, 2007; Callebaut & Rasskin-Gutman, 2005). Therefore, both cognition and memory pro-cesses are grounded in human experience and partake of areal-world environment impinging on perception and involv-ing action. From this new perspective, a given stimulus isstored in the sensorimotor pathways that were activated,shaped, and strengthened by the item when it was initiallyprocessed: memory processes are no longer higher ordercognitive activities totally detached from ordinary sensoryprocessing; on the contrary, they reactivate it, albeit partiallyor, anyway, in a slightly different manner.

This theoretical change occurred within a wider paradigmshift in cognitive psychology, that leads to the so-called em-bodied cognition (EC) approach. One of the central cores,among others, of this new perspective is that cognition Bisstrongly influenced by the body^ (Glenberg, Witt, &Metcalfe, 2013, p. 573). This review aims first at collectingevidence in favor of the claim that the memory system can beactually influenced by body manipulations. To this aim, it firstintroduces the sensorimotor model (SMM), according towhich during encoding people register perceptual and motorinformation, and later on, when the encoded event is recalled,these representations are reactivated (see the next section).Second, it reviews studies inspired to SMM, coming fromdifferent research fields: eye-movements, co-speech gestures,body posture, and bodily expression of emotion studies. Thereview is mainly focused on episodic memory rather than onsemantic memory. This choice has beenmade for two reasons.First, there are many reviews on concept representations andsensorimotor simulation (see, e.g., Meteyard, Cuadrado,Bahrami, & Vigliocco, 2012; Pulvermüller, 2013), and sec-ond, because the present article aims at addressing broader andmore general issues on memory, by focusing on the effects ofbodily manipulations at encoding and recall and the role of thebody in the encoding specificity principle.

What emerges is that a growing body of evidence seems tosupport the SMM model and the general notion that the bodyis able to affect memory. In particular, sensorimotor reactiva-tion it is not an epiphenomenon, but a privileged componentof the memory traces through which our cognitive system isable to retrieve information. Therefore, the EC approachseems effective. However, as pointed out by Goldinger,Papesh, Barnhart, Hansen, and Hout (2016), memory is influ-enced by the body is a claim too vague. To better delineate thequestion on how events are represented and organized in ourmemory, one needs to outline the exact degree by which sen-sorimotor involvement determines the efficiency of memoryprocesses. It is well established that when we remember, ourbody, and its potential actions, are involved in some nontrivialway, but at the same time, several questions still need to beanswered: How and how much would the body affect cogni-tion? Within which cognitive areas? Within what limits?

In the last part of the review, after presenting some conflict-ing studies, I defend an embodied approach on memory stud-ies, highlighting at the same time its limitations coming from acritical evaluation of the empirical evidences. Indeed, the stud-ies on the effect of body manipulations on memory detectedinterference effects that lead to a heterogeneous set of memoryimpairments, but, at the current state of research, the majorityof evidence concerns effect on memory processes rather thanmemory representations. I will highlight how these studiesseem to suggest that memory depends on sensorimotor pro-cesses, but only partially, and that in order to delineate the roleof body inmemory, one needs to exactly outline to what extentmemory depends on these sensorimotor processes.Limitations of the SMM arise regarding important areas stillunexplored. In this regard, one of the most compelling futurechallenges will be the investigation about the possible role ofbody manipulation in affecting the Bsense of recollection^—that is, the subjective sense of reliving the original event.

The sensorimotor simulation model (SMM)

Barsalou elaborated Glenberg’s idea further by developing atheory of knowledge called the perceptual symbol system(PSS; Barsalou 1999, 2008). The PSS assumes the existenceof a perceptual memory system through which the associationareas in the brain capture bottom-up activation patterns occur-ring in the sensorimotor areas, whereupon an opposite top-down process is initiated and the association areas reactivatethe sensorimotor areas to create perceptual symbols (see alsoDamasio, 1989; Edelman, 1989). Thus, memory traces arebetter understood in terms of sensorimotor encoding: theystore information on the neural states underpinning the per-ception of our environment, our body, and our movements.Since they are nothing but neural patterns, memory traces aredynamic—that is, plastic and bound to be modified by

Psychon Bull Rev (2019) 26:1747–17661748

subsequent encodings—so that what is retrieved in the futurewill not fully match in every minute detail what was at firstencoded (Barsalou, 1999; Edelman, 1987, 1989).

According to the SMM, a given event fundamentally con-sists in perceptual information so that a reactivation of thesame sensorimotor circuitry originally involved in its percep-tion is also at stake whenever the event is recalled or comes tomind. In this respect, remembering is tantamount to creatingmental simulations of bodily experiences in modality-specificregions of the brain. Memory consists in partial (or covert)reenactments of sensory, motor, and introspective states, notin amodal redescriptions of these states as suggested by thedigital computer-inspired theories of the mind that dominatedcognitive science during the late 20th century (e.g., Fodor &Pylyshyn, 1988). A growing body of literature has brought tothe fore how body movements and position can affect cogni-tion, especially mental simulation (see Körner, Topolinski, &Strack, 2015). In particular, some authors have hypothesizedthat mental simulations depend on the brain’s sensorimotorsystem to such a great extent as to call them Bsensorimotorsimulations^ (e.g., Dijkstra & Post, 2015). As stated byBarsalou (1999) in the PSS theory, the sensorimotor neuralcircuits responsible for encoding perceptual information alsostore that information. From this perspective, when we try toremember information, we Bmentally simulate^ the originalevent, and the cortical reactivation of the neural areas activat-ed when first encoding that event inheres in this cognitiveprocess as well (for a review, see Kent & Lamberts, 2008).

In fact, Kolers (1984) anticipated the SMM, arguing thatthe procedures by which information is encoded are alsostored in memory and could be used to speed up retrieval. Inthis respect, it would be impossible to track a clear distinctionbetween storage and processing processes. Essentially, theSMM predicts that memory processes draw upon all of theperceptual modalities contributing to our experience and spec-ulates that memory is distributed across the various modality-specific brain areas devoted to action and perception (e.g.,Brunel, Labeye, Lesourd, & Versace, 2009; Niedenthal,Barsalou, Winkielman, Krauth-Gruber, & Ric, 2005), includ-ing those responsible for proprioception and introspection (forthe role of introspection in cognition and the brain, see LeDoux, 2002; Solms & Turnbull, 2002). In short, memorytraces are multimodal in nature, so their reactivation throughmental simulation implies a multimodal reactivation. Recenttheoretical views on memory mechanisms, such as the Act-Inmodel (Versace et al., 2014), have stressed the pivotal roleplayed by the body in memory functioning, contending thatmemory traces capture and reflect all the components of pastexperiences (see also Damasio, 1989; Edelman 1987, 1989);specifically, their sensory properties are captured by sensoryreceptors, regardless of whether they are visual, acoustic (see,e.g., Wheeler, Petersen, & Buckner, 2000), or motor (see, e.g.,Nilsson et al., 2000) information.

Several neuroimaging and behavioral studies have gath-ered good evidence suggesting that shared modality-specificactivation is what actually happens between encoding andrecall. For example, Wheeler et al. (2000) found that retriev-ing vivid visual and auditory information reactivates some ofthe same sensory regions initially activated in its perception(i.e., the precuneus and the left fusiform cortex). The samepattern of results has been detected for encoding and retrievalof spatial information in the inferior parietal cortex (e.g.,Persson & Nyberg, 2000). Such evidence is often cited tosupport the so-called reactivation hypothesis (see, e.g.,Nyberg, 2002; Nyberg et al., 2001), which amounts to the ideathat experiencing an event in the recognition phase and men-tally reconstructing it at recall share the same brain modality-specific activation patterns.

To explore motor pattern reactivation, Nilsson et al. (2000)conducted an experiment in which the participants had tomemorize a set of verbal commands in three separate experi-mental conditions: (1) in the verbal condition, they just had tolisten to the commands; (2) in the imagery condition, theywere invited to imagine the action described by the commandswhile staying still; and (3) in the enactment condition, theywere asked to perform the action described by the commands.After each experimental condition, the participants had to en-gage in recall while being PET scanned. The results showedhigher activation rates in the primary motor cortex (M1) forthe enactment condition, lower rates for the verbal condition,and an average activation in the imagery condition (Nilssonet al., 2000). Similarly, Nyberg et al. (2001) ran an fMRI studyto directly compare brain activity during learning with brainactivity in subsequent recollections. They observed a substan-tial match between the cluster of regions activated in both thelearning and recall phases (markedly in the left ventral motorcortex), and concluded that memorization seems to depend onactivating and reactivating motor information (for similarresults, see Masumoto et al., 2006). This would suggest thatrecollecting memories is thus a sensorimotor simulationprocess—information retrieval occurs through simulating pastevents and simulation reactivates the same sensorimotor areasoriginally activated at encoding.

In the SMM, episodic memory retrieval is related to thebody because relevant sensorimotor aspects of the event anddetails on what it was about are reconstructed and packedtogether (Bietti, 2012). Yet although it is not the central focusof the present review, it is important to notice how the SMM isnot limited to episodic memory because the perceptual com-ponents of past experiences turn out to be reactivated even inthe case of semantic memory (e.g., Binder & Desai, 2011;Kiefer, Sim, Herrnberger, Grothe, & Hoenig, 2008; see alsoMeteyard et al., 2012; Pulvermüller, 2013). Such knowledgeis coded and grounded in sensorimotor brain systems similarlyto what happens for episodic memories (Tettamanti et al.,2005).

Psychon Bull Rev (2019) 26:1747–1766 1749

Further, it has been recognized that such reactivation worksboth ways: retrieval mechanisms reactivate encoding mecha-nisms, prompting encoding mechanisms to, in turn, facilitaterecall tasks (Dewhurst & Knott, 2010; Kolers, 1973). If one isto accept that recall is a simulation of encoding processes andstates—that is, repacking together the perceptual, affectiveand somatic components of human experiences, thenprompting mechanisms congruent to those affecting encodingwill speed up retrieval processes, while incongruent feelings,sensations or bodily movements will hinder them (see, e.g.,Dijkstra & Zwaan, 2007; Dijkstra & Post, 2014; Ianì &Bucciarelli, 2018; Riskind, 1983).

In the next section, we shall focus on encoding specificityand show how the mutual relatedness of body and memoryprocesses at the core of the sensorimotor simulation model ofmemory Bimplies that manipulations of the body and move-ment may result in memory changes^ (Dijkstra & Zwaan,2014, p. 298).

Sensorimotor memory pathways

According to the so-called principle of encoding specificity,encoding and recall are so interwoven that the more process-ing resources they share, the better information retrieval willbe. The idea that retrieval and encoding processes are strictlylinked to one another dates back to Tulving and Thomson(1973), who demonstrated that humans store very specificinformation about the context within which a stimulus isencoded. The literature has grown rapidly since then, and ithas emerged that cognitive resource sharing is mainly due tomemory. A classic example is the experiment by Baddeleyand colleagues, in which participants were asked to learn a listof words either on land or under water in full scuba gear(Godden & Baddeley, 1975). During the recall phase, the par-ticipants showed better memory when tested under the sameconditions as those under which they had learned the informa-tion (regardless of whether that was on land or in water) com-pared with when they were tested in a different environment.In other words, today we know that when a given stimulus isencoded into memory, the resulting stored representation con-tains information on the relevant stimulus plus information onany number of other situational and environmental cues pres-ent at the time the stimulus was encoded.

Kent and Lamberts (2008) argued that the SMM mightshed light on themechanisms underlying encoding specificity:the SMM holds that recalling perceptual informationreactivates the neural networks responsible for first processingthat information at encoding; so it quite naturally follows—and from memory architecture alone—that information avail-able at encoding may also be available at recall (Barsalou,2008; Slotnick, 2004; for semantic memory, see also Martin,2007; Thompson-Schill, 2003; for domain specificity, see also

Callebaut, Rasskin-Gutman, & Simon, 2005; Hirschfeld &Gelman, 1994). Indeed, at first, scientists considered the ex-tent to which cognitive resources happen to be shared in mem-ory processes with reference to environmental conditions on-ly; that is, limited to situational and environmental externalcues. However, based on evidence and hypotheses gatheredover time, predictions have emerged that the encoding-recallmatch primarily and most importantly affects sensorimotorinformation regarding the body and its movement. Memorytraces seem to contain detailed information on the posture,position, and movements the person underwent whileencoding a given experience, and that very same sensorimotorinformation is deemed to be reactivated during the retrievalphase.

Two main predictions have been drawn from the mutualrelatedness of body and memory processes:

1. The behavioral reenactment of processes involved in theencoding phase facilitates information retrieval; more pre-cisely, if memories are simulations reconstructing an orig-inal event along with its relevant sensorimotor compo-nents, then triggering those components at recall shouldspeed up the retrieval processes (Dijkstra & Zwaan,2014).

2. Behavioral tasks drawing upon the same neural resourcesas those reenacted at recall should slow down the retrievalprocesses; in other words, sensorimotor simulation maybe blocked by a concurrent task involving the same sen-sorimotor resources (Dijkstra & Post, 2015).

These predictions have been tested in several experimentalsettings.

Eye movements

There is growing consensus in the literature that the presenceof eye movements during retrieval can indicate that a reenact-ment of the original experience is taking place, and, in addi-tion, eye movements themselves seem functional to the re-trieval process. For example, Laeng and Teodorescu (2002)found that oculomotor movements performed at encoding toexplore certain given visual information were also reenactedat retrieval of the same stimulus, and crucially, the process ofimage generation was disrupted by forcing participants tomaintain a static oculomotor position (participants were askedto fix a central cross when they were asked to answer a ques-tion about a previously seen object): memory suffered whenspontaneous fixation was blocked at recall. More recently,Laeng, Bloem, D’Ascenzo, and Tommasi (2014) reportedthree experiments in which the participants first inspected astimulus and then were asked to retrieve it. In the first exper-iment, the authors found that imagining previously seen ob-jects (e.g., plain triangles with different orientation and form)

Psychon Bull Rev (2019) 26:1747–17661750

resulted in a pattern of eye fixations that mirrored the shapeand orientation of each object over the same region of thescreen where it had originally appeared. In the secondexperiment, they found that eye movements at recallsubstantially overlapped those used to scan the objects in theinitial phase of the study, even when more complex shapeswere used. Crucially, such an overlap predicted accuracy inmemory tasks in that those participants who reenacted eyemovements during recall more closely resembling theoriginal movements also showed higher scores in spatialmemory tasks. In Experiment 3, memory performancesignificantly decreased when gaze was forced to remain on afixation point distant from the original fixations. Interferingwith gaze during recall seems to decrease the quality of thememory. Johansson and Johansson (2014) addressed the samefundamental issue using four direct eye manipulations in theretrieval phase of an episodic memory task: (a) free viewingon a blank screen, (b) maintaining central fixation, (c) lookinginside a square congruent with the location of the to-be-recalled objects, and (d) looking inside a square incongruentwith the location of the to-be-recalled objects. They found thata central fixation constraint perturbed retrieval performanceby increasing reaction times needed for recalling such events.Secondly, memory retrieval was indeed facilitated when eyemovements were manipulated toward a blank area thatmatched the original location of the to-be-recalled object.The results were robust both in respect of memory accuracyand RTs. Their findings provide novel evidence of an activeand facilitatory role of gaze position in memory retrieval anddemonstrate that memory for spatial relationships betweenobjects is more readily affected than memory for intrinsicfeatures of objects (Johansson & Johansson, 2014).

These results seem to suggest that there is a high gazepattern correlation between perception and recall (see alsoHolm & Mäntylä, 2007). To develop empirically reliable the-oretical frameworks in light of this convincing evidence, Bit isessential that memory theorists . . . realize the importance ofthe encoding–retrieval relationship when designing experi-ments and building models of cognition^ (Kent & Lamberts,2008, p. 97).

Co-speech gestures: Simulating simulations

Hand and arm gestures are motor actions that often accompa-ny speech and are intertwined with spoken content (Kelly,Manning, & Rodak, 2008; Krauss, 1998; McNeill, 1992).Several authors have considered gestures essentially asplaying a pivotal communicative role (e.g., Kelly, Barr,Church, & Lynch, 1999): depending upon the context inwhich they are used, gestures accompanying speech can helpinterlocutors by adding information or disambiguating inten-tions. It is no coincidence that speakers’ motivation to com-municate affects the size of the gestures they produce: when

people are more motivated to communicate, they tend to per-form larger gestures (Hostetter, Alibali, & Shrager, 2011).Similarly, individuals who are more empathic (and thus moremotivated to communicate something clearly to their interloc-utors) tend to produce more gestures in order to facilitate com-munication between the speaker and the listener (Chu, Meyer,Foulkes, & Kita, 2014). Indeed, listeners automatically incor-porate the information coming from the speaker’s gesture intheir mental representation of the communicative message,thereby facilitating comprehension and learning (e.g., Ping,Goldin-Meadow, & Beilock, 2014).

However, the results of several studies seem to suggest thatthe role of gestures is not exclusively communicative.Humans also tend to gesticulate in contexts in which theirinterlocutors are not able to observe their gestures (forinstance, during a phone call; Rimé, 1982). Congenitally blindpeople use gestures too, even when they are knowingly speak-ing to a blind listener (see, e.g., Iverson & Goldin-Meadow,1998). Moreover, arm gestures actually have other roles inaddition to that of communication (Morsella & Krauss,2004): indirectly, they facilitate the maintenance of spatialrepresentations in working memory, and directly, they acti-vate, through feedback from effectors or motor commands,the sensorimotor information that is part of the mental repre-sentations. According to these claims, in a study by Wesp,Hesse, Keutmann, and Wheaton (2001), the participants wereasked to describe a painting either from memory or with itvisually present: the participants gestured more often whendescriptions were made from memory compared with whenthe spatial information was visually available. The authorsconcluded that gestures helped the participants by sustainingspatial and motor information associated with the mental rep-resentations stored in memory (Wesp et al., 2001).

Many results reported in the literature on gestures couldeasily be explained from the so-called gestures as simulatedaction (GSA) perspective, according to which gestures arisefrom cognitive processes engaged in simulating perceptualand motor states (Hostetter & Alibali, 2008, 2019). In fact,recent findings have suggested that gestures may be consid-ered as a type of simulated action that arises when motoractivation due to mental simulation processes exceeds a cer-tain threshold (Hostetter & Alibali, 2008; Kelly, Bryne, &Holler, 2011). Put more plainly, by simulating action, gesturesreflect sensorimotor mental simulations: they are externalplaceholders for internal processes. As a point of fact, themotor activation that results from the embodied simulationactivation exceeding the threshold is a gesture. In line withthis theoretical framework, the frequency of gestures duringthe verbal description of images seems to be influenced by theparticipants’ physical experiences with the stimuli portrayedin those images (Hostetter & Alibali, 2010): speakers gestureat a higher rate when they have specific motor experience withthe information they are describing compared with when they

Psychon Bull Rev (2019) 26:1747–1766 1751

do not. Gestures thus depend on the previous motor actionsperformed by participants: they represent the external sign ofthe internal motor simulation affected by the previous experi-ences. Further, several studies have detected that the form of agesture roughly mirrors the form of the underlying sensorimo-tor mental simulation: during the explanation of a previouslifting task, which could be a physical lifting action or usinga computer mouse, speakers produced more pronounced arcsin their gesture when they had experience of physically liftingobjects (Cook & Tanenhaus, 2009). Thus, the form of gesturepresumably depends on the nature of the underlying sensori-motor simulation.

In addition, many studies in recent decades have highlight-ed that gestures can also be considered as a component able toaffect and shape mental simulations. In this perspective, ges-tures can both reflect and trigger a Bsensorimotor simulation^(Dijkstra & Post, 2015), thereby inducing a beneficial effect interms of learning and memory. Although they are not able todirectly change the world, they are able to change and affectour mnestic processes (Madan & Singhal, 2012), both atencoding and retrieval. Because a stimulus is stored in themotor pathways that were involved when this stimulus wasinitially processed, gestures at encoding can thus enforce suchmotor pathways, and gestures at recall can facilitate theirreactivation.

Supporting this notion, a number of studies have found thatmemory is enhanced in participants who accompany the itemsto be recalled with gestures at encoding and in participantswho observe a speaker who gestures while uttering the items(e.g., Cutica & Bucciarelli, 2008; Cutica, Ianì, & Bucciarelli,2014). Cook, Yip, and Goldin-Meadow (2010) presented par-ticipants with a series of short vignettes asking them to givedetailed descriptions. The vignettes were then classified aseither eliciting gestures during their description or not. Theparticipants were given surprise free recall tasks after a briefdelay, and after a 3-week delay. No matter how long the delayspan, recall rates were higher for vignettes associated withgesturing when first described. In a subsequent experiment,enhanced memory performance was found even when partic-ipants were explicitly instructed to either gesture or not, ratherthan being allowed to gesture spontaneously (Cook et al.,2010).

Because mental simulations appear to support memory per-formance during retrieval and given that gestures can facilitatesuch mental processes, gestures could also be an effective wayfor improving memory at recall. In fact, a series of studieshave revealed that, during the retrieval phase, verbal memoryreports by both children and adults benefit from spontaneousgestures; in particular, children told to gesture when trying torecall an event remembered more about the event than chil-dren prevented from gesturing did (Stevanoni & Salmon,2005). In a recent study (Ianì, Cutica, & Bucciarelli, 2016),individuals who had constructed an articulated mental

simulation of a given text were more likely to accompanycorrect recollections with simultaneous gestures, comparedwith individuals who had constructed a less articulated modelof the text; vice versa, individuals who had constructed a lessarticulated model were more likely to accompany correct rec-ollections with anticipatory gestures (gestures for which thepreparatory phase starts before the word they accompany). Itis plausible that good learners have less need to produce an-ticipatory gestures to help them organize their thoughts. Onthe other hand, poor learners need to trigger their mental rep-resentation by using gestures. Similarly, Morrel-Samuels andKrauss (1992) investigated the relationship between the famil-iarity of a given imagine (i.e., the accessibility in memory) andthe synchronicity of gestures during a task involving narrativedescriptions of 13 photographs; again, the more familiar theimage (and thus the associated word), the smaller the asyn-chrony. Vice versa, the less familiar the image, the more ges-ture onset preceded voice onset. A possible interpretation ofthis pattern of findings is that less familiarity with a givencontent is usually accompanied by the presence of anticipatorygestures that are able to play a self-structuring role of theinformation when there is the need to compensate for lessaccessibility in memory: gestures are part of the actual processof thinking (Clark, 2007).

Body postures and movements

Dijkstra, Kaschak, and Zwaan (2007) devised an experimentto assess how congruent body posture might facilitate retrievalof autobiographical memories. Autobiographical memoriescan also be considered as a form of sensorimotor simulation,an embodied model of the original event through which peo-ple relive the same visual, kinesthetic, spatial, and affect in-formation of a given past experience. The authors argued thatif a certain body position was assumed during a given pastexperience, the retrieval of that very same experience shouldbe facilitated if the original body posture was reassumed com-pared with when an incongruent posture was assumed instead.That is exactly what they found. They asked the participants toretrieve autobiographical memories of specific events in thepast while holding several different body positions that couldbe either congruent (e.g., staying lying down on a reclinerwhile remembering the last dentist visit) or incongruent(e.g., staying lying down on a recliner while rememberingthe last football match) with the original one. The results dem-onstrated that response times were shorter in the congruentcondition compared with the incongruent one. Free recallsafter two weeks exhibited a similar effect: the participantsretrieved memories from the first run better in the congruentcondition than in the incongruent one. To conclude, having amemory-congruent body position appears to help participantsgain access to their memories (Dijkstra et al., 2007). Similarbody influences on memory have also been recently detected

Psychon Bull Rev (2019) 26:1747–17661752

in other cognitive areas. For instance, a recent study byMorse,Benitez, Belpaeme, Cangelosi, and Smith (2015) detected arole of the body posture in word learning.

Body movements seem thus to facilitate or hinder the ac-cess of past experiences to memory recall by activating orinhibiting relevant sensorimotor aspects of the experience. Afurther question has arisen: does body manipulation, besideshelping the process of recalling information in terms of reac-tion times (how information is retrieved) influence what isremembered?

Casasanto and Dijkstra (2010) asked whether simple motoractions might influence the quality of emotional memoriesand consequently what people choose to remember. Lakoffand Johnson (1999) had already observed how, when peopletalk about their emotions, they usually use linguistic expres-sions that are related to upward movement when referring to apositive emotional value (e.g., BMy spirits soared^) and, con-versely, if the emotional tonality is negative, they make use ofexpressions that are related to downward movement (e.g.,BI’m feeling low^): metaphors are grounded in embodied rep-resentations. Thus, upward actions are associated with posi-tive notions (good, virtue, happiness, etc.), and downwardactions with negative ones (bad, sadness, pain, etc.).Casasanto and Dijkstra (2010) investigated whether these as-sociations might also have an effect in memory processes, bytesting whether body movements were able to affect the re-trieval of specific emotional events. Specifically, the partici-pants in their study were asked to retell autobiographicalmemories with either positive or negative valence while mov-ing marbles upward or downward: retrieval was faster whenthe secondary motor task was congruent with the valence ofthe memory (i.e., moving marbles upward for positivevalenced memories and downward for negative ones). In asubsequent and crucial experiment, they did not prompt theparticipants to retell positive or negative memories, and foundthat they retrieved more positive memories when asked tomove the marbles up, and more negative memories whenasked to move them down. These results led the two authorsto conclude that there is a direct and causal link between actionand emotion, that positive and negative life experiences areassociated with schematic movement representations, and,most importantly, that body movements can also affect whatwe remember (Casasanto & Dijkstra, 2010).

Seno, Kawabe, Ito, and Sunaga (2013) extended these find-ings by showing that what is crucial for the modulation effectof emotional valence on recollected memories is self-motionand not visual motion per se. Participants underwent illusionsof self-motion (i.e., vection) by viewing upward and down-ward grating motion stimuli. Usually, observers illusorily per-ceive self-motion in the direction opposite to the observedmotion stimuli; thus, vection was supposed to help disentan-gle the effect of visual motion from the effect of self-motion.Indeed, the participants recollected positive episodes more

often while perceiving upward vection. Further, no modula-tion of emotional valence was detected when grating motionwas reduced to the point of not inducing any illusion ofvection. It could therefore be inferred that vertical vection, thatis, illusory self-motion perception, can modulate humanmood. Väljamäe and Seno (2016) further examined the pos-sibility by testing memory recognition using positive, nega-tive, and neutral emotional images with high and low arousallevels. Those images were remembered accidentally while theparticipants performed visual dummy tasks, and were present-ed again later, together with novel images, during verticalvection-inducing or neutral visual stimuli. The results showedthat downward vection facilitated the recognition of negativeimages and inhibited the recognition of positive ones. Thesefindings on the modulation of incidental memory tasks pro-vide additional evidence for vection influence on cognitiveand emotional processing (Väljamäe & Seno, 2016).

Interference effects While we are engaged in a sensorimotormental simulation, performing a different secondary actionthat involves the same sensorimotor resources needed for themental simulation (the action must be different from the sim-ulated action, and thus incongruent with the mental simula-tion) can have an interference effect on the mnestic processes,both at encoding and recall.

For instance, Ping et al. (2014) suggested that performingmovements that are incongruent with mental simulations thatare active during the formation of mental representations caninterfere with the formation of memory traces. In Experiment1, the participants had to carefully observe a series of videos inwhich an actor uttered a series of sentences while performing aspecific gesture. Immediately after each video, the participantssaw a picture of an object that could be in a position that wascongruent or incongruent with the gesture observed in thevideo (for instance, a nail in a vertical or horizontal position).Their task was to respond Byes^ if the name of the object in thepicture was mentioned in the sentence and Bno^ if it was not(filler trials). When the object in the picture was in a congruentposition (i.e., when the information conveyed through thegesture matched the information in the picture), the partici-pants were faster at responding correctly compared with whenthe object in the picture was in an incongruent position. Thismight demonstrate that participants automatically incorporat-ed the information coming from the actor’s hand gestures intotheir mental representation of the speaker’s message. InExperiment 2, Ping et al. (2014) investigated whether the sen-sorimotor simulation triggered by the gesture involved thelistener’s motor system. The participants performed the sametask as in Experiment 1, with the exception that they wereinvited to engage in a secondary motor task during the obser-vation phase, which was to be performed either with theirarms and hands (i.e., the same effectors used by the actor inthe video) or with their legs and feet (i.e., different effectors

Psychon Bull Rev (2019) 26:1747–1766 1753

from those used by the actor). The rationale of the dual taskwas to discover whether involving the participants’ motorsystem by asking them to move their arms and hands wouldhinder the creation of mental simulation and result in disap-pearance of the congruence effect. Consistent with thesomatotopic organization of the premotor activation duringaction observation, such interference should be specific tomotor resources controlling the effectors used in producinggestures, arms and hands in the case at hand. To test thisprediction, half of the participants in Experiment 2 were askedto plan and perform movements with their arms and hands,whereas the other half were asked to use their legs and feetinstead. The results revealed that the participants in the armmovements condition did not show the congruent effect in thepicture judgment task because they were prevented from cre-ating the mental simulation originally associated with thestimuli. In contrast, the congruence effect persisted when theparticipants were asked to move their legs and feet: theyresponded to pictures that were congruent with the speaker’sgestures more quickly than to those that were incongruent.Crucially for memory conceptualization, Ianì and Bucciarelli(2018) highlighted that the advantage of observing gestures onmemory for action phrases disappeared when at recall thelisteners moved the same motor effectors as those moved bythe speaker at encoding (i.e., their arms and hands), but waspresent when the listeners moved different motor effectors(i.e., their legs and feet). The results seem to suggest that thelistener’s motor system is involved both during the encodingof actions and at their subsequent recall during retrieval.

A similar effect of a secondary motor task involving thesame effector used to simulate the relevant action was ob-served by Yang, Gallo, and Beilock (2009). Past studies haveillustrated that the perception of letters automatically activatesa typing action motor program in skilled typists, so the authorsset out to explore whether the fluency arising from the motorsystem could affect recognition memory. The results showedthat expert typists made more false recognition errors for letterdyads that were easier and more fluent to type than fornonfluent dyads. To generalize, memory appears to be influ-enced by covert simulation of actions (e.g., typing) associatedwith the items being judged (e.g., letter dyads) to such anextent that the concurrent reactivation of motor programsleads to false recognitions. Vitally, a second experiment byYang et al. (2009) showed that such effects disappeared whenparticipants were asked to move the fingers they would use totype the dyads in an unrelated but concurrent secondary motortask, whereas it persisted when participants were asked toperform a noninterfering motor task. In other words, the par-ticipants who had in their mind a motor plan involving thefingers they would use to type the presented dyads could notsimulate the action usually needed to type them because therelevant cognitive resources were involved in the secondarymotor task.

Bodily expressions of emotions

Nonverbal emotional components like facial expressionsmight be critically implicated in the process wherebyrepresenting, during retrieval, experimental conditions andcues that were already present at encoding has an impact onmemory. The study by Riskind (1983) was amongst the firston the topic and concerned the effect of facial expressions. Ittested the congruence hypothesis about the priming effects offacial and body posture patterns onmemory retrieval, suggest-ing that an individual should be more likely to recollect pleas-ant experiences when smiling and assuming an expansivephysical posture. The hypothesis predicted that the accessibil-ity of pleasant experiences from one’s own life history wouldincrease when nonverbal expressive patterns were positive invalence rather than negative. Likewise, the accessibility ofunpleasant experiences from one’s life history would increasewhen nonverbal expressive patterns were negative in valence,such as when an individual frowns and/or assumes a slumpedposture (see also Laird,Wagener, Halal, & Szegda, 1982). Thelatencies with which the participants recalled pleasant andunpleasant life experience memories supported the congru-ence hypothesis: recalling an autobiographical memory witha smile and in an upright position improved access to pleasantmemories.

These pioneering studies have revealed that positive/negative nonverbal expressions can affect memory retrieval.The findings reflect the importance of both sensory and motorfunctions and affective valence in memory retrieval, and areconsistent with a view that conceptualizes cognitive processesas an integral part of the sensorimotor environment in whichthey are embedded. Bodily experience is more than just anemotional state exceeding a given threshold—it is part of thatemotional state; not surprisingly, Bin depressive patients thebody becomes conspicuous, heavy and solid^ (Zatti & Zarbo,2015‚ p. 2-3). According to the so-called somatic marker hy-pothesis (Damasio, 1996), an emotion is a change in the rep-resentation of body state, and the results of emotions are pri-marily represented in the brain in the form of Btransient chang-es in the activity pattern of somatosensory structures . . . thesomatic states^ (Damasio, 1996, p. 1414). Empirical findingsalso highlight the role of body posture and body image in theexpression and communication of mood. A study by Canales,Cordás, Fiquer, Cavalcante, and Moreno (2010) revealed thatduring depressive episodes, patients showed a series of pos-tural changes such as increased head flexion, increased tho-racic kyphosis, a trend toward left pelvic retroversion, andabduction of the left scapula. These alterations were not foundduring the remission phase. Memories, with their cognitive,affective, and sensorimotor information, then influence bodyexpression.

However, it is possible to deduce from the studies men-tioned above that the manipulation of body expressions may

Psychon Bull Rev (2019) 26:1747–17661754

in turn influence mood. On a clinical level, it means investi-gating, for example, whether people with certain disorderssuch as depression or other mood disorders, which tend tohave negative thinking and a greater reenactment of negativememories, might benefit from the manipulation of body pos-ture and movements.

Experimental research has demonstrated this bidirectionallink between nonverbal behavior and human feeling.Assuming a stooped rather than a straight sitting posturemay lead people to develop an increased level of helplessness(Riskind & Gotay, 1982). Similarly, inducing an upright orslumped posture in the laboratory setting can influence theamount of pride people express (Stepper & Strack, 1993):success at achieving an outcome led to greater feelings ofpride if the outcome was received in an upright position ratherthan in a slumped posture.

Recently, several experiments demonstrated that expansivecompared with contractive nonverbal displays (high-power-pose or low-power-pose condition) produced subjective feel-ings of power and increased risk tolerance (e.g., Carney,Cuddy, & Yap, 2010). In relation to memory processes, ithas been demonstrated that sitting in a slumped posture whileimagining positive or negative events associated with positiveor depression-related words leads people to refer to more neg-ative than positive words in an incidental free recall task:relatively minor changes in the motoric system can affectemotional memory retrieval (Michalak, Mischnat, &Teismann, 2014). It is not just slumped sitting that can inducenegative emotional processing but also slumped walking orgenerally walking with a forward-leaning posture that cannegatively affect memory. Using an unobtrusive biofeedbacktechnique, Michalak, Rohde, and Troje (2015) manipulatedparticipants to change their walking patterns to either reflectthe characteristics of depressed patients or a particularly happywalking style. During walking, the participants first encodedand later recalled a series of emotionally loaded terms. Thedifference between positive and negative words retrieved atrecall was lower in the participants who had adopted a de-pressed walking style compared with those who had walkedas if they were happy: walking style affects memory functions(Michalak et al., 2015).

Because straight posture is related to feelings of power andpride (see, e.g., Oosterwijk, Rotteveel, Fischer, & Hess,2009), an upright body posture might make it easier for peopleto recover from negative feelings. Conversely, recovery fromnegative mood might be impaired when people assume astooped rather than a straight sitting posture. This wasestablished by a study by Veenstra, Schneider, and Koole(2017), who asked participants to imagine a negative or aneutral situation; later on, they manipulated the participants’body posture, telling them to adopt a stooped, straight, or self-chosen body posture. After which, they asked them to list theirthoughts in order to support spontaneous mood regulation.

The results revealed that a stooped posture, compared withcontrol and straight postures, prevented effective mood recov-ery in the negative imagination condition and increased neg-ative affect in the neutral imagination condition. In addition,overall stooped posture induced more negative thoughts com-pared with straight or control postures (Veenstra et al., 2017).Body posture can play an important role in recovering fromnegative mood. Arminjon et al. (2015) asked participants tofirst read a sad story in order to induce a negative emotionalmemory, and then to self-rate their emotions and memoriesabout the text. One day later, some of the participants wereasked to assume a predetermined facial feedback (smiling)while reactivating their memory of the sad story. The partici-pants were once again asked to fill in emotional and memoryquestionnaires about the text. The results showed that partic-ipants who had smiled during memory reactivation rerated thetext less negatively than control participants who were notasked to assume any specific facial expression. In short, ma-nipulating somatic states can also modify the emotional as-pects accompanying any given memories; thus, the body andits morphology also appear to play an instrumental role inemotional information processing (Arminjon et al., 2015). Inother words, there is a reciprocal relationship between thebodily expressions of emotion and the way in which emotion-al information are experienced.

In this view, people’s bodily states can be considered as anopportunity for emotion regulation because it is no longer seenas a matter of the higher cognitive mechanisms controllinglower level systems—rather, it is conceptualized as an activityrelying on (and emerging from) a well-established interplaybetween the mind and the body (that is to say, the emotionalbrain, namely, the person; Damasio, 1994; LeDoux, 1996;Pessoa, 2017).

Depressed patients might be in a Btrap,^ in which negativethinking leads them to adopt a specific body expression,which, in turn, reactivates, in most cases, the sensorimotorcircuits that were activated in the perceptual coding of nega-tive events. Changing posture might break this vicious circle:it is a simple, highly acceptable, and low-risk interventionthat, applied together with other interventions, might counterthe depressive symptomatology. Wilkes, Kydd, Sagar, andBroadbent (2017) directly investigated this hypothesis. Theytested whether changing posture (upright vs. usual posture)could reduce negative affect and fatigue in people with mildto moderate depression undergoing a stressful task. Theyfound that postural manipulation significantly increasedhigh-arousal positive affect and fatigue compared with usualposture. Moreover, on the Trier Social Stress Test speech task(during which participants are asked to deliver a free speech infront of an audience), the upright group spoke significantlymore words than the usual posture group. Further, uprightshoulder angle was associated with lower negative affect andlower anxiety across both groups. Manipulations of body

Psychon Bull Rev (2019) 26:1747–1766 1755

posture can also be useful for counteracting fatigue in sleep-deprived individuals (Caldwell, Prazinko, & Caldwell, 2003).

Evidence against the SMM

Evidence in favor of the SMM is not uncontroversial. Readersshould not interpret the described effects as always well-replicable and accepted. A series of conflicting studies seemto cast doubts on the reliability of such embodiment effects.For instance, a recent replication attempt (Wagenmakers et al.,2016) of a pioneering study by Strack, Martin, and Stepper(1988) failed to replicate the original results about the role ofbody in emotional processes. Specifically, in order to demon-strate a direct influence of body on emotional processing,Strack et al. (1988) asked participants to hold a pen in theirmouths so that a smile was either facilitated or inhibited. Atthe same time, they were asked to rate the funniness of severalcartoons. Although the participants were not aware of themeaning of the particular muscle contractions (authorsinhibited or facilitated the muscles associated with smilingwithout explicitly requiring subjects to smile), their reportedamusement had been affected by the induced expressions.However, the results of 17 independent direct replication stud-ies (Wagenmakers et al., 2016) were inconsistent with theoriginal pattern of results. Out of 34 Bayes factor analyses(two analyses for each replication attempt; i.e., the defaultBF10 and the replication BFr0), only one provided evidencein favor of the alternative hypothesis.

Besides this failed replication attempt, some studies onshort-term memory show no evidence supporting a role ofthe motor system and sensorimotor simulations in memory.For instance, Pecher (2013) investigated the role of motoraffordances in memory for objects. Participants were askedto observe a series of pictures of manipulable andnonmanipulable objects. Because several studies showed thatperceiving manipulable objects (or their name), automaticallytriggers simulations of interacting with them (e.g., Tucker &Ellis, 2004), as well as motor areas activations (e.g.,Cardellicchio, Sinigaglia, & Costantini, 2011; Hauk,Johnsrude, & Pulvermüller, 2004), such manipulation wasdevised in order to elicit or not to elicit motor activation inthe observers. For each trial (i.e., each picture), after a 5,000-ms retention interval, participants observed a test stimulus andwere asked to decide whether or not it was the very same asthe previous stimulus. Crucially, during the interval betweenthe two stimuli, participants had to perform a motor-interference task, a verbal-interference task, both tasks, or notask. Because only manipulable objects were associated withmotor actions, the authors’ predicted that a motor interferencetask would have been detrimental, particularly for these ob-jects. However, no interaction was found. Similar no-significant results were obtained by Pecher et al. (2013) and

by Quak, Pecher, and Zeelenberg (2014) by using an n-backtask instead of a classic recognition task. Similarly, when par-ticipants had to memorize the name of a series of objects(rather than pictures), either manipulable or not, and wereasked to run a secondary motor task (tapping with hands orfeet), results did not reveal any interference effect (seeZeelenberg & Pecher, 2016).

Recently, Canits, Pecher, and Zeelenberg (2018) have in-vestigated whether motor representations triggered by the pro-cessing of manipulable objects may play a role in long-termmemory. In their experiment, participants were first asked torun a categorization task after a surprise free recall task. In thecategorization task, they had to decide whether a series ofobjects were natural or artificial. Crucially they wereinstructed to respond by grasping either a small cylinder (pre-cision grasp) or a large cylinder (power grasp). Because ob-jects could be large or small (thus eliciting a power or a pre-cision grasp affordance), such manipulation led to four condi-tions, two of which were Bcompatible^ (large objects/powergrasp responses; small objects/precision grasp responses) andtwo of which were Bincompatible^ (large objects/precisiongrasp; small objects/power grasp). Later on, participants wereasked to recall all the objects they had previously seen (such atask was unexpected). Results indicated that at the categoriza-tion task, responses were faster when the affordance triggeredby the object was compatible with the type of grasp requiredby the response. Nevertheless, at the free recall task, authorsdid not find better memory for objects for which the graspaffordance was compatible with the grasp response. They con-cluded that there is no evidence in support of the hypothesisthat motor action plays a role in long-term memory.

However, because these studies have not manipulated thebody movements during the retrieval phase, thus not creatinga strong mismatch between the sensorimotor information in-volved at encoding and retrieval, such results may not repre-sent strong evidence against a role of the motor system inmemory processes. In this regard, it has been found that whatis crucial is the matching of Bmotor context^ at encoding andretrieval (Halvorson, Bushinski, & Hilverman, 2019).Holvorson and colleagues (2019) found when participantsare engaged in an incongruent unrelated motor task both atencoding and recall, they do not suffer any interference effect.Such results do not rule out a motor interpretation, but ratherthat the secondary motor task at encoding may have justchanged the motor representation associated with the stimuli.Interference effects occurred only when at recall the secondarymotor task was different from the encoding phase task. It isworth noting that a sensorimotor mismatch can also be in-duced during a recognition task by showing stimuli triggeringmotor information incongruent with those involved atencoding (e.g., Ping et al., 2014). In other words, becausethe matching of the encoding and recall motor context matters,it could be that in the study by Canits et al. (2018), producing a

Psychon Bull Rev (2019) 26:1747–17661756

grasping movement incongruent with the object’s affordancedid not cancel the observer’s motor representations, but ratherjust biased them. In a free recall task, during which no sec-ondary motor instructions or incongruent stimuli were given,it is likely that participants had no difficulty in reactivating thesame motor representations.

More in general, all these studies have investigated a spe-cific kind of memory—namely, memory for objects or words.When people observe a given object or a word of an object,they activate relevant information about it, including the pos-sible use of this object and the resulting motor programs.Thus, memory for a given object or a givenword encompassesmultimodal and sensorimotor knowledge gained by humansthrough their personal experiences (Conca & Tettamanti,2018). As pointed out by Rosch (1978), inseparable fromthe concept of a given object are the ways in which humanshabitually use or interact with those objects. For instance, forconcrete objects, such interactions take the form of motormovements: the word chair triggers the action of sitting downon a chair and a sequence of typical body and muscle move-ments that are inseparable from the nature of the attributes ofchairs—legs, seat, back, and so forth. At the same time, theyactivate much more information than the motor ones (i.e., thepropositional and phonological representations or thesemantic information; for a discussion see, e.g., Mahon &Hickok, 2016). Further, as pointed out by Mahon andCaramazza (2008), the majority of findings revealing a motoractivation during language processing may be interpreted as asecondary and indirect activation, which would occur onlyafter the concept has already been understood. Language com-prehension relies on the propositional representation resultingfrom parsing processes and then on the construction of a ki-nematic mental model of the state of affairs described startingfrom the propositional representation (see also Kintsch, 2005).Instead, when we observe an agent acting in the world, wementally simulate what he or she is doing through anBautomatic, implicit, and non-reflexive simulationmechanism^ (Gallese, 2005, p. 117). This mental simulationis implicit and is triggered by the stimulus itself (the kinemat-ics of the observed movements). Conversely, the mental sim-ulation stemming from language comprehension is indirect inthat its input is the propositional representation resulting fromthe previously parsing processes. In other words, when aphrase is perceived, both the linguistic and simulation systemsare active, but the linguistic system’s activation peaks beforethe simulation system’s activation (see, e.g., Barsalou, Santos,Simmons, & Wilson, 2008). Thus, the sensorimotor simula-tion triggered by language processing is somehow indirect.Consistent with this claim, a recent study by Ianì, Foiadelli,and Bucciarelli (2019) highlighted how simulation from ac-tion observation prevails on simulation from action phraseswhen their effects are contrasted. Specifically, when atencoding phase the action phrases to be remembered were

paired with pictures portraying actions whose kinematicswas incongruent with the implied kinematics of the actionsdescribed in the phrases, memory for phrases was impaired.However, the reverse does not hold: when participants wereasked to remember the pictures portraying actions, their mem-ory was not affected by the presentation of phrasesrepresenting actions whose implied kinematics was incongru-ent with the kinematics of the actions portrayed in the picture.Therefore, the motor simulation stemming from languagecomprehension would seem more mediated compared withmotor simulation triggered by action production or action ob-servation (i.e., the contents of an episodic memory). A possi-ble reason could be that the memory system evolved primarilyto process perceptual and motor aspects of experience(Barsalou, 2008). Therefore, the processing of experience ismore central to human cognition than the processing of wordsof objects. In other words, although the motor componentleads to a richer memory trace, a memory for an object isnot confined solely to it. Different representational formats(e.g., the propositional or phonological ones) can easily com-pensate for the lacking motor information.

Put plainly, motor simulations could play a crucial role inremembering events, and a less pivotal role in rememberingobjects. Remembering episodes means substantially remem-bering actions (performed or observed), which, in most cases,have not been labeled with a propositional format before.

Indeed, in contrast to Pecher’s studies, Pezzulo, Barca,Lamberti-Bocconi, and Borghi (2010) found evidence in favorof a role of motor affordance in memory. In particular, thestimulus to be remembered was not a word or an object butinstead a climbing route. Participants were asked to memorizethree different routes: an easy-climb route, a hard-climb routeand a route impossible to climb. Such routes were indicatedtwice by a trainer on the climbing wall. Participants wereeither novice climbers or expert climbers. After the trainer’sindication, they were asked to perform a brief distractor task,and later on they had to mark on a climbing-wall picture thecorrect sequence of the learned route. Whereas for the impos-sible and easy routes the two groups did not differ, for the hardroute the expert climbers’ group perform significantly betterthan the novice group. These results have been interpreted asbeing in favor of a motor involvement in memory becauseonly expert climbers, the only ones to possess the necessaryskills to actually climb the hard routes, would have been ableto create a sensorimotor simulation of the route.

Further, even some studies about memory for objects orverbs have detected sensorimotor interference effects. For in-stance, Shebani and Pulvermüller (2013) reported that rhyth-mic movements of either the hands or the feet led to a differ-ential impairment of working memory for concordant arm andleg related action words. Specifically, hand/arm movementsimpaired working memory for words used to speak about armactions, whereas foot/leg movements hindered memory for

Psychon Bull Rev (2019) 26:1747–1766 1757

leg-related words (see also Dutriaux & Gyselink, 2016; VanDam, Rueschemeyer, Bekkering, & Lindemann, 2013).

Conclusions and directions for future research

A first, necessary conclusion that can be extracted from theseresults is that memory traces contain detailed sensorimotorinformation—for instance, on the body posture and move-ments the person underwent while encoding a given experi-ence. The very same sensorimotor information involved atencoding is thought to be reactivated during the retrievalphase. Mnemonic traces are not fully amodal mental represen-tations, independent of the body. Rather, they are at least part-ly reenactments of the original bodily and somatic states,which are simulated through the same sensorimotor pathwaysinvolved when the event was encoded. The collection of thesestudies highlights the importance of the motor system inmem-ory retrieval and supports the assumption that memory is foraction, action is for memory (Dijkstra & Zwaan, 2014). Thus,episodic memory, classically considered as forms of declara-tive knowledge (e.g., Tulving, 1995), seems also to containprocedural information. The declarative and procedural mem-ory systems have been intensively studied in humans, and thedemonstration of numerous double dissociations has shownthat the two systems are largely independent of each other(e.g., Eichenbaum&Cohen, 2001). However, from the resultshighlighted in the previous section, the concepts of declarativeand procedural memory appear to have thinner boundaries. Asalready demonstrated, the evidence suggests that the two sys-tems can, to some extent, acquire the same or analogousknowledge (e.g., Ullman, 2004) and mutually interact in anumber of ways (Poldrack & Packard, 2003). Procedural in-formation can trigger declarative information—for instance,the procedural information conveyed by gestures can activatedeclarative knowledge about an event that occurred previous-ly, thereby triggering episodic memory (Ianì & Bucciarelli,2017; Ianì et al., 2016). Further, many studies highlightedhow the motor system and procedural information providedby performing or observing gestures can improve declarativememory (Engelkamp & Zimmer, 1985). Action helps in re-membering verbal information in a procedural way, withoutthe use of intentional encoding strategies (e.g., Earles &Kersten, 2002). At the same time, incongruent procedural in-formation seems able to interfere with memory processes(e.g., Dijkstra et al., 2007).

Because memory is the reenactment of perceptual, motor,and somatic states acquired through experience with theworld, it can be conceptualized as a Bsensorimotor mentalsimulation.^ The notion of Bmental simulation^ has beenwidely used in cognitive psychology. Although it underlies avariety of other cognitive processes, such as mechanical rea-soning (e.g., Hegarty, 2004), deductive and abductive

reasoning (Khemlani, Mackiewicz, Bucciarelli, & Johnson-Laird, 2013), mental imagery (e.g., Moulton & Kosslyn,2009), and empathy (e.g., Niedenthal et al., 2005), it has beenused primarily in relation to social cognition (e.g., DiPellegrino, Fadiga, Fogassi, Gallese, & Rizzolatti, 1992;Keysers & Gazzola, 2007; Rizzolatti, Fadiga, Gallese, &Fogassi, 1996) and motion inference (Carlini, Actis-Grosso,Stucchi, & Pozzo, 2012). Understanding other people’s ac-tions involves the activation of the same areas devoted to theproduction of the same actions both in the agents and in theobservers. Indeed, a network of brain regions (i.e., the mirror-neuron system) is activated both when an action is performedand when it is observed in others (Buccino et al., 2001;Rizzolatti & Craighero, 2004; see, e.g., Prinz, 1990). Severalauthors (e.g., Gallese, 2007) have argued that such activationprovides an internal representation of the observed motor pro-grams that, in absence of overt movements, is usually calledBaction simulation^ (Jeannerod, 2001). As an action is ob-served, the motor system constructs a forward simulation ofthe action in order to predict and anticipate the action goal(Wilson & Knoblich, 2005). What is the relationship betweenthis kind of simulation and those implicated in memory pro-cesses? The literature lacks in differentiating the nature andthe relationship between them. It is likely that these types ofcognitive processes involve qualitatively different kinds ofsimulations and different kinds of phenomenological experi-ences (e.g., Kent & Lamberts, 2008; Moulton & Kosslyn,2004). At the same time, both of these mental processes relyheavily on the motor system, and both contain perceptual andsensorimotor information. Such simulations are usually de-fined as situated, because the situated nature of experiencein the environment is reflected in the situated nature of themental representations underlying simulations (Barsalou,2009). Further, simulations arising from action observationare modulated by the expertise and previous experience ofthe observer (Calvo-Merino, Glaser, Grèzes, Passingham, &Haggard, 2004): neural activation during action observation isgreater when participants are familiar with the observed actioncompared to the neural activation arising from either unusualactions or actions outside the reach of ordinary human motorskills. In addition, when participants are trained in a specificmovement, they are better than untrained participants at visu-ally recognizing similar movements (Casile & Giese, 2006).These latter findings seem to suggest that the effectiveness ofsimulations triggered by impinging stimuli is strictly depen-dent on previous experiences, stored in the sensorimotor sys-tem of memory. Barsalou (2009) proposed an uniqueBsimulator^ process underlying both these simulations: thebrain can be viewed as a coordinated system that generates acontinuous stream of multimodal simulation (Barsalou et al.,2007), which is the reenactment of perceptual, motor and in-trospective states acquired during experience with the world,body, and mind (Barsalou, 2008). Such process has two main

Psychon Bull Rev (2019) 26:1747–17661758

phases: storage in long-termmemory of multimodal states thatarise across the brain’s systems for perception, and partialreenactment of these multimodal states for later representa-tional use, including prediction. In this view, simulation trig-gered by action observation is also substantially a memoryprocess. Thus, we could conceptualize the simulation derivingfrom the observation of the action as a particular case of mem-ory, in which fundamental aspects related to the input stimulusare automatically reactivated. Recently, evidence for a causalrole of those brain regions in action comprehension has beenreported (Michael et al., 2014).

A second corollary conclusion that can be drawn from thereviewed literature is that human body manipulations are ableto affect memory processes by favoring or interfering withthem. In this regard, sensorimotor representations and theirmodality-specific cortical activations seem to be not just anepiphenomenon but rather constitutive elements of cognitiveprocesses. Focusing on the causal influences by which bodymanipulation affects memory, it appears that retrieval of mem-ory traces involves the activation of sensorimotor brain areasand, crucially, that interfering with them by performing a sec-ondary motor task leads to interferences in memory processes.Body manipulation results in changes in memory perfor-mance. It follows that memory is not fully independent fromthe body. In other words, the body, along with its sensorimotorstates, is at least partly a constitutive and inseparable part ofthe cognitive processes involved in the encoding and retrievalof mnestic traces.

Toward a more circumstantial outline

Having established that the body is involved in memory pro-cessing, the question that arises is to what extent does memorydepend on these sensorimotor processes? To better delineatethe solution of this issue, one needs to outline the exact degreeby which sensorimotor involvement determines the efficiencyof memory processes. Although it is clear how the bodyshapes memory traces, there is a risk of overstating the roleof motor processes in memory (see, e.g., Mahon &Caramazza, 2008). In this regard, the first considerationtouches upon the more obvious and evident limit of bodymanipulations: none of these, even the most incongruous withthe original experience or the most challenging for our motorsystem, are able to completely erase a memory trace, neitheran episodic nor a semantic one. For example, being relaxed orlying down, such as at a dental visit, is usual, it facilitates therecovery of the memory trace concerning my last dental visit,but it does not cancel the memory itself. Similarly, keepingarms and the hands behind our back does not imply the inca-pacity to catch the meaning of graspable object (e.g., a mug).In fact, several studies seem to suggest that patients with im-paired object-directed reaching and grasping due to motorareas lesions show intact object identification (e.g., Marotta,

Behrmann, & Goodale, 1997). More in general, patients withsensorimotor impairments due to lesions in motor areas do notshow great impairment in action understanding or action re-membering (for a discussion, see, e.g., Hickok, 2014; Mahon,2015). In other words, a nucleus of memory that is a Bpurelycognitive activity^ seems to exist (Goldinger et al., 2016).

Second, the facilitation provided by a body position com-patible with the content of the stimulus materials to be remem-bered seems to be only the result of a greater availability ofprocessing resources. In fact, most of the abovementionedstudies evaluated these congruence/incongruence effects interms of access speed or access facilitation (e.g., Casasanto& Dijkstra, 2010; Dijkstra et al., 2007; Ianì & Bucciarelli,2018), in terms of changes in the sensation of familiarity ofa given motor sequence (e.g., Yang et al., 2009), or in terms ofwhich manipulating somatic states can modify the emotionalaspects accompanying memories (e.g., Arminjon et al., 2015;Veenstra et al., 2017). In all these examples, manipulations ofthe body seem to lead to a modulation of the memory process,but not to a suppression thereof, nor a changing in memoryrepresentations. Body manipulations are not able to resetmemory processes to zero. They seem to be able to influencehow humans feel or process information, or to change theaccessibility of concepts associated with a given bodily state.This would seem to suggest how—although somatosensoryelements come heavily into play and, in part, constitute theontology of a memory—the latter is not confined solely tothem; there is a nucleus in which the memory processes re-main independent from somatosensory ones. In other words, itseems that sensory and motor information does not exhaustmemory contents (Meteyard et al., 2012). Further, in order todescribe the consequences of body manipulations in moredetail, three possible areas need to be considered.

Effect of body manipulations on the quality of memory tracesIn order to outline the extent to which the body affects mem-ory, one should differentiate how sensorimotor interferencesoperate on the retrieval processes rather than on the represen-tation of the memory traces. In other words, the first questionstill under debate concerns whether the congruence/incongruence effects can also affect the quality of a memorytrace and, more precisely, the accuracy with which some de-tails are reevoked, especially those that convey motor andspatial information. From the reviewed literature, such issueonly appears to have been addressed in eye-movement exper-imental settings: when, at recall, the participants were asked tomaintain fixation away from the original location of the stim-ulus shape, so as to interfere with the gaze reenactment pro-cess, recall accuracy decreased (e.g., Laeng et al., 2014).Oculomotor manipulations resulted in measurable costs inthe quality of memory. Would a similar cost emerge usingbody and movement manipulations? For instance, if duringthe retrieval of a previously performed action participants

Psychon Bull Rev (2019) 26:1747–1766 1759

are invited to perform a similar but not totally identical action(e.g., grasping a bottle in two different contact points), wouldthey be less accurate in reporting the original action (e.g., isthe estimation of where I grasped the bottle influenced by theaction performed at recall)? In other words, does performing asecondary task (involving the same sensorimotor resourcesinvolved in the mental simulation during retrieval) result indecreased precision with which participants report spatial andmotor information?

To sum up, from the quoted literature, evidence for a con-sequence on the accuracy of the memory trace arise only ineye-movement studies. In the other cases, at the current state,sensorimotor information seems able to shape cognitive pro-cessing rather than mental representation (i.e., the access to agiven memory trace). Because oculomotor manipulationsseem able to interfere with the quality of memory, other kindsof body manipulations might also result in similar costs. Morerecently, the results of two studies on co-speech gestures seemto suggest that body manipulations at the recall phase mayaffect the way by which participantsmanipulate a given mem-ory trace. In a study by Kamermans et al. (2019), participantslearned a bistable figure through touch for 30 seconds. Lateron, they were introduced to the idea of bistability and wereasked to reinterpret the figure—that is, to find the alternativeinterpretation of the learned figure by mentally simulate arotation. During the test phase, participants were divided inthree groups: in the gesture condition, participants were askedtomove their hands and arms as if actually having the figure inthe their hands. In the no gesture condition, participants wereasked to keep their hands still on the table. In the manualinterference condition, they were asked to drum their fingerswith both hands continuously on the table. Results revealedthat participants who were engaged in the secondary motortask were less successful in reinterpreting the figure, therebyindicating that by loading up the motor system it is possible tointerfere with the ability to mentally rotate the figure stored inmemory (Kamermans et al., 2019). Likewise, Nathan andMartinez (2015) have found that restricting gestures produc-tion during the test phase, by asking participants to tap withone hand a particular spatial pattern, results in a significantreduction in the ability to make inferences starting from thelearned material. However, these results seem to suggest thatbody manipulations may affect the way by which participantsmanipulate a given memory trace, rather than the memorytrace itself. In other words, they speak in favor of a causal roleof the motor system in handling a given memory trace.

Remembering another person’s actions through sensorimo-tor simulation A second open question concerns the specificmemory of other people's actions. Access and retrieval of agiven episodic memory can be facilitated (and memory con-tent affected) when specific action patterns are executed fa-voring the sensorimotor simulation of the event that initially

caused the memory. However, this effect has always beenobserved when considering actions performed by the subject.There are situations in which the gist of the event/memory isan action by another person, such as the kind of memoriestypically involved in eyewitness reports (e.g., rememberinghaving been mugged). This is, for example, the usual situationfor witness context, in which the memory to be recollectedfocuses on another person’s actions. Do similar mental simu-lations play the same facilitating role during the retrieval ofmemory traces for actions performed by another person asthey do in the case of actions performed by the subjects them-selves? And does the manipulation of body movements/posture affect the process?

There is some preliminary evidence in favor of this hypoth-esis. On the one hand, as previously mentioned, the literatureon action observation suggests that evenwhen a personwho isstaying still observes an action by another individual, theirmotor system is automatically activated (see, e.g., Buccinoet al., 2001). According to the reactivation hypothesis, thatactivation might play a pivotal role in recalling the event.Further, although using a specific experimental setting involv-ing pantomime gestures rather than real, everyday actions,Ianì and Bucciarelli (2018) found that memory for sentencesinvolving actions previously performed by an actor is greaterwhen the actor moves compared with when he stays still whilethe sentence is uttered. On the other hand, involving the lis-tener’s motor system during gesture observation cancelled thebeneficial effect only when the motor task involved the sameeffectors used by the speaker, as the beneficial effect indeedpersisted when the motor task involved different effectors.Now, supposing that these results were to go beyond thisspecific experimental setting, we could also expect themotor-based information to be used to simulate and recon-struct another person’s actions. But, then, a secondary motortask at recall would interfere with such memories, too.

Effects of body manipulations on sense of recollectionA thirdarea in which it is likely that sensorimotor interference para-digms might affect memory processes is that of the phenom-enal characteristics associated with memory. Since the senso-rimotor simulations convey mostly perceptual and sensorialinformation linked to the original event, it is likely that theycontribute to the sensation of Breliving^ the event itself. Whenwe remember a given memory trace, we engage in a mentalsimulation rich in details because not only visual informationbut also relevant motor states come into play. Memory doesnot keep trace of experience as if it were an abstract idea.Nyberg (2002) claimed that both perceptual and sensorimotorinformation is part of the memory trace and that brain regionsstoring such information are spontaneously reactivated at re-call in order to reactivate the same perceptual and sensorimo-tor feelings. In other words, the sensorimotor simulationmightsupport memory in terms of perceptual characteristics, motor

Psychon Bull Rev (2019) 26:1747–17661760

states, emotional richness, and the feeling of Breliving,^ thusin terms of its phenomenological characteristics (Mazzoni,Scoboria, & Harvey, 2010). For instance, when rememberingan event in the past, wemay Bsee^ in our mind the place wherethe event took place as well as the objects and the people whowere present, and relive the thoughts and the feelings we hadduring that event. All these details lead to the subjective ex-perience of mentally reliving a past event (D’Argembeau &Van der Linden, 2006), a feeling of Bwarmth and intimacy^(James, 1890/1950). Therefore, involving the motor systemshould result in changing the motor details reactivated at re-call, thereby modifying the phenomenological characteristicsassociated with recollection. In this respect, a growing body ofliterature has revealed that such a Bremembering^ system canbe dissociated from the Bbelieving^ one: the sensation of rec-ollection relies on different cognitive mechanisms with re-spect to the framework of beliefs about such an event(Mazzoni et al., 2010). In an anecdotal report, Jean Piaget(1951) described how, for much of his life, he had a detailedmemory of having almost been abducted in a park when hewas 2 years old, with his nurse, in the stroller. Piaget describedhis memory in a vivid and detailed way. Many years later,Piaget’s former nurse confessed that she had completely fab-ricated the story, and so he stopped believing it. Crucially,Piaget could not stop Bremembering^ it with a strong senseof recollection, even when he was certain that the event hadnot, in fact, occurred. What Piaget described is a nonbelievedmemory (NBM), a counterintuitive phenomenon in whichpeople report a vivid memory of an autobiographical eventalthough they believe the event did not occur. Despite thenewfound knowledge, Piaget remained able to Bremember^the scene, which continued to feel very much like a true mem-ory. This is a kind of Bpure memory^ for an event not accom-panied by the belief it really occurred: the sensation to be ableto mentally travel back in time and reexperience events withvivid perceptual and sensorimotor information, even thoughthe belief about its veracity has been lost. Indeed, mental rep-resentations tend to be labeled as Bmemories^ or asBrecollected^ when they are associated with high levels ofvivid perceptual and contextual information, thereby inducinga strong sense of reexperiencing the past (Moskovitch, 2012).

Scoboria, Mazzoni, Kirsch, and Relyea (2004) were amongthe first scholars to discover the existence of this kind ofBpure^ memory. The participants in their study were askedto fill out a series of questionnaires about memory, belief,and plausibility in relation to 10 events that might have oc-curred in their childhood. Although for the 96% of items,memory implied belief and belief implied plausibility (accord-ing to the so-called nested constructs hypothesis), the authorsfound that in a small percentage of cases (4%), the memoryrating exceeded the belief rating. These results indicate thatmemories are not always nested within beliefs (Scoboria et al.,2004). People can maintain a memory trace for an event

despite the autobiographical belief being lost. Subsequentstudies confirmed that such dissociation between memoryand belief can occur in natural contexts. In particular,Mazzoni et al. (2010) examined the frequency with whichNBMs usually occur, the factors that lead individuals to with-draw their belief that the events portrayed in these memoriesoccurred, and the phenomenological features of these memo-ries. Approximately 20% of a large sample of participantsreported having at least one NBM in their life. Those whoreported having at least one NBM were then asked to fill outa memory inventory assessing the reasons why these memo-ries were no longer believed and the characteristics of thesemental experiences. The participants reported having stoppedbelieving in their memories either because of a negative socialfeedback or because they perceived the event as being nomore plausible (for a recent and more detailed analysis, seeScoboria, Boucher, & Mazzoni, 2015). Further, the resultsrevealed that NBMs are subjectively experienced much likenormal believed memories. Although NBMs were rated lesspersonally important and less connected to other memories,the phenomenological ratings were similar to those of be-lieved memories, sharing with them many phenomenologicalfeatures such as Bstrong sense of recollection^ (see also, e.g.,Clark, Nash, Fincham, &Mazzoni, 2012). These phenomeno-logical features provide a Bmemory-like^ quality to mentalexperiences, regardless of whether or not these mental repre-sentations were believed. These memories enabled partici-pants Bto travel back in time mentally and relive the event,reexperience the same intense emotions, clearly recall visualand other perceptual details, and form a clear idea of whereobjects and people were in the original event^ (Mazzoni et al.,2010, p. 1339).

NBMs can be also induced in laboratory. The most com-mon technique to create NBMs is to induce false memoriesand then to provide a social feedbackwhich disconfirms them.Inducing a false memory in a participant and then telling themthe false event did not actually occur is very likely to under-mine the belief that it happened. But do the mental image andthe feeling of Breliving^ the event disappear as well? Severalfalse memories p