Cognition 110 (2009) 265–272

Contents lists available at ScienceDirect

Cognition

journal homepage: www.elsevier .com/locate /COGNIT

Brief article

Categorizing moving objects into film genres: The effect of animacyattribution, emotional response, and the deviation from non-fiction

Valentijn T. Visch a,*, Ed S. Tan b

a Industrial Design, Technical University Delft., Department of Arts, Vrije Universiteit Amsterdam, The Netherlandsb Amsterdam School of Communications Research, University of Amsterdam, The Netherlands .

a r t i c l e i n f o

Article history:Received 13 July 2006Revised 10 October 2008Accepted 24 October 2008

Keywords:GenreMovementEmotionAnimacyCategorizationMotion picture

0010-0277/$ - see front matter � 2008 Elsevier B.Vdoi:10.1016/j.cognition.2008.10.018

* Corresponding author.E-mail address: mail@valentijnvisch.nl (V.T. Visc

a b s t r a c t

The reported study follows the footsteps of Heider, and Simmel (1944) [Heider, F., & Sim-mel, M. (1944). An experimental study of apparent behavior. American Journal of Psychol-ogy, 57, 243–249] and Michotte (1946/1963) [Michotte, A. (1963). The perception ofcausality (T.R. Miles & E. Miles, Trans.). London: Methuen (Original work published1946)] who demonstrated the role of object movement in attributions of life-likeness tofigures. It goes one step further in studying the categorization of film scenes as to genreas a function of object movements.

In an animated film scene portraying a chase, movements of the chasing object were sys-tematically varied as to parameters: velocity, efficiency, fluency, detail, and deformation.

The object movements were categorized by viewers into genres: non-fiction, comedy,drama, and action. Besides this categorization, viewers rated their animacy attributionand emotional response. Results showed that non-expert viewers were consistent in cate-gorizing the genres according to object movement parameters. The size of its deviationfrom the unmanipulated movement scene determined the assignment of any target sceneto one of the fiction genres: small and moderate deviations resulted in categorization asdrama and action, and large deviations as comedy. The results suggest that genre classifi-cation is achieved by, at least, three distinct cognitive processes: (a) animacy attribution,which influences the fiction versus non-fiction classification; (b) emotional responses,which influences the classification of a specific fiction genre; and (c) the amount of devia-tion from reality, at least with regard to movements.

� 2008 Elsevier B.V. All rights reserved.

1. Introduction

The reported study follows the footsteps of Heider andSimmel (1944) who demonstrated the role of object move-ment in the attributions of life-likeness to figures. It goesone step further in studying the categorization of filmscenes as to genre as a function of object movements. Wepresented two animated abstract blocks involved in achase, varying five parameters of movement, and regis-tered: (1) animacy rating, (2) emotional responses, and

. All rights reserved.

h).

(3) genre categorization as non-fiction, drama, action, andcomedy.

The importance of genre knowledge and recognition forfilm viewers seems uncontested. Genre recognition primes(Roskos-Ewoldsen, Roskos-Ewoldsen, & Dillman-Carpen-tier, 2002) and directs (Zwaan, 1994) attention, affectsmodes of perception (Hawkins et al., 2005), generates nar-rative expectations (Grodal, 1997) and emotions (Smith,2003; Tan, 1996; Zillmann, 1988). Genres can be accuratelyrecognized by stylistic features as Hayward (1994) showedfor the literature, Dalla Bella and Peretz (2005) for musicand Visch and Tan(2007, 2008) for film. However, the par-ticular attributions and cues viewers use to arrive at agenre classification remain unclear. This study focuses on

266 V.T. Visch, E.S. Tan / Cognition 110 (2009) 265–272

the role of perceptual movement characteristics, animacyattribution, and emotional response in genre categoriza-tion. Four major genres were selected namely comedy, dra-ma, action, and non-fiction. Originating in mythology(Frye, 1957) and known at least since Aristotle ( Trans.,1988), these genres have stood significant historical periodand media shifts.

It may be assumed that in categorizing film as to genre,viewers use attributions to objects in motion at variouslevels of complexity. At a low level, animacy is involun-tarily attributed. Objects seen as animate are visually pro-cessed in the STS region (Beauchamp, Lee, Haxby, & Martin,2003), which in turn facilitates higher-order attributionsuch as emotions and intentions (Blakemore & Decety,2001; Scholl & Tremoulet, 2000). Automated animacydetection is an adaptive capacity, supporting early identifi-cation of living entities as prey, predators or mates (Allison,Puce, & McCarthy, 2000; Schultz, Friston, O’Doherty, Wol-pert, & Frith, 2005). Perceptual criteria proposed for anima-cy include changes of speed and direction (Tremoulet &Feldman, 2000), intention (Dittrich & Lea, 1994), goal ori-entation (Opfer, 2002), and displayed interactivity be-tween the objects ( Schultz et al., 2005). The stimuli weused meet all proposed criteria for animacy attribution,and in addition their motion patterns suggest a particularintention, namely chasing (Barrett, Todd, Miller, & Blyth,2005; Blythe, Todd, & Miller, 1999). A question this studypurports to answer is how animacy varies as a functionof movement characteristics. As to its role in genre catego-rization we predict that animacy enhances the recognitionof film realism, because moving objects that are clearlyartifacts seem to be realistically alive and socially aware(Heider & Simmel, 1944; Michotte, 1946/1963, 1950/1991). Hence, our animacy hypothesis: animacy attribu-tion correlates positively to non-fiction genre attribution.

Emotions may serve as a somewhat higher level cue togenre categorization. Genres are said to be organizedaccording to the particular emotions they typically elicit(Aristotle, Trans., 1988; Carroll, 2003; Grodal, 1997). Com-edy evokes mirth, action impresses the viewer, and dramaseeks to elicit tender emotions (Oliver, 2008), whereasnon-fiction does not try to evoke emotions, but insteadconveys an argument (Nichols, 2001). We expect that itis not paired with any emotion, though in our case thechase might invoke fear or admiration. Our emotionhypothesis then reads: genre categorization (comedy, dra-ma, action, and non-fiction) correlates with distinct emo-tion responses (respectively, funny, sad, impressive, andscary). To complete the view on the emotional responsesto movements, three emotions (aesthetic liking, surprise,and fascination) were added to the set of genre-specificemotions.

Finally, at the highest complexity level we propose thata distinction between fiction and non-fiction is crucial forgenre categorization. The distinction is pragmatically fun-damental in that non-fiction might have direct implica-tions for viewers’ personal lifes, whereas fiction allowsthem to lean back and appreciate the story and aestheticproperties of a film. The distinction may be based on theperceived degree of the transformation of reality. Non-fic-tion transforms the actual world only minimally (Branigan,

1992; Corner, 1995; Nichols, 2001), whereas fiction is sup-posed to transform it to a higher degree, with references toreality being indirect and metaphorical (Branigan, 1992;Dewey, 1934; Goodman, 1984; Singer, 1998). We propose,then, that the extent of transformation of the actual worldis perceived as a genre characteristic, with non-fiction act-ing as a baseline. Our realism hypothesis predicts thatuntransformed movements will be categorized as non-fic-tion in the first place, and as fiction, i.e. drama, action, andcomedy, in the second.

In addition, a more precise ranking in terms of fidelity torealism can be predicted. The drama genre’s transformationof reality is slight, thus facilitating empathy (Gaut, 1999).Drama portrays the reality of everyday life (Lacey, 2000;Neale, 2000) and is characterized by realistic acting (Hallam& Marshment, 2000). The action genre has a minimal simi-larity with reality allowing for thrilling conflicts, while cer-tain elements are enhanced in comparison to non-fiction(e.g. physical action, determination, and efficiency (Bor-dwell, 2006; Neale, 2000). The comedy genre’s relation tonon-fiction is the most remote: surprising deviations andexaggerations of reality abound (Neale & Krutnik, 1990;Sobchack, 2004). Our transformation hypothesis predictsthat moderate movement transformations from non-fictionscenes cue drama and action categorizations, while strongtransformations cue comedy categorization.

2. Method

2.1. Materials

A Basic chase scene between two abstract blocks wasanimated and varied as to five movement parameters. Achase was selected as the animations’ action script, be-cause the underlying intention is highly recognizable froma simple motion pattern (Barrett et al., 2005; Blythe et al.,1999). Animations produced in 3D MAYA 6.0 lasted 15 seach. Camera movement served action visibility, and waskept unobtrusive – see Fig. 1.

The choice of movement parameters was inspired by ananalysis of 80 chase scenes from the films of various fictionand non-fiction genres (news, sport, documentaries, andanimals), personal interviews with animators, and the lit-erature. Selected movement parameters were velocity, effi-ciency, fluency, detail, and deformation, for each of whichfour versions were made deviating from the Basic scenein one or the other direction. Deviations could be moderate(+ or �) or strong (++ or ��). Differences between thedeviations were exponential except for the efficiencyparameter.

2.2. Velocity

Velocity refers to the speed of the two blocks movingover the track. All objects, blocks, floor, and camera werescaled, but the track was not. Thus, perceptual sizes ofthe blocks and the duration of the scenes remained con-stant over all versions, while perceptual velocity varied.The variation factor was 1.5 between each two adjacentlevels.

Fig. 1. Filmstrip per second of animated running chase. The chase follows the following script: (a) A moves to the position of B, while B moves away from A;(b) the distance between A and B decreases over time until A and B are unified at the end of the chase; and (c) as the distance between A and B decreases, Btakes increasingly more bends that become shorter and sharper toward the end of the chase.

V.T. Visch, E.S. Tan / Cognition 110 (2009) 265–272 267

Comedies, especially the slap-stick type, use extremelyfast or slow movements. Velocity is a strong factor in dra-ma and comedy recognition (Visch & Tan, 2007) and inemotion. Fast movements evoke joy, surprise and excite-ment, whereas slow movements evoke sadness, weakness,gentleness and sympathy (Hille, 2001; De Meijer, 1989;Michotte, 1963; Pollick, Paterson, Bruderlin, & Sanford,2001; Scherer & Ellgring, 2007).

2.3. Efficiency

The ratio between the energy a character uses in pursu-ing a goal and the energy minimally needed to achieve itdetermines perceived efficiency. In our stimuli, efficiencyrelates to the directness of the chaser’s track. In a negativeefficiency version, the chaser’s zigzag movements arewider than those of the chased – see Fig. 2.

The action genre displays highly efficient actions in or-der to impress viewers (Neale, 2000). Drama protagonists’lack of resolution is reflected in less efficient movement.Animation theory stresses inefficient actions’ potentialfor comic effects (Thomas & Johnston, 1981).

2.4. Detail

Detail pertains to the temporal density of velocitychanges. The number of velocity alterations of the chaserwas manipulated through the number of ‘‘keys” in anima-tion. (A ‘‘key” is a programmed change of velocity at a spe-cific moment in time.) Detail level �� has two keys set atthe beginning and at the end, with the block moving at asingle velocity. Level � has five keys, the Basic scene eight,detail + 14, and detail ++ 26 keys, giving rise to 25 velocityalterations. High alteration densities seem to be character-istic of the chaotic character of movements in reality ascompared to the movements in fiction. Ashida, Lee, All-beck, Sun, and Badler (2001) found that when artificialagents displayed small and chaotic movements, they were

perceived as more natural than when showing less-de-tailed movements.

2.5. Fluency

Movement fluency was manipulated by varying thesmoothness of velocity transitions. The smoothest transi-tion (fluency ++) was achieved by using a horizontal tan-gent at each velocity change, i.e. keypoint, resulting in agradual deceleration to an extremely short period of nomovement, immediately followed by a gradual accelera-tion. At the next two levels, the tangents become steeper,until they are vertical at the fluency–level, resulting in noacceleration nor deceleration. Further velocity abruptnessat this level (�) was achieved by inserting duplicate framesin between keys, resulting in movement arrests of 40 mseach. In the fluency �� condition, six frames were in-serted, resulting in a 240 ms. stop of the chaser just beforethe velocity change.

Abrupt movements seem to be characteristic for com-edy (stop-and-go chases of Laurel and Hardy and ChuckJones’ Roadrunner) but also for the action genre wherethe action hero moves like a robot (Neale, 2000). Fluentmovements, in contrast, appear to be more so for drama.As regards expressive effects, Wallbott (1998) found thatnon-fluent body movements express anger, fear, and joy,while fluent ones express sadness, boredom, andhappiness.

2.6. Deformation

Deformation refers to dynamic shape alterations of thechaser object when changing its course. The ratio of heightto the width of the object was exponentially scaled by one-third per step, keeping depth, and volume constant. Wideobjects, level ++, cause large wedge-shaped deformationin negotiating a bend, see Fig. 3, whereas high and narrowobjects, level ��, are minimally deformed.

Fig. 2. Three of five efficiency tracks of the chaser: the extremely inefficient track (��), the Basic track (0) and the extremely efficient track (++). The chasedonly follows the Basic track.

Fig. 3. Example of deformation level ++.

268 V.T. Visch, E.S. Tan / Cognition 110 (2009) 265–272

Deformation is akin to the animation technique of‘‘squash and stretch” (Thomas & Johnston, 1981), yieldinga funny impression of elasticity of cartoon bodies. Thisis also a trademark of comedy actors, from Buster Keatonto Jim Carrey. In contrast, action heroes seem to possessextremely controlled and rigid bodies that scare andimpress.

2.7. Dependent measures and experimental design

There are 12 dependent variables grouped into fourtypes: genre categorization (comedy, drama, action, andnon-fiction), genre-related emotional responses (funny,sad, impressive, and scary), aesthetic emotional responses(fascination, aesthetic liking, and surprise) and animacy.All dependent variables were measured using five-pointlikert scales. Two data sets were obtained, animacy ratings

(‘‘animacy study”) and the remaining measures (‘‘mainstudy”).

The design consisted of five non-crossed independentvariables, the movement parameters, each with five levels(��, �, 0, +, and ++). Level 0, the Basic scene, was one andthe same scene for all five parameters. Because of its an-chor function, the Basic scene was presented and ratedtwice by each participant, the average being taken as thedata. Thus, in all, the stimulus set consisted of 22 scenes.Presentation order of scenes was randomized. For the mainstudy two orders were created, one the reverse of theother. In the animacy study, the 22 scenes were randomlypresented to each participant.

2.8. Procedure and participants

Each scene was shown twice followed by the rating ofthe participants. The main study consisted of two ratingsessions each with a different randomized order, one forgenre categorization, and the other for the aesthetic andemotional responses. The animacy study consisted of oneanimacy rating session in which the stimuli were random-ized. In all, each of the 22 scenes, including the dual Basicscene, was presented four times to the participants in themain study and two times to participants in the animacystudy.

Fifty-two participants (19 male, 33 female, ages rang-ing from 18 to 31), students of industrial design and ofarts, joined in the main study and were rewarded witha cinema ticket. They were tested in groups from six toten. Stimuli were presented using a beamer and ratingscollected through questionnaires. Twenty-two psychologystudents (11 male, 11 female, ages ranging from 17 to29) participated in the animacy study and receivedcourse credits. In the animacy study, the stimuli werepresented on 15.1 in. computer screen and ratings weretaken on-line.

Table 1Optimal Movements. Optimal levels within a specific genre or response arethose categorized significantly stronger, than at least one of the other levelsof that parameter. Horizontal ties do not differ in effect and representoptimal sets. For instance, focusing on animacy ratings, velocity levels 3, 4,and 5 do not differ among themselves, but all do have significantly higherratings than levels 1 and 2, rendering the set optimal for animacyattribution. It should be noted that this Table does not present movementoptimal between the dependent variables.

Velocity Efficiency Fluency Detail Deformation

Animacy 3, 4,and 5

1, 2, 3,and 4

2, 3, 4,and 5

1

Comedy 1 1 and 5 1 4 and 5 1, 4, and 5Drama 1 4Action 5 1 and 2Non-fiction 3 3 3 3Funny 1 and 5 1 4 and 5 1, 4, and 5Sad 1 and 2 5Impressive 5 1 and 5 1 5 4Scary 4 and 5 1, 4, and

51 and 2

Fascinating 5 1 and 5 1, 4, and5

2, 4,and 5

1, 2, 4, and 5

Aestheticliking

1, 4,and 5

1 and 5 4 and 5

Surprise 5 1 and 5 1 and 5 2, 4,and 5

1 and 5

Legend: Level 1 = ��; level 2 = �; level 3 = 0; level 4 = +; and level 5 = ++.Note: Significance for the optimal levels is measured using MANOVAGames Howell post-hoc tests with p < .05.

V.T. Visch, E.S. Tan / Cognition 110 (2009) 265–272 269

3. Results

3.1. Animacy ratings

Data of the animacy study (N = 22) were analyzed usingrepeated measures and showed no significant effect ofgender. The analysis did show that all movement parame-ter–level interactions, except that for ‘detail’, had signifi-cant effects on animacy ratings: velocity (F = 4.81,p < .01); efficiency (F = 5.67, p < .001; fluency (F = 7.68,p < .001); and deformation (F = 3.90, p < .01). Fig. 4 pre-sents the animacy ratings as a function of movementparameters and levels. Significant linear trends were ob-tained for velocity (F = 7.59, p < .02), fluency (F = 11.05,p < .01), and deformation (F = 15.39, p < .01). Increases invelocity and fluency resulted in higher animacy ratingsand increases in deformation resulted in lower animacyratings. Increases in the parameter efficiency from level 1(��) to 4 (+) did not have a significant effect on animacyratings, but level 5 (++) was rated as being significantly lessanimate than any of the former levels (F > 6.63, p < .018).The effect of efficiency on the animacy ratings was best de-scribed by a quadratic contrast (F = 10.57, p < .01).

3.2. Genre and emotion ratings

The data of the main experiment (N = 52) were sub-jected to MANOVA showing that control variables genderand order had no main effects on the dependent variables(F = 2.27, p = .13; F = 0.63, p = .43). Table 1 presents theoptimal levels of the movement parameters for each ofthe 11 categorizations and responses plus the animacyratings.

3.3. Correlations

Correlations between animacy-, emotion-, and genre-ratings were obtained by calculating Spearman rank corre-

1.6

1.8

2

2.2

2.4

2.6

2.8

3

3.2

3.4

3.6

3.8

- - -

Level of Movem

Mea

n A

nim

acy

Rat

ing

Fig. 4. Effect of movement parame

lations between the means of the dependent variable foreach of the 25 level–movement parameter combinations.Animacy correlated strongly and positively with non-fic-tion categorization, validating our first animacy hypothe-sis: (rs) = .63. An additional significant correlation ofanimacy attribution was found with action categorization(rs = .51). Animacy did not correlate significantly withemotion ratings.

The emotion hypothesis predicting correlations be-tween distinct emotions and genres was tested using not

Velocity

Efficiency

Fluency

Deformation

0 + + +

ent Transformation

ters on animacy attribution.

2.2

2.4

2.6

2.8

3

3.2

3.4

0 1 2

*

*

*

*

*

*

*

Comedy

Drama

Action

Non-fiction

Extent of Movement Transformation

Mea

n G

enre

Cat

egor

izat

ion

Rat

ing

*

* *

* p < .03

Fig. 5. Effect of movement transformation on genre categorization.

270 V.T. Visch, E.S. Tan / Cognition 110 (2009) 265–272

the means, but the individual ratings obtained in the mainstudy. Significant correlations were found with the fictiongenres only, for the pairs comedy genre and response‘‘funny” (rs = .49), drama and ‘‘sad” (rs = .24), and actionand ‘‘impressive” (rs = .33). Non-fiction was not correlatedwith any emotion, including ‘‘scary” or ‘‘impressive”.

3.4. Genre classification

Fig. 5 presents the effects of movement transformationon genre categorizations. In order to test the realism pre-diction, the levels of all movement parameters were col-

2.2

2.4

2.6

2.8

3

3.2

3.4

*

*

Level of Movem

-- -

Mea

n G

enre

Cat

egor

izat

ion

Rat

ing

*

*

*

*

Fig. 6. Effect of movement transformation on

lapsed from five to three levels: 0, 1 (+ and �) or 2 (++and ��). Significance was established using post hocGames Howell tests. Analysis showed that unmanipulatedmovement scenes were categorized significantly strongeras non-fiction than as any other genre. Moreover, theywere categorized as non-fiction stronger than were manip-ulated movement scenes. Our third realism hypothesis wasvalidated in that unmanipulated movement scenes werecategorized strongest as non-fiction, second to strongestas drama, second to weakest as action, and weakest ascomedy. The transformation prediction also received sup-port. Moderate transformations of movements increased

** *

ent Transformation

Non-fictionComedy

0 + ++

*

*

*

*

* p < .05

non-fiction and comedy categorization.

V.T. Visch, E.S. Tan / Cognition 110 (2009) 265–272 271

classification in all three fiction genres significantly com-pared to untransformed movements. Strong transforma-tions only significantly increased categorization in thecomedy genre. It is noteworthy that categorization as thecomedy genre was affected by movement transformationin exactly the opposite direction as non-fiction categoriza-tion (rs = �.31) – see Fig. 6.

4. Discussion

This study showed that the movements of abstract ob-jects, varying on velocity, efficiency, detail, fluency, anddeformation, elicit consistent higher-order categorizationsof film scenes as to genre, as well as lower-order emotionresponses and animacy attributions.

Animacy attributions increased, first, by a relativelyhigh velocity of the object probably resulting in more ‘li-vely’ movements. Second, animacy was increased by thelack of deformation of the object when it negotiates abend. This result is unexpected in light of the wide-spread occurrence of squash and stretch deformation inanimated cartoons (cf. Thomas & Johnston, 1981). How-ever, object deformation may have been confounded withobject proportion: strong deformations featured horizon-tally oriented object shapes, not resembling living bodies,whereas weak deformations were paired with tall uprightobjects, resembling human body shapes, which contrib-utes to animacy attribution (Lange & Lappe, 2006). Third,movement efficiency affected animacy. This could be ex-pected since efficiency in a chase is closely tied to pro-posed animacy cues: (a) goal-directed movement(Blakemore et al., 2003; Opfer, 2002) and (b) intentionalrelations or interactivity between objects (Neri, Luu, &Levi, 2006; Schultz, Friston, O’Doherty, Wolpert, & Frith,2005). However, we found a significant drop of animacyattribution when the chaser’s efficiency was extremelyhigh. This effect may be due to seeming anticipatorycapacities of the chaser enabling machinelike precisionin predicting where the chased will end. Fourth, a similareffect of negative animacy attribution to machinelikemovements was found with extremely low-fluent move-ments, which featured exclusively abrupt starts andstops. The lack of acceleration or deceleration also re-minds one of machine behavior rather than the conductof sentient beings. The prediction that animacy attribu-tion is associated with film realism, involving animate ac-tion, is validated by the considerable correlation betweenanimacy ratings and non-fiction categorization. We sug-gest that animacy attribution is not only functional foradaptive purposes like proper social interaction (Allisonet al., 2000) and avoidances of dangerous species(Schultz, Friston, O’Doherty, Wolpert, & Frith, 2005), butit seems also to confer ‘reality status’ upon percepts ofmotion pictures.

Specific stimuli movements that were varied in thisexperiment induced specific emotional responses by theviewers such as low velocity generating sadness – in linewith De Meijer (1989) and Pollick et al. (2001). Concerningthe role of emotion in genre categorization, we found thateach of the fiction genres correlated significantly with a

specific emotion, according to expectation: comedy withfunny, drama with sad, and action with impressive. As ex-pected, non-fiction did not correlate with emotion ratings.The contrast between functions of fiction and non-fiction,one affective and the other informative, was already pro-claimed by Aristotle (Trans., 1988).

As to the effects of movement transformations, untrans-formed movements seemed to give the viewer an impres-sion of realism as they were categorized as an instance ofthe non-fiction genre, as the realism hypothesis predicted.Moreover, these realistic movements were categorized indeclining order from non-fiction to drama to action tocomedy, conforming to the transformation hypothesis.The degree of transformation from the unmanipulated‘‘realistic” movements determined as what particular fic-tion genre scenes were categorized: moderate deviationsresulted in categorization as drama and action, large devi-ations as comedy.

The results have two implications for a cognitive theoryof genre. Firstly, they suggest that genre knowledge con-sists in part of representations that can be evoked by thespecific types of movement. Secondly, the results suggestthat genres are not separate mental categories but orga-nized as to their deviation from non-fiction. The knowl-edge involved may be expected to be implicit (Schacter,1996): people might use it in classification without beingable to explicitly describe the movements or their transfor-mative distance from non-fiction movements. In line withtheorizing by Mar and Oatley (2008) and Goldman (2006)we believe that the embodied simulation of charactermovement is inherent to typical responses to fiction. Itcan be expected that film viewers match their perceptionsas a default with their large repertoire of non-fiction motorexperiences (Barsalou, 2008). The extent of successfulnessof such a match could function as an implicit cue for clas-sification processes (Beilock, Lyons, Mattarella-Micke, Nus-baum, & Small, in press; Goldman & Sripada, 2005; Helbig,Graf, & Kiefer, 2006) of fiction. Future research, using on-line measurements such as brain imaging, might revealwhether matching with real life motor repertoires occursin fiction processing. On-line measurements could alsoshed light on dependencies between genre categorizationand attributions to movement percepts.

In summary, we have shown that naïve subjects areable to categorize genres, including fiction and non-fic-tion, consistently according to object movement parame-ters. Such a genre classification is achieved by, at least,three distinct cognitive processes: (a) animacy attribution,which influences the fiction versus non-fiction classifica-tion; (b) emotional responses, which influences theclassification of a specific fiction genre; and (c) theamount of deviation from reality, at least with regard tomovements.

Acknowledgements

We would like to express our gratitude to the journaleditor Gerry Altman and three anonymous reviewers fortheir stimulating and helpful comments, which havegreatly improved the quality of this manuscript.

272 V.T. Visch, E.S. Tan / Cognition 110 (2009) 265–272

References

Allison, T., Puce, A., & McCarthy, G. (2000). Social perception from visualcues: Role of the STS region. Trends in Cognitive Sciences, 4(7),267–278.

Aristotle (1988). Poetica (N. van der Ben & J.M. Bremer, Trans.)Amsterdam: AthenaeumPolak and Van Gennep (Original workwritten between 335-323 BC).

Ashida, K., Lee, S., Allbeck, J. M., Sun, H., & Badler, N. I. (2001). Pedestrians:Creating agent behaviors through statistical analysis of observationdata. IEEE Computer Animation 2001: Proceedings of the 14th Conferenceon Computer Animation, 84–92.

Barrett, H. C., Todd, P. M., Miller, G. F., & Blyth, P. W. (2005). Accuratejudgments of intention from motion cues alone: A cross-culturalstudy. Evolution and Human Behavior, 26(4), 313–331.

Barsalou, L. W. (2008). Grounded cognition. Annual Review of Psychology,59, 617–645.

Beauchamp, M. S., Lee, K. E., Haxby, J. V., & Martin, A. (2003). FMRIresponses to video and point-light displays of moving humans andmanipulable objects. Journal of Cognitive Neuroscience, 15, 991–1001.

Beilock, S. L., Lyons, I. M., Mattarella-Micke, A., Nusbaum, H. C., & Small, S.L. (in press). Sports experience changes the neural processing of actionlanguage. In Proceedings of the national academy of sciences, USA.

Blakemore, S.-J., Boyer, P., Pachot-Clouard, M., Meltzoff, A., Segebarth, C.,& Decety, J. (2003). The detection of contingency and animacy fromsimple animations in the human brain. Cerebral Cortex, 13(8),837–844.

Blakemore, S., & Decety, J. (2001). From the perception of action to theunderstanding of intention. Nature Reviews: Neuroscience, 2, 561–567.

Blythe, P. W., Todd, P. M., & Miller, G. F. (1999). How motion revealsintention: Categorizing social interactions. In G. Gigerenzer & P. M.Todd (Eds.), Simple heuristics that make us smart (pp. 257–285). NY:Oxford University Press.

Bordwell, D. (2006). The way Hollywood tells it: Story and style in modernmovies. CA: University of California Press.

Branigan, E. (1992). Narrative comprehension and film. London: Routledge.Carroll, N. (2003). Engaging the moving image. New Haven, CT: Yale

University Press.Corner, J. (1995). Television form and public address. London: Edward

Arnold.Dalla Bella, S., & Peretz, I. (2005). Differentiation of classical music

requires little learning but rhythm. Cognition, 96, 65–78.De Meijer, M. (1989). The contribution of general features of body

movement to the attribution of emotion. Journal of NonverbalBehavior, 13(4), 247–268.

Dewey, J. (1934). Art as experience. NY: Berkley Publishing Group.Dittrich, W. H., & Lea, S. E. (1994). Visual perception of intentional motion.

Perception, 23(3), 253–268.Frye, N. (1957). Anatomy of criticism: Four essays. Princeton: Princeton

University Press.Gaut, B. (1999). Identification and emotion in narrative film. In C. R.

Plantinga & G. M. Smith (Eds.), Passionate views: Film, cognition andemotion (pp. 200–216). Baltimore: John Hopkins University Press.

Goldman, A. I. (2006). Imagination and simulation in audience responsesto fiction. In S. Nichols (Ed.), The architecture of the imagination(pp. 41–56). Oxford: Oxford University Press.

Goldman, A. I., & Sripada, C. S. (2005). Simulationist models of face-basedemotion recognition. Cognition, 94, 193–213.

Goodman, N. (1984). Of mind and other matters. Cambridge, MA: HarvardUniversity Press.

Grodal, T. (1997). Moving pictures. Oxford: Clarendon Press.Hallam, J., & Marshment, M. (2000). Realism and popular cinema.

Manchester, UK: Manchester University Press.Hawkins, R. P., Pingree, S., Hitchon, J., Radler, B., Gorham, B. W., Kahlor, L.,

et al (2005). What produces television attention and attention style?Genre, situation, and individual differences as predictors. HumanCommunication Research, 31(1), 162–187.

Hayward, M. (1994). Genre recognition of history and fiction. Poetics, 22,409–421.

Heider, F., & Simmel, M. (1944). An experimental study of apparentbehavior. American Journal of Psychology, 57, 243–249.

Helbig, H. B., Graf, M., & Kiefer, M. (2006). The role of action representationsin visual object recognition. Experimental Brain Research, 174, 221–228.

Hille, K. (2001). Synthesizing emotional behavior in a simple animatedcharacter. Artificial Life, 7, 303–313.

Lacey, N. (2000). Narrative and genre: Key concepts in media studies. NY:Palgrave.

Lange, J., & Lappe, M. (2006). A model of biological motion perceptionfrom configural form cues. Journal of Neuroscience, 26(11),2894–2906.

Mar, R. A., & Oatley, K. (2008). The function of fiction is the abstractionand simulation of social experience. Perspectives on PsychologicalScience, 3(3), 173–192.

Michotte, A. (1963). The perception of causality (T. R. Miles & E. Miles,Trans.). London: Methuen (Original work published 1946).

Michotte, A. (1991). The emotions regarded as functional connections. InG. Thinès, A. Costall, & G. Butterworth (Eds.), Michotte’s experimentalphenomenology of perception. Hillsdale: Erlbaum (Reprinted fromFeelings and emotions: The Mooseheart Symposium (pp. 114–125), byM. L. Reymert (Ed.), 1950, New York: McGraw-Hill).

Neale, S. (2000). Genre and Hollywood. London: Routledge.Neale, S., & Krutnik, F. (1990). Popular film and television comedy. London:

Routledge.Neri, P., Luu, J. Y., & Levi, D. M. (2006). Meaningful interactions can

enhance visual discrimination of human agents. Nature Neuroscience,9, 1186–1192.

Nichols, B. (2001). Introduction to documentary. Bloomington, IN: IndianaUniversity Press.

Oliver, M. B. (2008). Tender affective states as predictors of entertainmentpreference. Journal of Communication, 58, 40–61.

Opfer, J. E. (2002). Identifying living and sentient kinds from dynamicinformation: The case of goal-directed versus aimless autonomousmovement in conceptual change. Cognition, 86, 97–122.

Pollick, F. E., Paterson, H. M., Bruderlin, A., & Sanford, A. J. (2001).Perceiving affect from arm movement. Cognition, 82, 51–61.

Roskos-Ewoldsen, D. R., Roskos-Ewoldsen, B., & Dillman-Carpentier, F. R.(2002). Media priming: A synthesis. In J. Bryant & D. Zillmann (Eds.),Media effects (2nd ed., pp. 97–120). Mahwah, NJ: Erlbaum.

Schacter, D. (1996). Searching for memory: The brain, the mind, and the past.New York: Basic.

Scherer, K. R., & Ellgring, H. (2007). Multimodal expression of emotion:Affect programs or componential appraisal patterns? Emotion, 7(1),158–171.

Scholl, B. J., & Tremoulet, P. D. (2000). Perceptual causality and animacy.Trends in Cognitive Sciences, 4(8), 299–309.

Schultz, J., Friston, K. J., O’Doherty, J., Wolpert, D. M., & Frith, C. D. (2005).Activation in posterior superior temporal sulcus parallels parameterinducing the percept of animacy. Neuron, 45(4), 625–635.

Singer, I. (1998). Reality transformed: Film as meaning and technique. MA:MIT press.

Smith, G. M. (2003). Film structure and the emotion system. Cambridge, UK:Cambridge University Press.

Sobchack, V. (2004). Thinking through Jim Carrey. In C. Baron, C. Diane, &F. P. Tomasulo (Eds.), More than a Method: Trends and traditions incontemporary film performance (pp. 275–296). MI: Wayne StateUniversity Press.

Tan, E. S. (1996). Emotion and the structure of narrative film. Mahwah, NJ:Erlbaum.

Thomas, F., & Johnston, O. (1981). The illusion of live: Disney animation.New York: Disney Editions.

Tremoulet, P. D., & Feldman, J. (2000). Perception of animacy from themotion of a single object. Perception, 29, 943–951.

Visch, V. T., & Tan, E. S. H. (2007). Effect of film velocity on genrerecognition. Media Psychology, 9(1), 59–75.

Visch, V. T., & Tan, E. S. H. (2008). Narrative versus style: Effect of genre-typical events versus genre-typical filmic realizations on film viewers’genre recognition. Poetics, 36, 301–315.

Wallbott, H. G. (1998). Bodily expression of emotion. European Journal ofSocial Psychology, 28, 879–896.

Zillmann, D. (1988). Mood management: Using entertainment to fulladvantage. In L. Donohew, H. E. Sypher, & E. Tory Higgins (Eds.),Communication, social cognition and affect (pp. 147–171). Hillsdale, NJ:Erlbaum.

Zwaan, R. A. (1994). Effect of genre expectations on text comprehension.Journal of Experimental Psychology: Learning, Memory and Cognition,20(4), 920–933.