A Proposed Study: 

The Brain and Kinetic Art


Kristen Turner

Stephen F. Austin State University

  • The disciplines of science, philosophy, psychology, theology, andart have each addressed fundamental issues of life such as humannature and death. Although the outcomes of their inquiries aresometimes drastically different, collisions among these perspectivessuggest that the real differences lie in the means used to obtain theends. For example, according to the neurobiologist Semir Zeki (1994),the artistic genre of Futurism unwittingly discovered the optimalstimulus to influence the firing of neurons in area V5 of the visualcortex, and anticipated science by forty years. Thus, science and artare not mutually exclusive. Nevertheless, some critics areuncomfortable about the intermingling of art and science. It raisesthe issue of what it means to be human (Goguen, 1999). Aphysiologically based theory of aesthetics seems reductionistic. Thatis, behavior is reduced to known physical determinants. Thisthreatens the assumption of free will and choice. Just as the brainplays a role in aesthetics, however, aesthetics can also teach usabout the brain (Zeki & Lamb, 1994).

    There has been interest in the relationship between the brain andart since the classical Greeks, through the Italian Renaissance, andtoday. Skoyles (2002) describes how Western culture's concepts aboutthe perception of art can be traced back to Plato's concept ofMimesis, the idea that perception is an internal copying of the imageof the external environment. This theory of aesthetics has beendominant from 400 BC until now. Mimesis is oversimplified, however,because art is not merely an image on the two retinas. Attributes ofa visual scene are processed separately and then recombined into aunified perception. Garcia-Cairasco (1998) also points out that artand the brain have not historically been considered mutuallyexclusive. Leonardo da Vinci, renown for his paintings, was ananatomist and a mathematician as well as an artist. He producednumerous sketches of the brain, although not all were accurate. LikeZeki, da Vinci combined his training in science with a passion forart.

    Leontiev (2002) has conducted numerous studies about theperception of art over the past decade. In reflection, he emphasizesthe role of attention in the perception of art. If the viewer doesnot approach the work with the proper mindset then, likeinattentional blindness, the intricacies of the piece will be lostand its effect will be superficial. Attention is necessary forperception. The viewer can recognize colors, shapes, and patterns,but recognition is not the same as perception. Perception involvesthe deconstruction and reassembly of the image. For Leontiev, then,the perception of art is an active process requiring an attentiveobserver. This active involvement of the viewer differs from Plato'smore passive idea of Mimesis.

    Semir Zeki is a dominant figure in the field of research of theperception of art. He is a prolific, eloquent writer with a largebody of work and an equally large share of criticism. Despite hiscompelling arguments, critics accuse Zeki of championing areductionistic theory of aesthetics. His views are more moderate,however, than his critics assume. He repeatedly asserts that artists,unwittingly and instinctively, act as neurologists and that sciencecan and should learn from art. The title of Zeki's (1998) article,Daedalus, aptly describes his conception of the symbioticrelationship of art and the brain. In Greek mythology, Daedalus wasan artisan and inventor who wished to leave Crete and designedfeathered wings that were tenaciously held together by wax. Daedalusinstructed his son, Icarus, to fly a middle course between the sunand the sea to avoid the pitfalls of either extreme. Unfortunately,Icarus flew too close to the sun, which consequently melted the waxand he plunged down to a watery grave. Zeki, however, heeds themessage of Daedalus and he maintains a steady course between histraining in neurobiology and his enthusiasm for art. Zeki argues thatsince it is the brain that directs the creation of and appreciationof art, then art is subject to the laws of the brain. The reason wecan communicate with one another about shared responses to art isbecause we share the same brain structure. He suggests thatneurologists can learn from art. Art, therefore, can function as analgorithm to study the brain.

    Both art and the brain seek to acquire knowledge and to discoverperceptual constancies. Mondrian anticipated science by declaringthat there are constant forms in nature, horizontal and verticallines, and that nature builds upon these to create all else. Thisdeclaration was made decades before Hubel and Weisel's discovery ofcomplex and end-stopped cells that are tuned to respond best to linesof specific orientations. Thus, the artist intuitively discerned thatto which the scientist demanded proof.

    Artists act as neurologists when, over time, their work becomesmore and more tailored to the optimal stimulus of the receptive fieldof a neuron in a particular module by de-emphasizing all otherattributes in a visual scene. Zeki (1998) compares Cubism toview-invariant neurons. Cubist art is conceptualized as freezing amoment in time and representing it as seen from various angles, muchlike a view-invariant neuron. Point of view, the observer'sperspective, was eliminated by the presentation of ambiguous figuresthat could be perceived as facing not one direction but several.Correspondingly, the brain sees objects from different angles anddistances and fuses these different images into a unified perception.Zeki also points out the role of top-down knowledge in decipheringthe elements of Picasso's work; one begins with the title of thepiece and searches for its justification in the contents of the work.Additionally, Zeki asserts that Michelangelo left his sculpturesunfinished because the brain would fill in the appropriaterepresentation. The colorful artwork of another genre, the Fauvists,is the optimal stimulus for area V4 of the visual cortex. Artworkthat emphasizes a particular quality, such as motion, mustde-emphasize all other qualities to produce the desired effect. Thisis similar to neurons that are finely tuned to respond to particularstimuli.

    Zeki's (1998) ideas have generated much commentary. Sudol (2002)compares and contrasts two models of the brain's response to art: therepresentational and the non-representational models. Therepresentational model refers to Plato's concept of Mimesis, whichSudol argues is an inadequate model. He dismisses Zeki's approach asrepresentational. The non-representational model corresponds toLeontiev's idea of an active perceiver that interacts with the workof art.

    Sudol (2002) attacks the representational model partially becauseof the contributions of Zeki. Since Zeki (1998) discusses how theprocessing of visual information as it is carried through themagnocellular and parvocellular pathways is unconscious, Sudolconsequently assumes that Zeki represents the absolute antithesis ofchoice and free will. In other words, Sudol shies away from areductionistic theory of aesthetics in an existential panic.

    Sudol (2002) also attacks the representational theory on the basisof a lack of parsimony. This also stems partially from Zeki's (1998)comments regarding the role of top-down processing in perception.Sudol argues that if perception is determined by comparing thestimuli with our previous experiences and memories, and this givesthe representation its meaning, and memories and experiences aresimultaneously compared to the stimulus, then we have a vicious cycleinvolving representation and top-down processing. The alternative isthat memory and previous knowledge can be modeled withoutrepresentation. If so, then it begs the question of why the processis selective; why is some rather than all information processed inthis manner? Sudol argues that the representational model isnecessarily inadequate as a theory for aesthetic perception becauseit excludes the perceiver; it is a misnomer to discuss a theory ofperception without taking into account the perceiver.

    Sudol (2002) cites the phenomena of change blindness as evidenceagainst Mimesis. The observer doesn't register what is actuallyoccurring in the environment, but instead perceives what he or sheattends to. If the theory of Memesis was correct, then changeblindness would never occur because we would always perceive exactlywhat is going on in the environment.

    For Sudol, (2002) a theory of art must account for simultaneousawareness both of the conscious perceiver and the factors of top-downknowledge brought to a perceptual situation. This explains how onecan find a degree of coherence in otherwise seemingly irrational Dadaart. Dada art is an arrangement of unrelated objects, but since theyare familiar objects we are free to brainstorm and search for ameaning that explains their interrelationships. This involves anactive, conscious perceiver searching previous knowledge, memories,and expectations. Similarly, Surrealistic art achieves a sense ofuneasiness because objects are removed from their usual context andjuxtaposed with other displaced objects. This conflicts with ourstored knowledge of the stimulus and its usual connotations.

    The non-representational view postulates that, instead ofgenerating an internal representation of the external world, weimpose meaning on a stimulus. This satisfies Sudol's (2002) tworequirements. The structure and meaning of the work of art is basedon top-down knowledge and carried out by a conscious perceiver. Thenon-representational view emphasizes the interaction between theperceiver and stimulus.

    Sudol (2002), then, sees the non-representational view as a sortof phenomenological approach, concerned with the immediate awarenessand observations of a conscious perceiver. He sees Zeki's work,however, as a representational approach that attempts to engineerart. Such an attempt is doomed because the response to art is toocomplex to be reduced to a formula. Sudol noticeably avoids the topicof autostereograms, computer-generated art, and the use of thedrawing system of linear perspective that flourished during theRenaissance, all of which engineered art to different degrees.

    Tibbets (2001) is much more positive about Zeki's work than Sudol,but he raises important questions about Zeki's theories. If thefunction of the brain is to acquire knowledge and art provides thebrain with knowledge, then in what way are images and objects in amuseum distinct from the objects we perceive in day-to-day living?Tibbets, therefore, raises the question of what constitutes art.Despite his reservations, Tibbets praises Zeki's work for relatingsimple lines, colors, and edges in Cubist and Impressionistic art tocomplex and end-stopped cells.

    There has been a recent discovery that supports a representationaltheory of aesthetics. Christopher Tyler, a neuroscientist, discoveredthat if a vertical line is drawn down the center of a portrait it isnearly always on one of the subject's eyes. This placement held trueeven for Picasso, the rest of whose subjects' facial features mayappear all over the canvas. Since there is no such eye centeringprinciple taught in formal art classes, Tyler concluded that it isthe product of innate aesthetic principles in the brain (Moss, 1999).Clearly, this supports a representational theory of aesthetics.

    Multimedia art has provided a new dimension to the debate ofrepresentational versus non-representational theories of aesthetics.Kunz (1997) asserts that the future of art lies in multimedia.Interestingly, Kunz compares multimedia software such as the AdobePhotoshop to the Cubist's use of presenting a figure from multipleangles. Although the software is normally used as a digitalpaintbrush, the images it generates can also be broken down intocomponent parts. If the user repeatedly zooms in closer and closerthen the image begins to break up into disparate areas of color;zooming in closer to these images reduces them to basic geometricshapes such as rectangles and circles. Kunz describes this as atheory of aesthetics based on multiple scales and likens it to theCubist's approach of multiple angles. This, therefore, seems like anonrepresentational method that imitates a representational theory ofaesthetics.

    According to Zeki & Lamb (1994), artists, operating oninstinct and based on their visual perceptions, concluded that motioncould be depicted in art independently of other attributes. Thepurpose of kinetic art is not simply to simulate motion but rather tosubordinate other attributes, such as form and color, to the dominantelement of motion. In the genre of Futurism, artists depicted motionto express the tumultuous political and social upheavals of Europe inthe early nineteenth century.

    Zeki & Lamb(1994) distinguish 3 stages in the development ofkinetic art. In the first stage, artists expressed the importance ofphysical movement in manifestos but represented motion in staticform. In the second stage, physical objects that are actually inmotion are incorporated into art. Laszlo Moholy-Nagy designedLight prop for an electrical stage, a machine that consistedof moving parts as well as moving mirrors. The machine produced theeffect of moving light reflected in all directions. The third stagewas an attempt to separate motion from form and color. Tinguelydesigned a machine called Homage to New York that wasexplicitly built to self-destruct and thus liberate motion from form.Spectators gathered in the garden of the Museum of Modern Art towitness the spectacle. The machine was set into motion and, afterabout half an hour, nearly fulfilled the destiny its creator hadintended by setting itself on fire to self-destruct but, alas,firemen were present and promptly extinguished the flames. Kineticart never progressed further than this third stage until Zeki &Lamb (1994) studied the link between brain activity and viewingkinetic art.

    Zeki & Lamb(1994) tested the hypothesis that, while viewingkinetic art, the brain responds as if the stimulus is an actualobject in motion. Zeki & Lamb used a PET scan to record brainactivity from participants while they viewed an image very much likea Moire pattern, a static image that nevertheless gives a slightsense of movement (see Figure 2). The result was brain activity inarea V5, which is a module for processing information about motion.Thus, the brain responds to motion depicted in kinetic art as if itwere an actual physical object in motion.

    The present study will expand on Zeki & Lamb's (1994) kineticart study. What is original about this study is that it will not onlyattempt to replicate their results with static representations ofmovement, but it will also measure brain activity while viewingfigural art with actual motion that will be produced by acomputer-generated art software program. Participants will alsoindicate the detection of movement while simultaneously being brainscanned. The independent variable is degree of motion with fivelevels being a static image, a static image showing a figure inmidstep, an image with actual motion that starts and stops, a kineticabstract image, and a non-kinetic abstract image. The firsthypothesis states that figural images with actual motion willgenerate more brain activity in areas V3 and V5 than figural imageswith static depictions of motion. The second hypothesis states that akinetic image will result in higher brain activity in areas V3 and V5than a non-kinetic image. The third hypothesis states thatparticipants will indicate higher detection of movement for thekinetic image and the figural image with motion than for the otherconditions. Thus, this study will attempt to match data from thepsychophysical as well as the physiological levels of analyses.

  • Method
  • Participants

    The participants will be 10 male and 10 female undergraduatepsychology students at Stephen F. Austin State University. Thestudents will volunteer to participate in the study in order toreceive extra credit in their respective psychology classes. Theparticipants must be willing to fast for one day prior toexperimentation since it is a requisite of using a PET scanner tomeasure brain activity.


    An Adobe Photoshop 7.0 software program will produce thecomputer-generated art on a 1998 Packard Bell computer. This softwareprogram is appropriate because images can be manipulated to actuallymove across the screen. The software program will generate threefigural images and two abstract images, all of which present varyingdegrees of motion. The three figural images depict a man in a park.In the first image a man is sitting on a park bench, thus a staticimage of a static figure. The second image presents the same scenewith the only difference in that the man is no longer sitting on thepark bench, but rather he appears frozen in midstep. This image,therefore, depicts a static representation of a moving figure. In thethird figural image the man is actually set into motion to moveacross the screen. The two abstract images will be kinetic andnon-kinetic, respectively. The non-kinetic image will be anillustration of the monocular depth cue of texture gradients (seeFigure 1). The kinetic image will be a Moire pattern generated by aMoire kit (see Figure 2). A button is provided for the participantsto indicate when they see motion. A neuroimaging technique, positronemission tomography (PET), will measure brain activity in response tothe pictures.


    Prior to experimentation, the participants will fast for one day.The participants will be injected with a small dose of radioactivetracer that travels through the bloodstream to the brain andindicates the volume of blood flow. Next the participants will lie ona table and be placed into the scanner. Inside of the scanner thereis a mirror above the head that will enable the participants to seethe images. The computer-generated images will be projected onto ascreen that is located in front of the PET scanner, thus enabling theparticipants to see the images by their reflection in the mirror. Theparticipants will press the button every time they believe that theysee actual movement or have a strong impression of seeing movement ina static image. Each trial will last 25 seconds. The first imagepresented will depict the man sitting on a park bench. Theparticipant will either press the button or not, depending upon theirsubjective impressions of motion. A scan will be performedsimultaneously and the radioactive tracer will highlight the areas ofthe brain that are activated by viewing the image. This procedure ofboth scanning and participant's indications of motion will berepeated for the second image, the man frozen in midstep. For thethird image in which the man actually moves across the screen,however, motion will start and stop at irregular intervals. Thescanning and response indication will be repeated for this image aswell as for the kinetic and non-kinetic abstract images. The scanswill then be compared using the subtraction method to identifychanges in the blood flow in specific areas of the brain. The brainactivity in response to the non-moving stimuli is subtracted frombrain activity generated from the stimuli in motion.


    The design of the experiment will be a single-factor completelywithin subjects design. The independent variable will be degree ofmotion with five levels including a stationary figure, a figure inmidstep, a figure with on and off motion, a kinetic, and anon-kinetic image. The study will measure two dependent variables,both the brain activity recorded from the PET scan and the subject'sresponses to perception of motion by pressing a button. The studyexamines 3 hypotheses. Based on Zeki's results, the first hypothesisstates that the real on/off motion condition will elicit more brainactivity in area V5 than in the other figural images. The secondhypothesis states that the kinetic image will generate more brainactivity in area V5 than the other, non-kinetic, abstract image. Thefinal hypothesis states that the participants will indicate a higherdetection of motion in the on/off condition and the kineticcondition, thus matching psychophysical and physiologicalresponses.

  • Results
  • A likely outcome is that an examination of the results willsupport all three hypotheses. The brain activity will be analyzed andit is probable that subtracting the brain activity generated from thestatic figural images from that of the on/off motion condition willshow activation in area V5 as well as area V3. Another likely outcomeis that subtracting brain activity elicited from the non-kineticimage from that of the kinetic abstract image would also showactivity in areas V5 and V3. The activity, however, will likely beless than that generated from the on/off figural image. Finally, itis probable that participants' indication of detection of movementwill be highest in response to the on/off motion condition and thekinetic abstract condition rather than the other conditions. Theparticipants' responses will be analyzed using a single-factoranalysis of variance with a probability level of .05. Thus, thepsychophysical results will match the physiological results and thisstudy will be another example of how studying perception at thepsychophysical level can provide information about the physiologicallevel of analysis (see graph on attachment in which the label yescorresponds to actual movement).

  • Discussion
  • The likely results will clearly expand on Semir Zeki's body ofwork. Zeki chose to experiment with a figure such as the Moirepattern. Although this figure gives a sense of motion, it is a staticrepresentation of movement. Zeki's study, then, is comparable toStage I of the Futurists movement, representing movement in a staticfigure. This study will correspond to Stage II by incorporatingactual movement into art. If one follows this line of thought, thenfuture research should push the envelope and attempt Stage III, thatis, separating motion from form. An engineer can redesign Tinguely'sHomage to New York, and a PET scan can record brain activitybefore and after it self-destructs in a blaze of glory, safelycontained within a firewall in the laboratory. Less zealousresearchers could record such a spectacle and then measure brainactivity with a PET scan while the participant watches it onvideotape.

    For future research, Skoyles (1997) suggests using PET to examinethe response of the motor cortex of subjects as they view classicalGreek sculptures from the period of anatomical realism. Skoylesproposition was sparked by the discovery of bimodal neurons in themonkey premotor cortex that fire while performing a hand action andwhen watching another monkey or human perform the same hand action.Since the sculptures of the period portrayed taut muscles andrealistic postures, then corresponding neurons in the human cortexmay be recorded using a PET scan.

  • Reference Page
  • Garcia-Cairasco, N. (1998). Art and Science. RetrievedMarch 5, 2002 from http://www.epub.org.br/cm/n10/opiniao/cairasco/art_i.htm

    Goguen, J. A. (1999). Editorial Introduction: Art and the Brain.Journal of Consciousness

    Studies, 6(6/7), 4-15.

    Kunz, G. A. (1997). The aesthetics of scale. Artweek, 28,12-13.

    Leontiev, D. A. (1992). The perception of art: Psychologicalmechanisms, factors, and processes.

    Journal of Russian and East European psychology, 38(3),45-63.

    Moss, L. E. (1999). Nuts and Bolts. American Artist, 63,62-65.

    Skoyles, J. R. (2002). Motor perception and art. RetrievedApril 5, 2002, from http://www.skoyles.greatxscape.net/mdart.html

    Sudol, J. (2002). Can Art be Engineered?. Retrieved March5, 2002, from http://www.rpi.edu/web/campus.News/may01/may_14/home2.htm

    Tibbetts, P. (2001). Inner Vision: An Exploration of Art and theBrain. Quarterly Review of

    Biology, 76(3), 389.

    Zeki, S. (1998). Art and the Brain. Daedalus, 127(2)71-104.

    Zeki, S. & Lamb, M. (1994). The neurology of kinetic art.Brain, 117, 607-636.