Mask Design and Human Perception:   Using our Senses to Make Better Masks

 by Wanda Strukus

  Buy other USITT Publications here:

Dressed for the Photographer : Ordinary Americans and Fashion, 1840-1900

Julie Taymor: Playing with Fire

Costume Design : Techniques of Modern Masters

Cutting for All! : The Sartorial Arts, Related Crafts, and the Commercial Paper Pattern : A Bibliographic Reference Guide for Designers, Technicians,

Footnotes:

 of sensation and perception; which is to say that the first thing that happens to the spectator is that she or he has a neurological response to the lines, edges, textures, shapes, and colors we place before them.  Almost immediately thereafter, the spectator will begin to have what we call cognitive responses of memory, emotion, and reaction, but none of these higher cognitive functions can occur until the object is perceived.  Because of this simple fact, it is worth turning our attention to how our brain processes and organizes visual information and to the kinds of visual information that promote the greatest amount of stimulus.   It seems a foregone conclusion that the more we know about how we see, the more specific and calculated our design strategies can be, and the greater chance we have of engaging our audience on a profound level.  We cannot control how an audience responds to our work, but we can thoughtfully and knowledgeably give them work that will engage them perceptually, with the belief that the more they look and want to keep looking, the more they will feel.

            We make our creative decisions based on a wide range of factors.  The world of the play, the character traits of the particular role we are designing for, our knowledge of design elements such as line, shape, texture, and color, our prior experience, and the vision of the director all influence the choices we make.  Part of our job is our ability to explain these choices.  We are often called upon to articulate, for example, why we have decided that blue is the appropriate color for an element.  Perhaps our choice is in response to mood and atmosphere, or perhaps it addresses the more pragmatic concerns of time of day and the other colors present in the visual scene.  These choices are made because of our knowledge and experience, but at certain times they seem to come from a place of "intuition".   Because of our body of knowledge and our highly exercised creative impulses, we often know what "works" or what is "right" simply because we know.  In the converse situation we know that something is "wrong", and we parse through our memory and body of experience to identify the errant factor.  Usually we are successful, but sometimes we are not, and we send something out on stage that is fine, but we are left with a nagging sense that we missed something somewhere.  An increased awareness of how our visual system works can provide some answers for why some of our design choices work better than others, and can also help us to pinpoint where we have gone wrong and why.

            How do we know what will provide the audience with an engaging perceptual experience?  We begin by observing our own perceptual responses.  I have my own perceptual experience as a means of measuring how an audience might perceive a mask, and this is one of my most powerful resources as I design.  The more I know about my own perception and my own processes of visual organization, the more I can grasp about what I am making, and the greater control I have in terms of effectively communicating my ideas.  We will find that many of the laws governing visual perception are familiar to us because they share striking similarities to basic design principles.  It should be unsurprising that our visual design elements are inextricably linked to the way we see; my goal here is to bring those links into our working space so that we can use them not as coincidences, but with intentionality.  By taking perceptual laws out of the esoteric arena of the unconscious or intuitive and into a place where we are consciously aware of them, I think we can learn to facilitate the moments of "inspiration" that we all value so highly, because we have a better understanding of how our own visual processes work. 

            The concepts addressed in this paper call upon some of the basic physiological processes of the brain, with particular emphasis on how the brain organizes visual information.  The organization of visual information has direct implications for mask design because on a fundamental level, a mask is simply visual information --shapes, lines, edges, textures, colors--that the brain needs to organize into an identifiable object.  This focus on physiological processes does not undermine the importance of the "content" of the mask.  Certainly we long for more than an intense perceptual experience in the theater; we want to be moved emotionally, intellectually, and sensually as well as perceptually.  If we think of "content" as the emotional quality or qualities that are delivered by the mask in performance, we should acknowledge that this emotional aspect is largely inseparable from the visual aspects of the mask.  We do not perceive the form of the mask through our visual system and then perceive the emotional content through one of our other senses.  The personal responses or feelings that the audience experiences derive directly from how much they perceive visually--the more we see, the more we feel.

            In this paper, I give a general overview of how perception theory and design principles overlap, and then focus on two particular aspects of perception theory, geon structural description theory and the theory of cue-conflict, to identify specific ways in which we can apply the work of cognitive scientists to our designs.  I do not advocate a rigid adherence to cognitive science as a means of designing, but I do think that under certain circumstances and within certain models the laws we find in perception theory are incredibly useful to us.  What follows is an outline of the three conditions or assumptions that I use when applying perception theory to mask design:

 1.  What is a "good" mask?    

            We already have a general set of criteria that we use to identify a good mask.  A good mask fits well, has good visibility and ventilation, distills and defines character using design elements, and is playable; it is able to take on a variety of expressions in performance.  In addition to these factors, I define a good mask as one that prolongs or extends the length of time needed to perceive it, or that makes us want to keep looking at it.  When we look at something for the first time, we really "see" it.  Shortly thereafter, we assimilate it and it loses its perceptual impact and often our attention. We might know that the object is still present in our visual field, but we ignore the sensation of it.  This is a natural tendency on our part as human beings.  We have to assimilate and ignore some of the sensations affecting us or we would never get anything done because we would be busy having intense perceptual experiences of the details of our world (imagine if every morning when you awoke you were struck by the perception of the texture of your ceiling, and that texture continued to stimulate your visual senses for as long as you looked at it.  Why would you turn away or bother getting up?).  As natural as this kind of assimilation is, it is not what we want to happen to the masks we design.  We want our audience to keep seeing it and to keep responding to the visual stimulus the mask provides, and not simply see it, assimilate it, and forget about it.  We make masks for many reasons but regardless of each mask's particular function, our first order is to make the mask visible, and to prompt our audience to continue to see it as if for the first time.  We can think of this as making the making the familiar unfamiliar.  We take the familiar, ubiquitous human face and replace it with an unfamiliar object: a mask.  This notion of making familiar unfamiliar was used by Irwin Piscator, called alienation effect (verfremdungseffekt) by Bertolt Brecht, and was even earlier espoused by Russian formalist Viktor Shklovsky, who stated that the purpose of art is to make the stone stony, to make us see the object in all of its perceptual texture, shape, detail. 

 If we start to examine the general laws of perception, we see that as perception becomes habitual, it becomes automatic.  Thus, for example, all of our habits retreat into the area of the unconsciously automatic; if one remembers the sensations of holding a pen or of speaking in a foreign language for the first time and compares that with his feeling at performing the action for the ten thousandth time, he will agree with us.   Such habituation explains the principles by which, in ordinary speech, we leave phrases unfinished and words half expressed.  .  .  .  .   By this "algebraic" method of thought we apprehend objects only as shapes with imprecise extensions; we do not see them in their entirety but rather recognize them by their main characteristics.  We see the object as though it were enveloped in a sack.  We know what it is by its configuration, but we see only its silhouette.  The object . . . fades and does not leave even a first impression; ultimately even the essence of what it was is forgotten. . . .  The process of "algebrization," the over-automatization of an object, permits the greatest economy of perceptive effort . . . [1] 

Shklovsky describes a common occurrence in our perceptual experience of the world.  We move about our world surrounded by a sea of objects.  Once we have seen and experienced an object, we rarely give it the same attention again as we did that first time.  How many of us have noticed a scratch on our car and wondered, "When did that happen?"  We may see the car every day, but we don't give it our full perceptual attention.  However, there is the occasional object--often an artwork--that seems to jump out at us, engaging our attention no matter how many times we have seen it.  That response is subjective and personal.  But some aspects of the way we respond to the work rely on the way the object's design engages our processes of perception.  This stand-out quality is what we want to achieve when we make a mask, not because we want the mask to call attention to itself to the detriment of the rest of the production, but because the perceptual significance of the mask is going to increase attentiveness and engagement in what the performer is doing, which is presumably a goal in any masked event.  We want to avoid "the greatest economy of perceptive effort" identified by Shklovsky, and instead increase the effort needed to see and experience the mask.  A good mask is one that engages us and makes us look and re-look.  It keeps the neurons in our brain firing, and keeps us engaged on an experiential, perceptual level rather than an intellectual or analytical one.

 2.  The multi-dimensionality of masks made for multiple views.

Many masks are designed with the idea of frontality: the mask is primarily viewed from the front with some thought given to profile or three-quarter views.  Most of my work has been for

theater-in-the-round, and/or for directors or choreographers who need more versatility than a frontal mask provides.  As a result, I design masks that are visible and must be playable for their full three hundred sixty degrees.  If we assume that the back of the mask is going to be seen and played, that the top of the mask is going to be seen and played, and that every possible angle and view of the mask must engage the audience, then we have to embrace the idea of making the mask perceptual engaging and identifiable from all possible views.  This means that every area of the mask has to provide visual stimulus, and this visual stimulus needs to be consistent for the entire mask.    

3.  We construct what we see. 

            One dominant assumption in perception theory is that we do not really "see" anything.  We do not see a house, or a dog, or a mask, and there is no region in the brain (at least thus far discovered) where an image from the outside is projected like a film image on a movie screen.  The visual system is a process of taking in visual data from our three-dimensional world, transforming them into electrical signals, separating them into discrete elements and then somehow regrouping them so that the world becomes visually available to us.  The electrical signals that pass through our brains contain information about edges, line, lightness and shadow, color, texture, and shape, as well as orientation in space, and from this information our brains construct what it is we think we see. 

            We can illustrate this theory with the following images take from Paul Klee's notebooks (Figure 1).[2]  In the first image, we "see" horizontal lines.   We would not even necessarily call this a "grouping" of horizontal lines, but simply lines.  In the second image, we see a configuration of vertical lines.  When people are asked what they see in the third image, most say, "a city", or something similar.  Why do we see a city in the third image?  With very little modification, the third image is simply a combination of images one and two, where we identified nothing but lines.  In the lower or early perceptual areas of our brains we have neurons that respond solely to vertical line orientation.  These are simple neurons, and they do not multi-task.  If there is a vertical line, they fire, and if there is no vertical line they are dormant.[3]  In this same area, we also have neurons that respond only to horizontal lines.   Image three, resulting from the combination of one and two, necessarily involves the firing of at least twice as many neurons as the other images.  I say "at least" because the increased brain activity at the lower level cause activity at the higher regions of the brain, where information is organized.  We could "see" image three as a combination of vertical lines and horizontal lines, but we do not.  In addition, even if we want to see it merely as a collection of lines, it is very difficult.  Our brain, "likes" (for lack of a better term) to organize visual information into something that makes visual sense.  Image three is not an image of city, but that our brain constructs it as such. 

            The three conditions that I have outlined above describe the conditions under which we can best apply the following perceptual information to mask design.  We will start with some of the basic rules of perception, called the Gestalt Laws.  The Gestalt laws were founded early in the twentieth century and continue to influence theories for how we organize visual information.  These laws are the first layer of perception theory, and they are sometimes considered weak because they attempt to describe rather than explain what happens in perception.  None-the-less, they identify areas of concern and provide a foundational vocabulary for perception.  Gestalt laws focus on characteristic stimulus, which is the raw data presented by the perceived object.  Perception is explained via factors such as lightness, shape, color, size, and the spacing of small units that create larger stimulus patterns in the brain (just as the small units of horizontal and vertical lines in the Klee notebooks create larger stimulus patters to construct "city").[4]  The basic principle behind Gestalt theory is that the whole is greater than the sum of the parts (the city is greater than the combination of vertical and horizontal lines).  There are a number of laws, or "clues" for helping us organize visual information, but none of them alone are responsible for, or capable of, the organizing process.  While some of these cues carry more organizational weight than others, we cannot usefully say they are hierarchically arranged as the visual situation and our own cognitive processes will determine which cues come into play.   It is the context and the combination of laws or clues that lead to organization and identification of the object.  These laws derives from regularities that occur in our natural world.  As theater designers, we are somewhat less concerned with naturally occurring environments then we are with the artificially created objects and environments of the stage.  This being the case, we will be focused on harnessing and manipulating these laws, rather than simply observing them. 

            I have briefly outlined a few of the Gestalt laws that I feel have the most bearing on mask design, and have tried to point out how these principles of visual organization resonate in terms of design principles:

            1.  Pragnanz, also called "good form" or "simplicity".  This law states that we organize information into its simplest form.  It might also be called the law of least resistance in that if we see a combination of lines, or edges, we will construct the combination as shape that is simple and familiar, rather than as something difficult and unlikely. This law is what allows us to use the concept of less-is-more in our designs, and enables us to work with the idea of essences or ideograms.  Our visual systems are designed to construct familiar objects out of what we see, not to ignore and dismiss arrangements of lines and shapes as a meaningless jumble.  Our brain's tendency to organize information simply and easily also provides us with the phenomena of "cue-conflict".  Sometimes the visual data available to us can be easily organized in more than one way.  In this kind of situation the brain can construct two completely different versions of what we think we see, and both versions are constructed with equal ease and familiarity.  An example of this is found in the well-known figure/ground problem of the Edgar Rubins face/vase (Figure 2).  Some viewers identify the central area of the image as the face, and others identify the left and right sides as face silhouettes.  Most people can construct both the face and the vase, but the critical point is that we cannot see (or construct) both at the same time.  Our visual systems can quickly organize the information as first one and then the other, but not as both.  Since the two constructions are equally simple, we have what we call cue-conflict or ambiguity: The brain cannot organize the information as definitively one object or the other, and so our visual system remains in flux.  This fluctuation leads to a lengthening of perception time and interest, and we will later explore how we can use the phenomenon of cue-conflict in mask design.

2. Similarity--similar things appear to be grouped together.  By similar things, we mean elements that have one or more piece of data (or design aspect) in common, e.g. shape, color, lightness, texture.  We use similarity with great regularity as designers.  We typically design a mask for a character with some sense of consistency.  We choose a dominant line--horizontal, vertical, diagonal, or circular--and reference that line as we complete the features.  There is no particular reason that we cannot make a mask that has strong vertical lines--a long sharp nose, upward-slicing eye brows, and upward slanted eyes--and also strong circular lines such as big round cheeks and drooping jowls.  However, the dissimilarity of the lines would establish a visual inconsistency in the character.  This visual inconsistency would ultimately effect how the performer works with the mask and how the audience responds to it.  If we do use dissimilar elements, it is for a specific purpose, such as to show the dual nature of the character.  However, since masks often require us to find the "essence" of the character--the core personality--in visual terms, the use of similarity gives us a clear, essential design that is easily read by the audience.  As we create the raw data for the character, the Gestalt law of similarity helps us keep the design clear and simple and provides perceptual information that is easily organized.

3. Good continuation.  This law can easily be equated with "line" in terms of design elements.  If we can connect points to form a straight or smoothly curving line, we do.  This law explains our ability to follow the flow of light in a painting.  In mask work, this law provides one of the links between what we design and how the mask performer works with the design.  Part of what connects the mask to the human actor is the continuity of the lines in the mask to the lines in the performer's body.  When we provide clear, strong, unambiguous lines, we give the performer access to the mask by providing clear information about the way the mask can "play."  We cannot determine how the performer will use the mask, but we can provide lines that make good continuation possible. 

            The three Gestalt Laws I have described provide a basic description for how people tend to organize visual information.  With regard to masks, we are particularly interested in how these tendencies are affected when the visual data is seen from multiple viewpoints.  How do these laws of pragnanz, similarity, and good continuation function when the mask is seen from the back, and from the top, as well as from the front and the sides?  The mask, like a face, has different features and surfaces, and the visual data we are given from the feature-filled front, for example, is going to be very different from the information provided from the back of the head.  Because we are interested in keeping the spectator's visual attention engaged in seeing all facets of the mask, we turn now to some additional theories that can aid us in this task.

            Geon Structural Description Theory (GSD), also called Recognition by Components theory (RBC) offers an explanation for how we see the three dimensional object, and coincides with some of the operative principles of design theory.  The focus of GSD theory is on how we recognize both natural and artificial objects in our environment, but it easily lends itself to theories about how we create objects (masks) to be recognized by the audience.  GSD assumes that a given view of an object is represented as an arrangement of simple, primitive volumes, called geons.  Geons are cubes, cylinders, wedges--basically three-dimensional shapes--defined by either straight or curved edges, parallel or expanding lines, and symmetrical or asymmetrical shape.  All visual objects can theoretically be identified in terms of their unique configuration of geons.  Figure 3 provides an illustration of some familiar geons and how they appear in everyday objects.  Irving Biederman, one of the primary proponents of this theory, has identified thirty-six such geons.  These thirty-six geons are enough, by Biederman's calculation, to yield one hundred and fifty four million objects made from three-components (his liberal estimate of known objects (object categories) in the world is 30,000).[5]  Whether or not there are thirty-six geons is not our primary issue; there could be any number of geons, but the point is that the number is finite, and that they are volumetric shapes that, when entering our visual field, cause a certain pattern of neurons to fire.  Because geons are viewpoint invariant, the pattern of neurons fire regardless of the attitude of the geon when it enters our field.  Whether I see the object from the front or the back, above or below, and whether or not it is rotated or inverted, the same pattern of neurons for object identification should fire in my brain.  It follows that an object made up of such viewpoint-invariant geons will also be viewpoint invariant, and will cause the same kind of brain activity regardless of the viewpoint from it is perceived.

            There is some dispute with GSD theory's applicability to everyday object recognition.[6]  For example, if we were to apply GSD theory to recognize a bird, we would have to acknowledge that a bird perched in a tree would have a different geon structure than a bird in flight, and that a bird seen flying in profile would have a different construction than a bird that we see flying over head.  We would have to include geons for wingspan for the flying bird, but these geons would be unidentifiable in the bird at rest.  Which geon structure defines a bird?  Since we are engaged in making objects rather than explaining how naturally-occurring objects are recognized, the bird-geon problem offers us not an obstacle, but a lesson in the complex use of geons for design purposes, and we will return to this issue and what we can learn from it momentarily.

            Despite disputes with the universal applicability of GSD theory to naturally occurring objects, we find that GSD is extremely strong when we consider it in terms of artificial or human-made objects.  One of Biederman's oft-cited examples to explain GSD theory is his nonsense object (Figure 4).  This is an object that Biederman created to make an obvious point about GSD theory, and it is true that this nonsense object, while perhaps not exactly identifiable, is probably recognizable from any viewpoint.  In other words, we might not know what it is called, but we would recognize it if we saw it again regardless of  viewpoint because of its unique geon configuration. By making an illustration of this artificial nonsense object, Biederman suggests the applicability of GSD to not only three-dimensional object recognition, but to sculptural object creation as well.

            It seems that when we create a three-dimensional art object, we like to use a process similar to GSD theory combined with similarity and repetition.   Biederman has noted that many human-made categories, such as lamps, have geon configurations that differ from one model to another (for example, a floor lamp versus a ginger-jar lamp) and these different configurations serve as a primary means of distinguishing objects within a shared category.[7]  We can take as another example the Standard poodle, (Figure 5) which is a naturally occurring animal that has been shaped and designed through human activity to have a geon configuration different from any other dog, and certainly from any other object.  More importantly, we could confidently assess this object as viewpoint invariant, no matter how unusual the view.  By cutting and grooming, a spherical geon is added to the natural shape of the dog, and it is added repetitiously, all over the dog, so that if only one limb of the dog were visible, or if we should find the dog suspended from the ceiling, we would easily identify the object as a Standard poodle. 

            The Standard poodle is a somewhat irreverent application of GSD theory to the creation of the artificial object, but it serves to illustrate something frequently observable in human-made sculptural forms, ranging from architecture to artfully arranged bowls of fruit.  We like to create forms and see forms that approach viewpoint invariance; in part because of the ease of recognition, but more importantly, because viewpoint invariance maintains the firing of certain neural patterns in our brain.  If we consider a building that has a facade characterized by columns--a long, symmetrical, cylindrical geon configuration--the pattern of neurons that respond to such cylinders fire in our brain.  This enables us to identify this building as distinct from other buildings and engages our perceptive attention.  If we walk around to the side of the building and discover that the promise of the facade is not realized in rest of the building, and that the sides are smooth and square, the pattern of neuronal firing initiated by the column-filled facade ceases.  Certainly, we still have neural activity with regard to the texture and horizontal or vertical lines, but this activity is entirely dissimilar to our first experience.  The side of the building is going to stimulate less complex activity than the geon-filled facade, and relatively speaking, this side view will be dead or empty perceptual space.  Perception theory assumes that we construct our world.  If we look at the side of the building and there is nothing visible to recall the front of the building, then we cannot construct the front of the building from looking at the side.  This moment of dead perceptual space, and the inability to construct the identity of the object from an alternate viewpoint  is precisely what we want to avoid with the creation of the mask. 

            Of course, viewpoint invariance is not always our objective when we create or enjoy visual objects.  If we return to the above-mentioned bird geon problem, we can observe that we enjoy seeing forms that are not viewpoint invariant because of the surprise they offer us.  A robin sitting outside your window has a simple and specific geon configuration of a sphere (head and body), a triangle (beak) and perhaps two lines or thin cylinders (legs).  When the bird launches into flight via the sudden addition of two triangular wing geons, we have a quick and unexpected stimulus to our visual system that often results in feelings of surprise.  Cats, which perhaps do not have a good memory for geon structure, are invariably astonished and dismayed when their spherical prey sprouts extra geons and flies away.  My point here is that we do not necessarily want to dismiss a geon structure that is not viewpoint invariant, but that we want to use such configurations for specific reasons.  Another example would be a character whose front view reveals the costume of a formal, proper, puritanical minister, with simple, clean lines and no extraneous geons to his human form.  If the back view of this same character reveals a devil's tail, we are amused and delighted by the surprise and irony.  We would call the tail (as we would the wings of the bird) a viewpoint or circumstance-dependent geon: we do not see it unless we get a particular view of the character, and we use it appropriately to reveal another "side" to this character.  There are times in mask design that we may want to reveal a dual nature to the character, and in these cases, viewpoint-dependent elements are useful.  If duality of character or surprise is not on our agenda, however, we are more successful in creating a unified mask if we follow the impulse to create viewpoint invariant objects.

            If we use Figures 6, 7, and 8 as an experiment, we can see how using GSD theory to provide viewpoint invariance might be applied more directly to the mask in performance.  Figure 6 shows a mask for the character Kapila in the play Hayavadana.  Kapila is strong, athletic, muscular, good, and simple.  He is neither bright nor handsome, but he has magnetism.  I conceived broad and circular design for this character.  I was consciously working with  GSD theory when I designed the mask.  I went through a typical brainstorming design process, asking myself: What is the essence of this character?  What are his character traits, and how can I reveal them sculpturally in a mask?  What kind of recurring design motif can I use?  I was thinking in terms of recurring motives, not of geons, which is to say that my visual thinking was analytical, and not consciously perceptual.  Even though I was not thinking about geons, the questions I asked myself, particularly those regarding visual motifs, engaged a thinking process that implicated GSD theory.  If we interpret my search for a visual motif as a search for a geon motif, we observe another one of the strengths of GSD: it speaks to patterns and ways of thinking that we already use, and the questions that we ask of ourselves in our design process are exactly the same questions that we need to think in terms of GSD theory.

            Because this mask was going to be used for theatre in the round, I could not think of the mask as frontal.  I knew that every angle of the mask would be seen by someone, and that every person in the audience would see an angle or side that was different from their neighbor's view.  Part of the excitement of theater-in-the-round is the fact that everyone has a different view of the performance, but we do not want these different views to result in completely different experiences.  The mask had to provide consistent visual stimulus from every angle presented.  I did frontal, profile and three-quarter view renderings so that I would be sure to keep the bulging circularity of the design constant.  As I sculpted and later painted the mask, I continued to check it from multiple viewpoints.  I erred, however, in not making a  rendering of the back of Kapila’s head, and in being less than conscientious about the stimulus that would be provided by this back view.  I knew that there would always be some people in the audience who would see the back of the head, and that for the period in which they saw this view, it would be the defining characteristic of Kapila in their visual field.            Despite my knowledge,  I ignored the potential problem.  Kapila would wear a hat, and the back of his head would simply be the hat.  The hat would have a color and a pattern, and this color and pattern would provide the visual stimulus.  The back of the head, even for a human actor, is not particularly engaging, so why should a mask be any different?  By thinking this way, I set up a situation that was tantamount to the one created in my example of the building with the facade of columns and the geon-less side and back.  My design was saying that we are not supposed to "see" the back of Kapila's head, just as we are not supposed to see the back of the building.

            The back of the Kapila mask was fine.  If we look at Figure 7 we understand, intellectually, that this is the back of Kapila because we have seen the hat from the front.  While there is nothing particularly wrong with this solution, as I sat and watched the show during tech week, I knew that the mask was not “right.”  As I watched Kapila from the front, I was engaged in perceiving the shapes of his face; they repeated one another, worked together, and created a playable mask for the performer.  I had the same experience when seeing the mask in profile and in three-quarter view.  From all of these viewpoints I was engaged not on an intellectual level but on a sensation/perception level.  But as soon as Kapila's back was turned toward me, I could feel my perceptual interest wane.  It is not that I did not know it was Kapila.  Of course I knew who the character was, but the perceptual experience of Kapila disappeared in the absence of the circular, character-defining geons.  I could not construct Kapila from seeing the back of his head.  I needed add something that would cause the pattern of neurons that now meant “Kapila” to keep firing.  Figure 8 shows my simple solution.  I added circular loops, which suggest hair, but more importantly, satisfy the perceptual expectations I set up in the front of the mask.  The loops are circular geons that are similar to the circular geons that make up Kapila's facial features.  This is a case of using repetition of geons to make a mask more viewpoint invariant.  Using the revised mask, we can construct Kapila from all views and the blank spot on the object and on the brain is gone.

            The above example suggests something else about how we organize visual information and what kinds of things keep us perceptually engaged.  The problem with the back of Kapila’s head was that we could not construct the identity of Kapila by seeing it.  We solved this problem by adding Kapila’s distinctive features—the circular geons—to the area in question.  The back of Kapila’s head still had color, and the yellow and orange stripes of the hat are another design trait of the mask.  However, color alone did not, in this case, provide enough perceptual information for us to continue perceiving the identity of Kapila.  Color is a powerful element in design, and we use color to provoke an intense emotional response to what we see.  However, color is a surface characteristic, rather than a structural characteristic, and while it works effectively in conjunction with structural characteristics, it cannot bear the entire weight of identification and recognition on its own.[8]  If we remind ourselves that color is always the color of something: at the very least, it is the color of a shape, field, or area.  With the notable exception of the color of the sky, color defines a finite surface area or space area; we see the point or edge where one color stops and another begins.  Much or our process of "seeing" color depends on the adjacent or background areas of differing color, as well as on illumination and reflectance values.[9]  In the original version of Kapila, the colors on the back of his hat defined only the limits of the surface area of the hat.  The color enabled us to locate Kapila when his back was turned toward us, but could not actually help us construct Kapila.[10] 

            Was the Kapila mask as successful in practice as it is in theory?  One of the other masks used in the performance was a mask for the character of Devidatta, Kapila's friend and complement.  I designed this mask with a greater emphasis on vertical lines, because Devidatta is a poet and an intellectual.  I had the same problem with the back view of Devidatta as I did with Kapila, but I unfortunately was unable to come up with a successful solution.  As a result, the back of Devidatta presented dead perceptual space, which I was very much aware of during the performance.  No one ever said that they felt that the Devidatta mask provided inadequate stimulation, but many people did remark that they "liked" the character of Kapila better than Devidatta, that they enjoyed Kapila's presence, and that there was something extremely vital about Kapila.  They responded to Devidatta with admiration, but not with excitement.  The actors in both roles were good, strong, performers, and their work with the masks was exemplary, so I do not think that the weaker response to Devidatta can be attributed to a weaker performance.  I believe that Devidatta was less engaging because he was only engaging from the front, which means that every time we saw him from the back--and everyone in the audience experienced the back view with some

frequency--our perceptual interest waned.

            When we design almost anything for the stage, the first questions are, "What is the significant element?  What is the single line/form that will establish and define the essence of the object?" While a designer typically thinks in terms of line or form, this question might be practically answered by GSD theory.  We are looking for viewpoint-invariant parts that will prompt us to construct the object.  GSD theory tells us how an object can be more or less engaging by suggesting how an object can involve our perceptive processes to a greater or lesser degree.  A mask that presents perceptual dead space from certain viewpoints is likely to find the viewer less responsive.  Part of the "magic" of mask performance is the ability to keep the audience in an intensely perceptual and experiential moment.  The strength of GSD theory is that it gives us a convincing explanation for this magic, and also a provides a template for how to make this magic happen.

            We return now to our third assumption: we construct what we see.  We perceive edges, shapes, geons, lines, movement, and texture, and we assemble this information, combine it with memory and imagination, and come up with a coherent image of what the object might be.  We have simultaneity of bottom-up and top-down processes, which means that we have raw data out in the world that our senses are designed to extract, and at the same time we have cognitive processes of knowing, intellect, and memory that construct this raw data into some "thing" of the visual world.  Designers provide the material to be perceived in order to get the viewer to construct a certain object or face.  Line, shape, color, and texture are important because they provide visual "cues" that help us to construct what we see.  The Gestalt laws outline the basics for these visual cues.  Once we know what some of these cues are, we can start  experimenting with the idea "cue conflict" that I identified earlier with the Rubins face/vase (Figure 2).              Figure 9 is a painting by wilderness artist, Bev Doolittle, called Pintos.  This "camouflage" painting is an excellent example of manipulating cue conflict.  What makes it more than simply a good painting of horses is that the use of line, shape, and color thwarts our attempts to organize the objects in the scene.  What engages our perceptive attention is our inability to find a stable organization of the visual elements.  The pintos move in and out of the landscape, preventing us from separating figure and ground.  Our visual system wants to organize this image with stability because that is what it is designed to do.  The cue-conflict in the design of the painting holds our perceptual attention and keeps our brains active in looking at the object.  In terms of prolonging the time it takes to perceive an object, cue conflict is an extremely valuable tool in fulfilling our (and Viktor Shklovsky's) definition of the function of an artwork.

Because Doolittle's art is a significantly more complex example of cue conflict than that presented by the Rubins face/vase, it gives us a good idea of the sophistication of our visual system and suggests a model for using cue conflict in our work.

            Before we try to apply cue conflict to mask design, we should examine whether or not it is something that is useful for us.  As captivating as cue-conflict might be, does it not defeat our purpose if applied to mask design?  We are not creating a static image for a viewer to enjoy at his or her leisure; we are asked to design character-defining three-dimensional forms for performance.  Does a perceptual trick serve this purpose?  It is possible that cue conflict is something to be avoided rather than sought out?  We can assert that, tricky or not, cue conflict is highly effective in prolonging the viewer's perception time.  Our working theory says that prolonging the perception time is the goal of our art object, with the presumed result of greater emotional and intellectual engagement.  In Bev Doolittle's complex example of cue-conflict, the conflict is not simply a perceptual circus trick.  The Pintos are not camouflaged in a city or outer space; they are camouflaged in a landscape.  As our perceptual system fluctuates between the organization of the Pintos and the organization of their background, our cognitive processes may be prompted to consider the pinto's place in their landscape, and we may "see" the Pintos as inseparable from their environment.  Theoretically, the ambiguity provided by cue conflict better serves Doolittle's goals as a wilderness artist than an unambiguous painting of pintos in a landscape.  The ambiguity is greater than the sum of the two cues and we can hypothesize that well-executed cue conflict does not simply stay on the perceptual level.  It seems entirely possible that the intense brain activity caused by cue conflict stimulates the imaginative and interpretive areas of our brains.

            How can we use cue conflict to good effect in the making of masks?  Because we are dealing with objects rather than a painted canvas, our challenge is to translate the technique suggested by Bev Doolittle and by the Rubin's face/vase into a three-dimensional illusion.  The easiest tool at hand lies in our choice of materials.  Masks can be constructed from almost anything, and we can use surprising materials to add perceptual weight to our designs.  If we construct a mask out of found objects, part of our impulse to do so results from the associations we make (and thus anticipate that our audience will make) when we see these objects as part of a face.  These associations can occur only after they are perceived (I am not arguing for a linear mode of visual processing; our perception of shapes and objects happens quickly.  We then make a cognitive association, which may focus our perceptual attention on particular aspects of the object, which leads to further cognitive processes of interpretation and response.  This process continues in the what cognitive scientist Erich Harth calls "the creative loop"[11]). Figure 10 shows an illustration for a costume for a cook by Mannerist painter and designer Giuseppe Arcimboldo.[12]  If Arcimboldo's costume design were realized, and if it entered our visual field, our initial perceptual response would be prompted by the edges of the objects, the texture of the metal, the colors, the shapes, and the facial features.  The delight in seeing this object would likely come from a form of cue-conflict; what we see is a face and figure, and yet what we see is a pot and spoon.  We are free to see both, but not necessarily at the same time, and so our attention, and our perceptive mode moves back and forth between the two, thereby lengthening the process of perception and prolonging our perceptual engagement and our interest.

            Figure 11 shows the realization of three-dimensional cue-conflict.  The object is not a mask, but a performing object; it is the dismembered body of Pentheus that I created for a production of The Bacchae.  Pentheus is a good example of the grotesque, something we are often called upon to create in mask work.  The grotesque raises the issue of a balance between attraction and repulsion, and between beauty and horror.  It is very easy to horrify or disgust an audience, but if the audience does not want to look at what we have put on stage because it repulses them, we defeat our purpose of trying to extend their perception time.  The challenge of the grotesque is that it must be difficult to look at and extremely engaging at the same time.  The examples of cue conflict provided by Bev Doolittle's artwork and the Rubins face/vase can provide us with clues for how to approach this problem.

            For the Pentheus project, I had the grotesque challenge of creating pieces of a body that would be assembled on stage in an intimate arena space.  My first thought in the design process was: "What is the least I can show to make the audience construct a body?"  My second thought was: "What materials can I use to make the audience want to keep looking at something inherently horrible?"  I did not want to trigger an emotional response of disgust, but rather one of sadness.  I knew that I could hope to get the audience to experience pity only if I could convince them to look at the object for an extended period and thus have an intense experience of the body.  My goal was to get the audience to construct (loosely): "There are pieces.  These pieces form a body.  The body is Pentheus.  How sad."

            I needed to steer clear of artificial materials such as foam, wire, or latex because  although we can use these materials to create a "realistic" body, the play was not realistic, and space was far too intimate--most of the audience being no more than twenty feet from the action--for these lifeless materials to have a powerful visceral effect.  Our perceptive powers are very strong: we can and do perceive the difference between synthetic materials and organic materials, and the perceptual difference results in a different kind of response.  I decided to use materials taken primarily from nature; broken branches for bones, bark and hand-made, high-fiber paper for skin, and vines, leaves, raffia, and flowers for muscles, tendons, and organs.  The rational behind these choices was: 1) They were logical within the context of the play, because Pentheus's last living moments are spent hiding in a tree and spying on the Bacchae.  The organic materials would recall this moment;  2) These materials suggested a supernatural transformation of Pentheus from a human being into a non-human aspect of nature;  3) The form, texture, color, and line of these materials are similar to the form, texture, color, and line of different parts of the human body.  I thought that the similarity would prompt the basic perceptual function of association.  4) I suspected that this "natural" visual data might prompt the spectator to respond with a memory of something familiar and thus make another association:  These materials once contained life; this body also once contained life.  One way of giving animation to an object is by using materials that once contained and aspect of animation: life.

            My goal was to construct a body that was like a Bev Doolittle painting, and in doing so, find a balance between "trees" and "body" to create a situation in which there would be cue conflict.  I wanted to trigger perceptive and cognitive processes that would be constantly fluctuating over what it is that we are seeing.  How did I go about doing this?  First, I went for a walk in the woods to engage in a live experience of a Bev Doolittle design.  I had to see if I could find faces and bodies in the forest.  This task seemed hopeless.  For the first ten minutes, I was certain I had made a mistake; there were only trees in the forest, and as it turned out, trees did not look anything like bones.  But as I focused my attention on the task, thereby focusing my perceptive processes on trying to see branches as something other than what they are, I found them; femurs, scapulae, ribs, and a spiky stump that would become vertebrae.  I gathered these materials together and to make sure it was not just my own flight of fancy and that other people would see what I saw, I brought some of them to a meeting with my director.  I held up a branch, a shattered, knobby piece of wood, and said, "This is Pentheus's femur."  The director looked at the branch, adjusted his perceptual organization of the object, and said, "I get it."  I later showed him a concept rendering of Pentheus, but I wanted him to see the branch first to ascertain whether he could make the perceptual adjustment that I wanted the audience to make.  The shape of the branch and the shattered quality of the old, broken wood were huge factors in the success of this test.  This was the beginning of the balancing act of Pentheus's body.  I would create something that was neither definitively Pentheus, nor definitively forest debris, but something that would prompt the construction of one, and then the other, creating a perceptual tension and dynamic.  When the body was laid out, it was seen as a pile of branches, then a body, then a pile of branches, then a body, ad infinitum.  The object served its purpose of increasing the length of perception time, and keeping the audience constantly engaged in the experience of the moment.

                The question remains, why do we want to remain on the perceptual and experiential level with a mask?   Is there anything wrong with approaching our design analytically and intellectually?  The answer depends on what we are trying to do with the mask in performance.  If we are trying to achieve that special level of engagement that masks are capable of, we need to think about designing masks that ask more from us as designers and as audience members.  If we value the mask because it engages our imagination or  because it enables us to project ourselves onto the character, we probably need to think about what it is that makes this possible in a good mask.   Whatever the mystery of the mask, it begins with perception.  The raw visual data that the mask provides is different from the data provided by an unmasked performance.  When we say that mask performance is "interesting," what we really have to mean is that there is more activity going on in the brain, or at least that there is a different kind of activity going on in the brain than what we usually experience.  The more active the perceptual experience, the more active the cognitive and emotional experience, and the more active the cognitive and emotional experience, the greater chance we have of tapping into the magic of the mask.  When I think of some of the great masked performances I have seen, they have been explosive sensory experiences.  I have not left the theater thinking, "What good masks!"  I have left the theater in silence, speechless and simultaneously empty and full from a kind of sensory catharsis.  Later, after the performance is over, I can think back through the details: what happened, what the masks looked like, and how the masks were used.  I have not missed anything even though I was unable to analyze the performance as it was happening.  The perceptual experience of the performance was so powerful that is was branded in my memory.  I think this is our goal when we design and construct our masks.  We want to give the audience a profound experience to take with them, not an analytical exercise to undergo in the theater.  This is a tall order for the designer, but I think we can use knowledge of perceptual experience to bring us closer to this goal.