This book is primarily pratice oriented, aimed at those who wish to practice game design. The goal is go develop a psychological understanding of game characters. The emphasis on psychology comes in opposition to the method of designing game characters based on cinema and other passive and linear media. Because games are interactive, players interpret characters in richer ways, so a perspective anchored on social psychology is much more apporpiate.

Part I: First Impressions

The early discussion covers visual cues that affect the player’s creception to characters. The qualities that are explored are: attractiveness, baby face, and stereotypes. This is not about social roles yet, but about presentation and expression, which may or may not relate to player expectations. When player expectations are fulfilled, this creates a smooth and consistent experience, but when expectations are broken this can be surprising or confusing to the player. Broken expectations can lead to depth of experience, but if not done properly, the effect can simply be jarring.

This first section is on personality traits which are expressed by characters. These are studied with respect to the building of first impressions. The most important personality traits discussed are dominance and agreeableness. Characters can express dominance through pose and demeanor. Generally, player characters are expected to be dominant to some degree. Isbister cites Maslow’s hierarchy of needs as an example for how to think about how a player will react to other characters. Because safety is one of the more fundamental needs, dominance and agreeableness are the first elements of character that are read by the player. Dominance also resembles the idea of “status” that is derived from Johnstone’s Impro. Dominance and agreeableness may be seen in a spectrum relative to the player character. The attributes are relational, not absolute. Changes in agreeableness and dominance are evidence of significant changes, or are expressions of communicative symbols.

In addition to dominance and agreeableness, Isbister explains that there are five common personality traits that are common across cultures (pulling from McCrae and Costa). These are denoted by the acronym OCEAN, for Openness, Conscientiousness, Extroversion (dominance), Agreeableness, and Neuroticism.

Part II: Focus on the Player

The first section covers culture, from the perspective of design. Design is relevant from the perspective of the audience in terms of how culture is portrayed. The section is primarily oriented around thinking about the culture of the player, and seems to have an underlying goal of teaching designers how to think about culture for games that are going to be marketed in cultures different from the designers’ own. My interest extends beyond this a bit, thinking about cultural representations for characters, as opposed to the cultural world of the player.

Cultural differences are described in terms of:

1. Expressions and physical characteristics. Gestures and expressions have different values across cultures. This category affects the reception of signals cast in body, face, voice, touch, and social distance.
2. Social norms and expectations. Roles and expectations of the individual vary among cultures. For example, American culture has a strong ideology that everyone must fend for themselves, whereas within Japanese culture there is an expectation of interdependence. Roles have much to do with games because of the role of the player and the corresponding hero archetypes.
3. Local media contexts. When marketing to other cultures, one must be aware of the media contexts and traditions there. This suggests consideration of the cultural “grand” narratives which have formal and structural characteristics, in addition to ideology.

Isbister describes gender, but again from the perspective of thinking about the player. The discussion is aimed to encourage designers to think beyond the andronormative perspective that tends to plague the game industry. Gender is the element of sex that is learned in culture, not inherited. It is culturally defined and time dependent, variable, and constantly changing. Discussion describes 3 bullet points aimed at thinking about gender in design.

1. Play style: what girls like to do in games versus boys. This is an open topic, but has had much research. Isbister lists several findings (p. 112):
   • Girls tend to enjoy games that allow for open-ended play and exploration that does not necessarily require completion of one goal or level to get to the next (Gorriz and Medina 2000, Kafai 1996).
   • Girls play games with puzzles or mysteries over games that involve physical accuracy and acuity (Children Now 2001, Gorriz and Medina 2000).
   • Girls would rather spend their time creating things instead of destroying things (Gorrriz and Medina 2000, Bruner, Bennet, and Honey 1998).
   • Girls enjoy everyday life activities and metaphors just as much if not more than fantasy adventures (Subrahmanyam and Greenfield 1998).
   • Girls may be less comfortable than boys about just jumping in and exploring a game to learn how to do things; they may do better with more explicit mentoring and instruction at the beginning of a game (Subrahmanyam and Greenfield 1998).

   There is furthermore another set of bullets that discusses how girls relate to others (whether players or NPCs): (p. 114)

   • Girls prefer collaboration to violence against others in games (Gorriz and Medina 2000, Subrahmanyam and Greenfield 1998).
   • Girls tend to prefer working in smaller teams than boys do (Subrahmanyam and Greenfield 1998).
   • Girls enjoy forging relationships–visiting other characters, writing letters, learning about how other characters feel about what is going on (Subrahmanyam and Greenfield 1998, Greenfield, Bruner, Bennet, and Honey 1998).
   • Girls are interested in the story behind the story–motivations and interrelationships among characters (Gorriz and Medina 2000).
   • Girls enjoy communication with other girls, and games that encourage or incorporate chat and social activity in conjunction with the game can support this (Gorriz and Medina 2000).
2. Roles: comfort and fantasy. The essence of discussion on roles is to include a variety of roles for female characters. It is important to have a variety of gender roles that the player may explore and relate to.
3. Expectations: what is expected in interactions. It is important to have gender aware reactions, and especially not get it wrong. Having characters treat the player with the expectation that she is male is alienating and dissociative. Gender plays a strong role in everyday life reactions, and to ignore the differences in these reactions tends to encourage the expectation that the player is always male.

Part III: Using a Character’s Social Equipment

This part covers the means of expression of a character’s emotion, intention, and personality. The subsections are the face, the body, and the voice. I do not go over this last section in detail, but the first two are fairly relevant, especially from the perspective of simulation and embodiment.

The face is very important for psychological understanding, and the muscles and expressions that act in the face can be divided into action units (Eckman, Freisen, and Hagen 2002).  Isbister also discusses the four basic emotion types, which are so considered basic because they are universally recognized across cultures and share common physical expressions in the face. These are anger, fear, happiness, and sadness. The gaze is also read in the face, and the gaze can also have functional properties. One’s face is an expressive object which can convey emotion and personal state. Finally, faces are tied to emotional feedback, and this produces mimicry. Here are some bullet points that address the communicative symbols of the gaze: (p. 145-146)

• Dominance or submissiveness (patters of holding or avoiding direct eye contact)
• Where a person’s attention is at the moment
• Flirtation
• Interest in beginning a conversation (or desire to avoid one)
• An invitation for one’s conversation partner to take a turn in the dialogue
• Active listening
• Pondering of a point

Body language may be used in character design. Body language is an expression of embodiment, so describing body language formally and developing a symbolic language for the body is a productive goal. Isbister divides body languagee into four categories: distance, touch, imitation, and posture. The first element of body language is distance, how far people stand from each other. There are four types of distance:

1. Public distance (more than 12 feet)
2. Social distance (12-4 feet)
3. Personal distance (4 feet to 18 inches)
4. Intimate distance (less than 18 inches)

Touch falls into four functional categories. An interesting thing is that these are used fairly effectively within, for instance, The Sims. These are again culturally dependent, and will operate differently in modern America, versus modern Japan, or Regency England.

1. Function (a doctor’s examination, or having one’s coat removed by a servant)
2. Social ritual (a handshake)
3. Friendship building (friendly hug or a pat on the shoulder)
4. Intimacy (sexual interest or emotional connection)

Imitation is the quality of imitating someone’s posture or gestures. This occurs most frequently when the imatated person is more dominant, the rest of the group will mimic his or her behavior. Conversation also produces mimicry among its participants. Posture is very rich for expressive signification. Isbister references Gallaher (1992) to isolate four categories of posture. These correspond strongly with the dominance and agreeableness dimensions, and they also are affected by culture and gender.

1. Expressiveness (variety and energy in expressions)
2. Animation (energy in movement)
3. Expansiveness (occupation of space) — this is most strongly aligned with dominant personalities.
4. Coordination (smooth movement and grace)

Part IV: Characters in Action

The first section in this part discusses player psychology, and the types of engagement the player has with the game world. This relates to thinking of player character design as an extension of the player. Isbister outlines four categories of experience. These can be aligned with the player expectations, and to do so produces coherent and smooth gameplay.

1. Viseral is the sense of being in the world and having augmented engagement. Visceral experience in games is usually the type of experience that the player would not be able to do in real life.
2. Cognitive is the way that we map our own problem solving onto the character. When our plans map easily onto the abilities of the character, the expierience is smooth.
3. Social is the element of interpreting others socially, and enables the player to wear personal masks.
4. Fantasy is the capacity and appealingness of exploring new identities within the game world.

There are three types of player characters according to Isbister: tools (no PC), puppets, and masks. I think there are hybrids in this mapping. Puppet characters are like characters in platform games, where the character has a distinct identity, and expresses that identity through its behavior in the world. Masks are commonly used in MMOGs, where the player is given rich tools for customization and for performance of the character, especially via emotive expressions. This last category though gets ambiguous with first person games, and games like Fallout 3, which have a lot of mask like qualities (customization, decision making), but few elements of performance.

Discussing NPCs, Isbister uses the language of roles. Role awareness is important in interaction, and consistent role views and expectations create smooth experiences. Misunderstanding of roles will lead to trouble (both socially and within games). This can be productive (when one wishes to create tension), or simply jarring for the player. There are three elements to roles, which are interdependence, power dynamics, and obligations and investment. Interdependence is the objective and ability of the NPC. Even hostile NPCs provide interdependence by providing opposition and conflict for the player. Isbister reviews many common NPC types and discusses them all in light of these three categories, as well as by the defining manner in which they interact with the player.

Mindstorms was originally published in 1980, and later republished in 1993. A lot of time has passed since both editions. Papert is writing on the use of Logo as a cognitive tool, and encouraging the idea that computation can help the way that children learn and think. Having learned programming through Logo myself, I have found Papert’s conclusions to be remarkably resonant with my own work. In the second edition foreword, Carol Sperry, an educator, remarks on how Papert was hugely influential on education though this book and the promotion of Logo in the classroom. She observes that computation reveals the arbitrariness of language, as well as the variance of potential answers to questions. This book is not about teaching a certain method of thought, a straightforward way of investigating a problem and finding answers. Rather, the book explains how computers might be used to teach metacognition. There is a pervasive theme of embodiment, practice, and activity within the text, making it seem very progressive from a cognitive standpoint.

In the introduction to the second edition, Papert makes a fascinating and outstanding claim: debugging is the essence of intellectual activity (p. xiii). As he explains the issues, omissions, and lack of clarifications of some topics in the original edition, Papert describes these flaws as bugs. The goal is to represent procedural thinking (and literacy) generally, but not tie it down to androcentric or mechanical modes. Papert’s chief examples of learning through computers are about mathematics, but this was not meant to imply that math is the only or most important issue that could be addressed, rather, it was the most accessible example in his perspective.

There are generally two dimensions to my inquiries into cognitive science. One is the study of how the minds of simulated agents might work. The other is to understand how people might interact with simulations in order to learn, feel, and benefit from them. This book falls squarely into that latter category. Papert gives an example of how he played with gears in his childhood, and this helped him learn mathematics in a very tactile manner, referencing the embodied experience with gears in memory while understanding math problems, even at an emotional level. The example of the gears is a review of learning based on practice. This is practical embodied experience working with a system. This understanding is continued and extended into other domains. This sort of argument would, on the surface, support the classic position of knowledge transfer in cognitive science. However, an alternative perspective is that the transfer is metaphorical. The study of learning involves the genesis and origin of knowledge, which is derived from Piaget. Papert studied under and worked with Piaget, so this influence makes a great deal of sense. The understanding of knowledge is cast in terms of models.

Introduction

The proposal is that computers can help people change (and expand) the way they think. Papert argues that the reason why people are limited in the way they think comes from cultural and economic obstacles. One such obstacle to spreading of scientific knowledge is that people see scientific objects as belonging to others. Frequently, scientific and technical objects may be available, but their operation remains closed and forbidden. A very similar argument can be made with literature. Papert argues that computers will penetrate where other approaches have failed. And, in terms of the pervasiveness of computers, this is fairly accurate.

The political element here is very notable and important. Ultimately, this is about expanding mental models. Lakoff would ague that these come from metaphors which are also often very political, and connect to cultural modes of thinking. Papert makes a metaphor of scientific learning as compared to the learning of foreign languages. The comparison is that learning best occurs via immersion than classroom experience.

Piaget argues that children build their own intellectual structures. They learn without being taught. This may be considered or compared to model formation. This idea connects  nicely to Lave in terms of mathematics. Children build with things from the culture, those things are meanings. Meanings understood in this way can also be seen as symbols, specifically cultural symbols. People do learn math practically, but they do not see it as math, because it is not approached the same way that math is taught in the classroom. Papert attributes the difficulty in learning math and science to cultural factors, which is a difference from Piaget.

The turtle in Logo is an object to think with. They are embodied and projected references that help make use of real world understanding and experience. Children can use the turtle easily because it resonates with the way that they might think of a tangible object. Papert spends some time defending the idea of teaching children programming, as programming was generally considered (in the 1980s when the book was first published) to be a complex and difficult skill, best handled by experts.

Computers and Computer Cultures

Two types of thinking are described by Piaget: Concrete and formal. Concrete thinking is more tactile and embodied, whereas formal thinking relies on abstractions. Papert claims that the computer can concretize and personalize the formal. (p. 21) I might argue that this would be because of the computer’s representative and simulative power.

Papert discusses some of the fears of the negative effects of computers. Common fears are that computers would brainwash children and program them, or plug in and be totally engrossed and cut off from the rest of the world. Papert is optimistic in the face of this criticism, and I think that is because he seems to see procedural commands and constructive activity as the default mode of interacting with computers. The historical result is naturally a mix. Computers are pervasive and growing, but interaction sometimes is literate and other times is not.  Papert notably does not argue that the ideal future is inevitable or a given, but that computers can be positive. More positive developments require children to learn epistemological reflection (metacognition), which requires an intervention in the way that education works fundamentally.

Mathophobia: The Fear of Learning

Poetically, the trouble that adults have with understanding the world of children is described as that adults “forget” the wonder of what it is like to be young. Instead, it seems that what is the case is that children do not have strongly formed models of how the world works. Papert discusses the fact that children must learn the conservation of volume in pouring water from a wide glass to a narrow one. The idea that volume is conserved is a model, and a view of the world. Children’s worlds are open, and no models have set in yet, yielding a certain flexibility. Because children are learning and observing constantly, they are more comfortable with the diversity of and incompleteness of models. Maturity is essentially the entrenchment of established models, and the cutting off of others. Instead of seeing adulthood as a state of forgetting childlike innocence, adulthood may be seen as the reduction of the world to a select few models.

Turtle Geometry: A Mathematics Made for Learning

Papert looks at “styles” of mathematics. These are differentiated: Euclid’s mathematics is axiomatic, Descartes is algebraic, and Logo’s is computational. These ostensibly represent the same abstract domain of math, but their perspectives, methods, and metaphors are all very different. The formulation of math as styles relates to the authority of knowledge, that there is one right way to view a situation or a problem. Logo is explicitly playful, and it encourages freedom within its computationally bound space. Interestingly, Logo’s approach to math is differential: the computational manner of drawing shapes relies on changing values with respect to one another, and taking small steps. This approach is the same approach to math used by calculus, but works in a concrete and embodied sense, not a formal one.

Languages for Computers and for People

The emphasis here is on being open to understanding things partially. It is important to acknowledge the value of partial understanding, and not expecting to have the full knowledge of a system. The educational system encourages a view of the world as full of right and wrong answers, and teaches people to fear and reject mistakes.

Microworlds: Incubators for Knowledge

Microworlds are Papert’s term for a constrained system which has some interfaces for play. Working with and constructing microworlds is analogous to theory building. Building incorrect theories is important, especially as false theories are transitional to building better ones. The view that theories are true or false is pervasive, and is a major obstacle to teaching science which is mistakenly thought of in this way. Logo encourages children to build and test their own theories, and Papert describes a method for simulating Newtonian mechanics. Even working with an incorrect model reveals the way in which objects ascribe to physical principles.

This book does not seem to describe or endorse a particular theory of narrative on its own, but it does give a historical account of narrative developments and theory, focusing on modern theories especially. In this case modern generally means post 1960, after the emergence of the poststructuralist movement. It may be useful using this to understand the historical influences that shape the perception of narratives and novels as story worlds.

Introduction

Theory of narrative is tightly coupled with theory of the novel. Before narrative theory was explored in general, the topic of study focused on the novel, which is relatively recent as far as traditional media go. Study of the novel primarily seems to have come from formalism and New Criticism. Martin argues that the shift in study to narrative as a whole, including the complexities and ambiguities that arise from such a change of focus, is indicative of a Kuhn-like paradigm shift.

Theories in general have much to do with historical circumstances as with the subject matter itself. The theory of narrative and the novel has a lot to do with the trends and influences in psychology and philosophy. Early theories of the novel were preoccupied with realism, form, and moral situations. When mass production and mass society arose, the conventional class structure was broken down, and it was believed by many that the novel would die out. This is an interesting prediction in pre-1960 narratology, and it is indicative of other fears of catastrophe spurred by social changes. The novel did not die out, but it did change from its social realist upbringing and took on new forms. Beat generation fiction, “new American Gothic,” and other literary movements took the place of realism instead.

More recent studies (as in the first half of the 20th century) use a structuralist approach, and view novels and narratives as ascribing to some mythic structure. This focus is aimed at novelists who broke from the realist tradition, but could still be viewed as representing the monomyth, albeit in many ways. Northrop Frye and Wayne Booth both form alternate views. Frye broadens fiction beyond the novel, into four categories: novel, romance, confession, and anatomy. These categories could be blended to create some hybrid forms. Booth argues for the study of rhetoric in fiction, arguing that fiction is inherently representative and rhetorical.

Martin includes a diagram that illustrates the elements of narratives and the axes which connect them. Different axes are the critical focus of different theories. Martin argues that the multiplicity and variety of literary theory is its strength. This suggests that the full picture and all axes are important.

Martin's narrative axes

From Novel to Narrative

Frye’s work on fiction aims to classify and form categories. Classifications are problematic because they aim to compartmentalize and judge works based on certain determinants and criteria. Frye’s schema is useful because it removes barriers between poetry and prose. The classic and artificial categoires are drama, poetry, and fiction. These are formal differentiations, having nothing to do with the content of the narrative itself.

Frye's narrative modes

Frye's narrative categories. These represent the four main categories and their hybrids.

The result of this discussion is that the diversity of narrative forms within the space of the novel leaves the holistic conception of the novel in ruins. The diversity challenges the understanding of the novel alone, instead the study changes its focus from the novel to narrative. When conventions and restraints are applied, the novel works to push these boundaries and break through them. Much like technology and art, the novel opposes itself to poetics and defies its constraints. The study of form is therefore at odds with the understanding of the medium’s history. These ideas were studied by Shklovsky as literary defamiliarization, and then picked up by Bakhtin.

From Realism to Convention

The novel itself is generally wrapped up as being realistic, and perceived as potentially real. This term is ambiguous, though. One perspective, from James, is that the novel should be written and read as a history. Believability and realism are tied to the attitudes of readers, which tend to fall under three categories: credulity, credence, and skepticism. Wellek and Becker see realism as a period concept, characterized by typicality, objectivity, and causality. For a narrative to be perceived as realistic, it must contain typical circumstances, be understood objectively (in the sense of authorial distance), and events must be connected causally.

Another element of realism is motivation, which indicates the inner lives of characters. The course of narration plots out what happens to characters as a result of motivation, causality, and the author’s intervention. It is in this case that the realist novel conducts itself much like a simulation. The characters are motivated agents, and the world they inhabit obeys causal rules. The author may intervene, but in a limited manner. As a result, unexpected emergent events may occur. That a novel may contain events unexpected to the author is a surprising claim, but is consistent with the view of narrative as simulation. The example given is with Tolstoy’s War and Peace.

Tolstoy decided to create a character who would be killed in a battle. Even if he exists only to die, the character must first be created and endowed with traits that make him interesting; in this case, Tolstoy made him brilliant. “Motivation” requires that such characters be firmly woven into the texture of the novel as a whole, as Tolstoy indicated in a letter: “Since it is awkward to describe a character who in no way is connected with the novel, I decided to make this brilliant, young man the son of old Bolkonsky” (a character important in the chapters that follow the battle). This puppet, born only to die, took on a life of his own. Wars do cause pointless deaths, but they are doubly pointless if they simply illustrate, once again, the horror of war, and the character involved has stimulated but not satisfied our curiosity. “He began to interest me,” Tolstoy wrote; “a role presented itself for him in the further course of the novel, an I had mercy on him, severely wounding him in the place of death.” The character’s survival led to events in the novel that Tolstoy had not originally planned, which themselves required further explanation. This process of motivation, which was well described by Victor Shklovsky and Boris Tomashevsky in the 1920s, is similar to what Frye calls “displacement.” But in Frye’s account of creation, the writer starts from a traditional, archetypal plot (such as is found in myths and romances), and then “displaces” it from its dreamlike unreality to make it plausible from a realistic point of view (134-40) (p. 65)

Narrative necessarily concerns the past. This is necessarily the truth because of the fact that the narrative is written and by the very process of narration. It works in a manner reminiscent to linear games, where local agency falls in the category of narrative compression and omitted inessential details. Histories are narratives and follow biases and trends. Realistic novel is therefore a kind of history.

Narrative Structure: A Comparison of Methods

There is a useful discussion of character here. Character is inseparable from fiction. This is a point that James, Propp, Tomashevsky, and Barthes all would agree on. However, in modern narratives, character and motivation are more important. Instead of characters being substitutable for one another (in the case of Propp), it is the action which may be substituted. Characters remain preserved. This influence makes sense in the sense of variance. In Propp’s account of the folktale, the formal structure is preserved, but characters vary across a bewildering diversity of characters and situations. In modern serial fiction (a great example is sitcom TV shows), characters are constant, but the situations and circumstances vary. The plot of the episodes may remain structurally similar, but the appeal to the audience is the characters reaction to the new situations. This idea is very relevant to the perspective of adaptation as well, as it describes what should be the focal point of the adaptation, plot or character.

The discussion of character and plot gives way to subtle hints at the notion of the story world. In the example of Huck Finn, we view the world more clearly through a flat character. Huck is a flat character because he is not complex and does not grow. As a result, the world that we see through his narration is less distorted. Beyond flat and round characters, there are static and kinetic characters. Kinetic characters may weave in and out of the plot because they occupy different spheres, beyond the sphere of the narrative. This indicates that the narrative itself is a limited view of a larger reality. If all characters were present in the narrative, then the story world would be the entire world that the reader observes. However, when characters exit, the story world is revealed to be larger than the view indicated by the narrative itself. The absence of these characters from the narrative is a point where ambiguity and openness may enter.

From Writer to Reader

Another useful diagram, illustrating the spectrum of understanding the dimension between the reader and the writer. This may be seen as a more detailed view of part of the diagram of narrative axes shown earlier.

From the author to the reader. Barthes would be proud.

Blumer was a strong follower of George Herbert Mead, clarifying, expanding, and extending Mead’s sociological psychology and philosophy into empirical work. This text works well as a follow up to Mead’s Mind, Self, and Society, because it makes clear many of Mead’s points and applies them practically to the study of interaction. This text clarifies the ideas into an explicit position that is used by symbolic interaction. This view of interaction works well with Erving Goffman’s approach to performance. It is my goal in this to build a bridge between symbolic interaction and computational representation of social characters.

The Methodological Position of Symbolic Interactionism

There is a very clear review of the principles of symbolic interaction. These three premises form the foundation and basis for all of this work. The principles are as follows: (p. 2)

1. Human beings act toward things on the basis of the meanings that the things have for them.
2. The meaning of such things is derived from social interaction.
3. These meanings are handled in, and modified through, and interpretive process used by the person dealing with the things encountered.

The first premise is straightforward and reasonable enough, but is generally not accounted for in other theories of social science and psychology. The meaning of the things involved is of central importance beyond the things themselves. To ignore that meaning (and the variability of the meaning to different individuals) undermines the study of behavior of subjects who might be interacting with the things. The meaning itself does not come from the object itself, but arises from interaction between people. The social origination of meaning for objects reverberates with Tomassello’s understanding of joint attention. It also opposes the idea of natural affordances that come from Gibson and Norman.

Symbolic interaction views action as the center of human society. Activity oriented approaches to modeling human behavior is totally consistent with the views here: “Culture as a conception, whether defined as custom, tradition, norm, value, rules, or such like, is clearly derived from what people do. Similarly, social structure in any of its aspects, as represented by such terms as social position, status, role, authority, and prestige, refers to relationships derived from how people act toward each other.” (p. 6-7)

Social interaction forms conduct. One must fit one’s activity into the space of others’ actions. There are two levels of interaction from derived from Mead: gestures and symbols. The difference is that gestures are non interpreted and symbols are interpreted. There is also a triadic nature of meaning that also derives from Mead. A gesture has three parts: “It signifies what the person to whom it is directed is to do; it signifies what the person making the gesture plans to do; and it signifies the joint action that is to rise by the articulation of the acts of both.” (p. 9) This triad and the unit of the gesture are an interesting and potentially useful target for simulation of social characters. Gestures are not interpreted, but they are dynamic, direct, and fluid.

Objects are posed generally, but the concept is well defined. Blumer explains Mead’s conception of objects: An object is anything that can be indicated or referred to. There are physical objects, such as trees, chairs, and other physical entities. There are social objects, such as a student, a mother, a friend. Finally, there are abstract objects, such as moral principles and ideas of justice or compassion. Objects involve commonality, perspective, and meaning. Objects take on different meanings according to the perspectives of the individual considering the object, where that perspective is determined by identity, role, and so on. Another interesting note on this is that this sense of objects strongly relates to the method of object interaction found in The Sims. Common objects are defined culturally, they have the same meaning to a class of people.

The possession of a self is the ability to treat oneself as an object. This works by taking the positions of others. The positions one can simulate correspond to stages of development: individuals (play stage), groups (game stage), and the community (generalized other). Self objects can be defined by roles. Role taking involves perceiving oneself as one might be seen by others. Given an object and activity oriented model for social behavior, this formulation gives a strong endorsement for the use of roles (and eventually performance) as integral to social interaction.

A bold chain of reasoning claims that human action comes from self-indications, rather than motives, needs, conditions, stimuli, etcetera. This is more in line with an identity oriented understanding of individuals. Observation gives way to interpretation, which is filtered through the roles and the frame of the self, and this interpretation creates self-indications. Action is made on the basis of these indications, not the stimulus itself. This approach rejects both the raw behaviorist position, and also the position of rational planning. Planning and thought may be considered, but they are formulated in the sense of self-interaction. Activity is a sequence of actions and situations, a formulation which is remarkably prescient and reverberates with Agre and Lave.

Blumer also challenges the dominant view which relies on the inherent stability of social structures. Participants build actions through designation and interpretation. Institutions are diverse sets of members who act according to some set of meanings. The institution itself is therefore an emergent phenomenon, and not inherently stable.

Sociological Implications of the Thought of George Herbert Mead

The self is a process, not a structure of internalized norms and values. On the act: the act is constructed by the self, based on self-interaction and indications. It is not constructed by responding directly to observations, which is the dominant view (Watsonian behaviorism), and is flawed because neglects the self.

Existing views of interaction pose it as about conflict, common sentiments, and so on. Mead’s symbolic interaction views that interaction is the interpretation and defining of one another’s acts. This position can accommodate a wide range of human relationships. Joint action interprets actions as having a common  shared meaning. A Joint act is a social act. It comes together through participants interpreting, defining, and fitting their actions. This idea is extremely relevant to the view of roles and performance. Social actions must be fit into the space of other actions and the current situation.

Society as Symbolic Interaction

Symbolic interaction poses that interpretation occurs between stimulus and response. Interpretation forms symbols. This still seems a lot simpler and realistic than the mode of communication posed by BDI. Arguably, the BDI approach considers the selves of its participants, and it involves interpretation, a great deal of it, but that interpretation is divorced from the dynamic traits of participation and interaction. “Fundamentally, group action takes the form of a fitting together of individual lines of action. Each individual aligns his action to the action of others by ascertaining what they are doing or what they intent to do–that is, by getting the meaning of their acts. For Mead, this is done by the individual “taking the role” of others–either the role of a specific person or the role of a group (Mead’s “generalized other”).” (p. 82) Blumer also explicitly criticizes views of society as “social systems” which are composed by the collected actions of individuals trying to meet their life situations, which seems to aim to dissuade against the work of Axtell.

It is extremely important that society or culture be understood as made of individuals with selves, that is, they engage in active interpretation. Failure can lead to either viewing interaction as composed of raw stimulus and response, or can occur where the observer injects his own meaning.

George Herbert Mead is one of the seminal influences in symbolic interaction, which is a semiotically oriented subfield within sociology. Symbolic interaction alone claims that action is based on meaning, and meaning is socially derived. I think that the pair of meaning and action is strongly reminiscent of the semiotic dyad. I would like to consider symbolic interaction from the perspective of symbolic cognitive science and AI. The flaws of symbolic AI are tempered by the social origin and construction of meaning, and the manner in which action and meaning are interweaved.

Mead describes his work as behaviorist, but he interjects an important layer between stimulus and response, which is the matter of interpretation. I think this layer is extremely important, but I suspect that the pattern of stimulus, interpretation, and response is still not sufficient to serve as a model for simulation of social agents. When considered in context of roles and performance, as described by Erving Goffman, the perspective seems much more complete.

This book in particular is a compilation of Mead’s theory of social psychology. The ideas were developed in 1900 and the book was published in 1932. The preface argues that the work is incomplete, but is the fullest volume of Mead’s on social psychology. Mead is something of a bridge between science and philosophy. The emphasis is strongly on the dependence on society, which is at odds with the more individualistic trends in early cognitive science and psychology. According to Mead, the mind is an emergent phenomenon. Reminiscent of Vygotsky, Mead argues that language is the means for this emergence. The transformation from biology to mind is given in terms of language, but the examples given are about symbols and contextualized meanings. Interestingly, meanings are not subjective, private, or mental, but they are instead shared and objective within the situation. This sort of reference is akin to Agre’s diactic entities.

Contrary to trends within cognitive science, the manner of reasoning goes from the outside in, with society before the individual. “Instead of beginning with individual minds and working out to society, Mead starts with an objective social process and works inward through the importation of the social process of communication into the individual by the medium of the vocal gesture.” (p. xxii) This is a dramatically different perspective than engineered by Newell. Mead would probably argue that Newell’s model of cognition fails because of its failure to sufficiently consider the elements of interpretation and selfhood within everyday life.

The Point of View of Social Behaviorism

Social psychology is interested in the effect of the social group on individuals. The use of behaviorism seems to be a matter of method: it is argued that the observation of conduct is sufficient to understand psychology. This is in opposition to the study of introspection (encouraged by Wundt), which attempts to imagine and infer the internal cognitive processes at work in the human mind. Introspection and behaviorism are two ways of approaching a black box problem. The behaviorist position does not argue that it can explain what is inside the box, but it can explain how the box interacts with the outside world. Introspection aims to piece together the insides from both observation and self-reporting. Newell and Simon make use of introspection through their think-aloud protocols in their research for GPS. As an aside, the difference between introspection and behaviorism is interesting within the context of fiction. Introspection gives insight into character’s inner lives, which may be unreliable, where without that view, the reader is left to guess.

Mead favors the behaviorist position, making the claim that the behaviorist method is sufficient to study social psychology. This argument does not devalue the internal mind, but it argues that it does not need to be considered in this case. Mead’s position is understandable because he is constructing a perspective of social psychology, which is dependent on interaction. Interaction relies on individuals considering each others’ conduct, and not each others’ internal minds. Mead does argue that we can only understand conduct in terms of language. I might adjust this claim somewhat, and argue that we might consider gestures, gazes, and other expressions part of that language.

Despite his behaviorist claims, Mead makes a significant break from the traditional separation of stimulus and response. He also foremost prioritizes the society before the individual. In both senses of ‘prioritize,’ society both comes first and is of foremost importance. “We attempt, that is, to explain the conduct of the individual in terms of the organized conduct of the social group, rather than to account for the organized conduct of the social group in terms of the conduct of the separate individuals belonging to it. For social psychology, the whole (society) is prior to the part (the individual), not the part to the whole; and the part is explained in terms of the whole, not the whole in terms of the part or parts. The social act is not explained by building up out of stimulus plus response; it must be taken as a dynamic whole–as something going on–no part of which can be considered or understood by itself–a complex organic process implied by each individual stimulus and response involved in it.” (p. 7) Also important in this passage is the sense of understanding individuals based on terms that are used to understand the society. This is the origin of shared social meanings, which are used to define individual thoughts and interactions. Mead’s approach at its onset rejects the idea of psychic unity because of his claim that individual minds are are formed through social interaction and must be understood in those terms.

Symbols and intentions are introduced as an awkward way of formulating communication. The model proposed is that symbols are presented in communication, and the goal of communication is for the intention behind the symbol in the speaker to be reproduced in the listener. This does not seem like a correct model of communication, as it misses elements where the intention is meant to be opposite and not shared, or misleading communication. One example of this is where one individual makes threatening gestures, and its intention is aggression, whereas the recipient’s intention (if the communication is successful) would be fear or submissiveness. This sort of example is discussed somewhat, but does not seem to be clear.

Mead is criticizing the Watsonain approach to behaviorism, which would involve raw stimuli and responses. Similarly, emotion is studied in terms of physical manifestations. Darwin treats emotion as a mental state (as opposed to a valenced reaction), and mental states necessarily depend on consciousness. Mead wishes to focus on the social, which precludes consciousness. The raw stimulus and response approach to psychology yields a study of psychoses that map to neuroses. Mead does not wish to emphasize or promote the difference between mind and body. Instead, he wishes to find a correlation between the experience of the individual and the situation, not the individual and the stimulus.

Mind

This section is on the symbolic nature of communication, which may be gestural (speech is seen as a kind of gesture). Communication, as described, works when one makes a gesture, and the intent of that gesture is shared and taken up by the recipient. This argument may be readily challenged. Interaction is applied to social symbols, and the recognition of a symbol generates a response.

Language does not reflect things that exist, but also makes it possible for new situations and objects to exist. This works in the sense that an object is dependent on circumstance and use. Social process enables new objects, and communication brings out new relationships between gesture and act. When the self and matter of interpretation is introduced, gestures become symbols. The general nature of objects is very useful and important. Mead has a somewhat idiosyncratic understanding of what an “object” is, but it is extremely relevant for representations of social interaction. An object is something that may be observed or referred to. In this sense, objects are different depending on circumstances, and new objects may be formed via communication and interaction.

The Self

Mead draws a distinction between the body and the self. I think this may still be challenged, but it is interesting that the distinction is not between body and mind. The “self” is reflexive: it may be either a subject or an object. Communication is enabled and operational by making use of the self as an object. With Mead’s nuanced understanding of an object, the idea that the self can be made into an object is actually quite remarkable. The self may take on different meanings according to the circumstances, and this is very important for the understanding of social behavior and action.

The self is primarily not physiological, but the physiological organism is essential to the self. The self as a social object is socially (and therefore culturally) determined. Mead does discuss the performative element of action, but only briefly. Thinking is required prior to social action, but thought is merely inner conversation. The complete self is a reflection of entire social processes: “In other words, the various elementary selves which constitute, or are organized into a complete self are the various aspects of the structure of that complete self answering to the various aspects of the structure of the social process as a whole; the structure of the complete self is thus a reflection of the complete social process.” (p. 144; emphasis mine) This bold claim ties back to both a challenge of psychic unity and a use of Vygotsky’s internalization. The individual is thus a reflection of the social whole, complete with its cultural meanings and values.

Mead discusses children and games and play. Development is posed as a process (at the kindergarten level), where children play and enact roles, which involve role taking of the other. This also calls for the individual to simulate, through play, the “generalized other,” which is still situational. The elements of simulation and role taking reverberate with trends in modern developmental cognitive science (think Tomassello), and also bridges this with elements of play and performance. Play works in the sense of “playing at” another, but also in the sense of free movement within a permeable imagined social landscape.

Simulation of the generalized other is necessary for thought, that is, it is necessary to enable internal conversations. Through role taking, it enables continuation and enforcement of social practices, standards, and values. This goes hand in hand with identity (in the sense of Holland) and indentification, accounting for the diversity of social groups.

There are two stages of the self: (1) The individual self, which has personal attitudes. (2) The social self, which engages in role taking and simulation. Mead’s example in the latter is a game. I would argue that games have significant potential for role taking on their own. The game is a structure for an organic system of meaning. It is also worth it to compare these two stages with the capabilities of characters in digital games. Characters in digital games tend to be bound to some (somewhat social) system of meaning by its very rules, but most game characters only have fragmentary data that reflects their individual standing within the world. It is by incorporation of social attitudes, the achievement of the second stage of the self, that social patterns are internalized into the individual.

The metaphor of gameplay as a means for social integration is valuable, especially from the perspective of games in development, and the establishment of models. It is important that “He has to play the game.” Mead’s example is of children playing a ball game as a metaphor for introduction into society. It is necessary for a child to play the game in order to understand it, similarly, it is necessary for individuals to participate in society in order to understand the meaning of social interactions. Games and social value systems are understood through play, participation, performance, and practice.

A final important note: “No individual has a mind which operates simply in itself, in isolation from the social life process in which it has arisen or out of which it has emerged and in which the pattern of organized social behavior has consequently been basically impressed upon it.” (p. 222) Mead’s position is very much in contrast with the theories of cognitive science that emerged later in the 1960s. Despite their independence, Mead resonates with Vygotsky as well. Mead’s theory of social dependence is remarkably progressive and ahead of his time.

Society

Society is dependent on communication, which, as was established earlier, is dependent on language. I would argue that human society as we know it is distinguishable because of symbols and meaning, although these are somewhat vague and abstract. Mead is interested in differentiating between human and insect society. The matter of symbol and communication are ambiguous in this context. Insect society is characterized by a certain direct mapping between stimulus and response. Human society is different because we possess selves, which allow us to interpret senses into symbols, rather than act directly on those senses themselves.

The development of society is dependent on the common experiences of its members. Communication is thus dependent on society: “You cannot build up a society out of elements that lie outside of the individual’s life-processes. You have to presuppose some sort of cooperation within which the individuals are themselves actively involved as the only possible basis for this participation in communication.” (p. 257) This exposes a complexity with Mead’s earlier claims about the individual’s dependence on social meaning. Here he claims that social meaning must originate from individual meaning. This is a cyclic argument, but accurate. The natural conclusion is that the individual cannot be separated from the social. In order to understand one, we must understand both. In order to understand meaning of symbols, it is necessary to consider the social meaning of the symbols, and the experience of those symbols by individuals.

Propp is one of the earliest formalist accounts of story structure. However, immediately in the preface, we see an interesting perspective that is not normally accounted for. Morphology as a practice comes from botany, which is a study of the component parts, and then the relation between individual parts and the whole. The botanical metaphor is interesting, and casts the flavor for Propp’s entire analysis. Morphology in botany is used for two things primarily: classification and study of function. The morphology approach in botany is generally usurped by genetics in modern practice. A genetic analogue to the folktale would observe the creation of tales and study their emergence, and the factors that caused them to appear in the way that they do. Morphology is a differentiation and classification among specimens, identifying how some tales may be alike in many ways but different in key elements.

From a computational perspective, Propp’s work is an excellent example of a descriptive or analytic model. This may be used for classifying or understanding a body of work in terms of its parts, but it is not sufficient to generate new material computationally. In fact, it is not generally possible to proceduralize the deconstruction of a tale into a grammar. Human knowledge and context is needed to both interpret a grammar from a story, and to turn a grammar into a new story. These dimensions have a great deal of choice and flexibility involved, and these details are not within the resolution of the descriptive model. Nonetheless, that does not mean that the model is useless or flawed. It may still work as a tool for interpreting and analyzing tales once they have been converted into the symbolic structure required by the model.

The History of the Problem

Propp’s account for the problem suggests that the goal of morphology is to investigate and catalogue, but also provide a history of folktales. The perspective of understanding history recalls the idea of the genealogy of folktales, understanding how the patterns emerged, although this does not seem to be discussed. The study of folktales have attempted to classify tales in various ways, differentiating between animal tales and fairy tales, and so on. Morphology is posed as an alternative to this sort of high level and hierarchical classification. Existing theories have made classifications based on category, theme, and motif. The flow of these ideas gradually transitions into a grammar that analyzes tales in terms of elements. The categorical approach gives a hierarchical, taxonomy, and that is the origin from which Propp wishes to relocate the study of folktales.

The Method and Material

Propp has targeted his material specifically and exactly. The subject of his study are the fairy tales classified by Afanas’ev, from the numbers 50 to 151 (although there was a more recent reordering, which makes the new numbers 93-270). Propp is forming a morphology of only these 100 tales, so his approach may be seen as maybe not a morphology of all tales, but an example of the means by which a morphology may be defined around a collection of works. The morphology is “a description of the tale according to its component parts and the relationship of these components to each other and to the whole.” (p. 19)

More exactly, the tale is broken down into the functions of the dramatis personae. The function is the unit of analysis from which the tales are composed. The set of characters is large, and the set of functions are small. The two dimensions (function and character) lead to a large combinatorial diversity, producing a large potential number of tales. It is important to clarify that the function is not the act performed by a character, but the meaning behind that action. “Function is understood as an act of a character, defined from the point of view of its significance for the course of the action.” (p. 21) Propp defines four theses relating the functional composition of the folktales: (p. 21-23)

1. Functions of characters serve as stable, constant elements in a tale, independent of how and by whom they are fulfilled. The constituted the fundamental components of a tale.
2. The number of functions known to the fairy tale is limited.
3. The sequence of functions is always identical.
4. All fairy tales are of one type in regard to their structure.

The Functions of Dramatis Personae

This section forms the bulk of Propp’s work. He catalogues with great detail the functions present in a tale, which are identified by Greek and Roman letters, as well as a few symbols. He classifies the functions in order of their appearance within tales, gives several varieties of each type of function, and examples of those variants. The botanical metaphor continues at this level: Propp explains that the task of identifying functions is the extraction of genera. Genera would give way to species, and then varieties. What is interesting is that Propp uses a genetic metaphor to examine the component structures of the folktales, not the tales on the whole.

A good example of a function which has an important role within the morphology is “VIII. The villain causes harm or injury to a member of a family. (Definition: villainy. Designation: A.)” (p. 30-34) Villainy takes on many forms. I will mention a few of them:

1. The villain abducts a person (A1). A dragon kidnaps the tsar’s daughter (131), a peasant’s daughter (133); a witch kidnaps a boy (108)…
2. The villain seizes or takes away a magical agent (A2). The “uncomely chap” seizes a magic coffer (189); a princess seizes a magic shirt (208); the finger-sized peasant makes off with a magic steed (138).
3. The villain pillages or spoils the crops (A3). A mare eats up a haystack (105). A bear steals the oats (143). A crane steals the peas (186).

There is a brief interlude where Propp explains that certain chains of functions have types of their own. Several types of functions work well with each other, or do not work well, and there is a great diversity of possible connections that occur within the catalog of folktales. Generally, illogical connections may exist, but require extra motivation or context in the tale. In computational adaptations of Propp’s work, these connections are extremely problematic, though. In the context of a pure morphology, that is not a problem. This is one point where it is important to realize that Propp’s study is a descriptive or analytic grammar, not a generative one. A human could perform the task of reconciling an illogical chain of functions, but that is generally beyond the power of a computational generative grammar.

For accommodating stories with multiple parts, villainous “moves” are extracted to form subsequences in the grammar. Some of these groups may be cycled and repeated. Additionally, some elements are grouped: prohibition is always paired with violation, for example. The sequence ABC↑ can be understood as a unit, the “complication.” The sequence DEF is the hero’s testing and reward, which also serves as a unit. There are combinatorial variations within these units, but they serve as logical groupings within the grammar itself.

Some Other Elements of the Tale

Tangentially, Propp mentions the issue of motivation. Motivation is extremely important in my own work, and it has an insubstantial role in the formalist analysis. Here motivation is addressed as following from the action (plot) itself. Some motivations are hatred, fear, jealousy, love, lack, and justice. Motivation adds to the quality of the action, but does not address the form of the action itself.

The Distribution of Functions Among the Dramatis Personae

Functions are grouped by spheres of influence, namely, spheres have have authority over certain functions. Characters may correspond to the spheres, or a single character may cover multiple spheres, or many figures may make up a single sphere. These spheres serve as roles within the format, which would conceivably be where agents would intervene if they were able to act within the story as a world. The spheres are enumerated as follows: (p. 79-80)

1. The villain. The villain performs the villainy, struggles with the hero, and pursues him/her.
2. The donor (provider). The donor gives the hero a magical agent.
3. The helper. The helper may undo the misfortune or lack, rescue the hero from pursuit, and transfigure the hero.
4. The princess (a sought-for person) and her father. This sphere assigns the difficult tasks, brands the hero, exposes and recognizes the hero, and also participates in marriage.
5. The dispatcher. Dispatches the hero.
6. The hero. The hero performs the departure and engages with the villain. Oddly, the conflicts are represented in the villain’s domain in this analysis.
7. The false hero. The false hero appears in some stories, engages in some of the activities of the hero, and has a special function of presenting false claims.

The Tale as a Whole

The methods of recombination are actually rather complex. The opening problem is how to distinguish one tale from another. The first step is to tell how a single tale is structured, identify “what is meant by a tale” (p. 92). Given here are methods of reconstruction, which are moves that may take place in the story. Each move is a conflict and resolution, which actually take on multiple formats by the grammar. Some tales have two full sequences of A-W, another example interrupts a tale so it has three sequences: A-G, a-K, K-W. This diversity of recombination is extraordinarily complex, and does not actually seem to be well integrated into the final grammar.

Some tales are differentiated by the sequences of H-I (the struggle with the villain and victory over him), and M-N (the difficult task and its resolution). Many tales diverge on these. The challenge is reconciling the study of the whole to include both systems of multiple moves, as well as this H-I and M-N distinction. Propp does explain that four categories result: tales with H-I, tales with M-N, tales with both, and those with neither. He observes that when there are multiple moves, then the fight (H-I) move always occurs first and is followed by the difficult task (M-N). The solution for reconciliation is to use a branching path. However, the grammar still does seem unusually inflexible.

The remaining discussion is on the variants of the morphology and the characteristics of the overal forms. Propp finds that many instances of functions are interchangeable, but some are not, i.e. H1 must always be followed by I1. Some functions may change places. Direct violations exist, but are usually exceptional cases. Theme is an ill defined problem in terms of story form.

A quick review of Jameson’s intentions: historicizing. The central theme and motivation behind The Political Unconscious is the desire to historicize narratives and to understand them within a Marxist framework of meaning. Texts come to us as already read, and interpretation weaves between previous interpretations. Interpretation is, essentially, allegorical. The goal of this study is to use the Marxist framework to understand the system behind interpretations.

Jameson’s chief influences are Frye, Greimas, Freud, Levi-Strauss, Lukacs, Barthes, and Deleuze. Jameson aims to tie these approaches together:

These divergent and unequal bodies of work are here interrogated and evaluated from the perspective of the specific critical and interpretive task of the present volume, namely to restructure the problematics of ideology, of the unconscious and of desire, of representation, of history, and of cultural production, around the all-informing process of narrative, which I take to be (here using the shorthand of philosophical idealism) the central function or instance of the human mind. This perspective may be reformulated in terms of the traditional dialectical code as the study of Darstellung: that untranslatable designation in which the current problems of representation productively intersect with the quite different ones of presentation, or of the essentially narrative and rhetorical movement of language and writing through time. (p. 13)

The concept of Darstellung seems like a potentially relevant idea that is worth investigating.

On Interpretation

Jameson gives a historical review of interpretation (through Althusser). These means of interpretation are ways of reading and connecting the model of the text productively to the real world. Specifically he explores medieval interpretation and how the biblical texts were analyzed according to four levels: (p. 31)

1. Analogical: political reading (collective “meaning” of history)
2. Moral: psychological reading (individual subject)
3. Allegorical: allegorical key or interpretive code
4. Literal: historical or textual referent

The semantic network behind most narratives is the political unconscious. Jameson is interested in texts, and the space common to all texts (or shared by texts) has a master narrative which is necessarily unconscious. Jameson is not interested in the acceptance of these master narratives, but rather challenging them and understanding them in historical context.

The idea is, in other words, that if interpretation in terms of expressive causality of of allegorical master narratives remains a constant temptation, this is because such master narratives have inscribed themselves in the texts as well as in our thinking about them; such allegorical narrative signifieds are a persistent dimension of literary and cultural texts precisely because they reflect a fundamental dimension of our collective thinking and our collective fantasies about history and reality. (p. 34)

Mediation is about the relationship between the text and its social and cultural context. Mediation is an inherent source of ambiguity. The modern variant of mediation is “transcoding” which illustrates a more literal semiotic process and relationship wherein meanings are encoded and decoded in different ways. An approach to transcoding could be to strategically use different decoding schemes to analyze texts at different levels.

Exploring mediation and transcoding: Expressive causality is a kind of mediation which seems to suggest a kind of audience creativity. The work of mediation is identification and differentiation. Mediation is explored by way of Marxist analysis.

Jameson gives an argument for the narrativization of the real. The real is understood as fantasy, especially within context of the Deleuzian libidinal or “desiring” apparatus. There is an interpretive power in ideology, exploring what the text represses. The notion of master texts is posed in the Freudian language of consciousness and repression. The argument that texts may be read and have adjusted meanings is reminiscent of postmodernism, but the rejection of inherent meaning given by postmodernists clashes with the claim of textual repression.

More than this, the very closure of the “semiotic rectangle” now affords a way into the text, not by positing mere logical possibilities and permutations, but rather through its diagnostic revelation of terms or nodal points implicit in the ideological system which have, however, remained unrealized in the surface of the text, which have failed to become manifest in the logic of the narrative, and which we can therefore read as what the text represses. (p. 48)

Jameson argues that the formalizing goals New Criticism serve to propagate a particular view of what history is. This is interesting in comparison with proceduralization of literature, which I would argue, propagates a view of what meaning an interpretation are.

To Frye, there are phases in reinterpretation, which is essentially a form of rewriting of texts. The fist phase is that  The second phase is a cultural object within

1. The text is an object created as a form of expression, and may be interpreted as a single and independent work.
2. The text is a cultural object situated within a cultural context. The smallest unit of study at this level is the “ideologeme.”
3. The text is read in terms of the ideology of form, which is a means for historicizing the text within multiple sign systems, which are themselves situated historicaly.

Persuasive Games is about procedural rhetoric, and exploring the use and potential of rhetoric in games. To Bogost, rhetoric is much more integral to games than expressiveness, or rather that expression is dependent on rhetoric. In my own analysis of games and systems, I use the term model, and in this context, models can be thought of as rhetorical systems. Rhetoric is about persuasion, which in turn uses (implicitly or explicitly) assumptions and relationships, which are the stuff from which models are made.

The context of Bogost’s work is the struggle of videogames for legitimacy. For reasons that echo Sutton-Smith, games have faced problems with acceptance because of two causes. One is the misperception that games are for children, which is overtly false. The second is a quandary that games are trivial and inconsequential. This misperception construes that games are trifles and do not have the weight of traditional media (citing James Newman). Henry Jenkins has argued that the struggle for legitimacy is characteristic of a new medium. Bogost argues that legitimacy must be obtained by a critical analysis of the rhetoric of videogames, which must explore how they work. How games work addresses the level of mechanics, but also the levels of expression and representation. This critical analysis is what Bogost calls procedural rhetoric. Serious and political games must explore procedure rather than content.

Procedural Rhetoric

Discussing an example of procedural rhetoric, Bogost explains the game “Tenure” (created by Owen Gaede in 1975). This is about high school education, the player is a teacher, and is faced with various options and choices throughout gameplay, which reflect choices that a real teacher might need to make. It reveals issues of interpersonal dynamics (with school faculty), and with the demands and expectations placed on new teachers. What is interesting to note, is that while the game makes many simplifications, and is not necessarily realistic, it makes claims about teaching and illustrates those claims. “Tenure makes claims about how high school education operates.” (p. 2)

Bogost examines procedurality in several senses. Procedure calls to mind computational procedures, but there are also human procedures, which are made prevalent by customs of practice and bureaucracies, and so on. A human procedure given is the way a retail clerk might handle a customer asking to return an item that is past the normal time that returns are accepted. Computation is symbolic and representative. Procedures are expressive with respect to representation. They reflect cultural values, because they are situated in a culture. Human procedures (such as the store return) are very flexible, whereas computer procedures are relatively inflexible. However, the human flexibility is not a flexibility within the procedure itself, but a way of accounting for and transferring to other procedures which might bear more strongly on the situation. When faced with procedures of any nature, a standard human reaction is to investigate its operation. We ask “how does this work?” and we explore procedures in order to explain them.

Rhetoric is inherently about persuasion and here is a review of rhetoric deriving from the Greeks. The term “rhetoric” is derived from the same root as “oratory,” and the term classically applied to the manner of speech used in political speeches. Historically, rhetoric has extended to literary and artistic practice. In these contexts, rhetoric means “effective expression” (p.19). Kenneth Burke is influential here, who argues that identification is the center of rhetoric, not persuasion. This echoes with the notion of shared assumptions and understandings. The term Burke uses to characterize identification is “consubstantiality”, which in its etymological root suggests that rhetoric encourages the speaker and audience to share the same stance or substance. A central principle of rhetoric is the enthymeme, which is a syllogism with the assumption omitted. It is omitted because it may be assumed to be shared. This assumption requires filling in the missing gap on the part of the listener. This can also be argued as the important shared substance on which both the listener and speaker form their identification.

A strong example of procedural rhetoric is Molleindustria‘s McDonald’s Game. This is a strong example because the game has a message (that corruption is a necessity in fast food, and that there are inherent problems with the practices used by fast food corporations), and it communicates these through illustration and simulation of how these processes work. This last step communicates because it presents a model of the system that it represents, and the messages emerge as logical conclusions when one plays the game. Other games with ostensible educational interests fail to effectively use this demonstrative illustration. Bogost gives two examples of games of this sort: “G!rlpower Retouch” and “Freaky Flakes.”  These are problematic because they illustrate the content of their message, they fail to connect the procedures in game to the procedures which they are attempting to represent. This critique in mind, procedural rhetoric seems to be about obtaining a synchronicity of process, representation, and message.

Regarding interaction and procedural rhetoric: Bogost gives several examples of interaction in games. One of these is in Chris Crawford’s game “Balance of the Planet,” which gives the user affordances for manipulating variables that affect the simulation through sliders. Another example is Grand Theft Auto 3, which enables the player to “do anything” within the “parameters of the game.” The claims of freedom are suspect because of the restraints of the due parameters. A player may assault anyone whom he likes, but cannot approach a random person and engage in conversation. However, this restriction is a form of expression within the model presented by the game. What is omitted from the model is just as important as what is present, and that omission may be seen as making a claim about the world represented in the game. It is important to remember those assumptions, though, as a subject of critique.

Chris Crawford has described interactivity as a tight loop of listening, thinking, and speaking. Games may be examined and judged based on how they listen to the player (what the player’s engagement may be), the complexity with which it interprets those inputs, and the expressivity with which they talk back to the player. Sophisticated interactivity is not necessarily more or more frequent interactions, but ones in which input, model, and representation are all tightly linked. Sophisticated interactivity yields an effective enthymeme. The simulation gap is posed in terms of rhetoric. It is the players’ synthesis, filling in the gap of the enthymeme. “A procedural model like a videogame could be seen as a system of nested enthymemes, individual procedural claims that the player literally completes thorough interaction.” (p. 43) Increased persuasion would be increased coupling of the model and representation. Crawford’s take on interactivity are all important for rhetorical argumentation.

There are two kinds of analysis: black box and white box analysis. A concern is with the lack of visibility in black box analysis. This is exemplified by Sherry Turkle’s critique of Sim City. The argument is that the game is a black box, and the model is not externally visible, so one cannot see how the game works. However, players can see how the game works through playing the game. However, what is missing is whether players will be able to make sense and observe objectively the claims and mechanics of the game. This ambiguous element is procedural literacy.

Procedural Literacy

Looking at games and learning, a predominant trend is to use principles of reinforcement from behaviorism. The behaviorist perspective is that games teach “basics” of represented material. For example: Flight Simulator, Sim City, Ninja Gaiden, GTA. Behaviorism forecloses the simulation gap, assumes that playing a role necessarily implies validation. An alternative view is constructivist, which comes from Montessori education and references Papert.

With constructionism, enthusiasm is a factor in play, but it is not in the sense of play as progress, but rather play as creativity (in regards to Sutton-Smith). James Paul Gee argues that games illustrate higher order thinking (essentially metacognition). This idea claims that games teach strategy, not individual actions. Bogost challenges this perspective as well, because it overemphasizes the general (that games teach mechanics and strategy), and that they forsake the importance of individual games and the expressive capability of their models.

The idea of procedural literacy comes from Papert (after Piaget), but also Perlin and Flanagan, and Mateas. Bogost wishes to discourage the idea of procedural literacy as merely programming. Programming is a relevant skill, but it alone does not imply literacy, nor does literacy imply being able to program. Expanding on Mateas: procedural literacy should not be limited to understanding how systems work in the abstract, but it should also be used to interrogate, critique, and use specific processes.

Systems that are strongly representative (like Playmobil as opposed to Leggo) encourage a stronger cultural reference and link into representative systems of meaning. Combining representations, or tying representations to systems allows for interpretive creativity. When performing this sort of creativity, the interpreter constructs models of the new combined system. “When a child constructs a Playmobil scenario combining HAZMAT-crew parts and pirate parts, he constructs an argument for how such a character would behave. This argument is carried out though the rules of play itself, the types of behaviors the child chooses to encourage or prohibit.” (p. 258)

To say procedural system, this implies something malleable and participatory, rather than just an artifact. Literacy involves reading and writing, and these tie into the sense of interaction, freedom, and play with the system itself. To be literate requires more than the ability to read alone, but the ability to write and create within the medium. Creation of models is natural, but procedural literacy demands a certain awareness in how one creates models, and how they are created generally. Procedural rhetoric asks the following questions: (p. 258)

• What are the rules of the system?
• What is the significance of these rules (over other rules)?
• What claims about the world do these rules make?
• How do I respond to those claims?

Bogost concludes the section with an excellent summary:

At the start of this chapter, I asked: if videogames are educational, what do they teach, and how do they teach it? To summarize the reply given here: videogame players develop procedural literacy through interacting with the abstract models of specific real or imagined processes presented in the games they play. Videogames teach biased perspectives about how things work. And the way they teach such perspectives is through procedural rhetorics, which players “read” thorough direct engagement and criticism. (p. 260)

Play is an inherently and deliberately ambiguous concept. Brian Sutton-Smith attempts in this book to conduct a review of studies of play. He is interested in defining play, but such a definition is unlikely to capture the full scope and meaning of the term. Instead, he explores play from the perspective of several rhetorics. These rhetorics are each culturally derived, and each has a certain intrinsic ambiguity. Understanding play is, for obvious reasons, integral to understanding games. However, literature and other works are also playful. I would argue that a certain degree of play is required in building models (creatively), so the rhetorics of play will go a long way in bridging games and literature.

Play and Ambiguity

This section primarily introduces the ambiguities of play and attempts to understand what play is. It explores the question of what the levels of play are, the kinds of play there are, and so on. A substantial list of examples is given here, which is very useful in interpreting the scope and flexibility of both play as a concept, as well as Sutton-Smith’s analysis. These topics are so diverse that it is challenging to imagine what sort of theory would tie them all together. In the following list, I have enumerated Sutton-Smith’s categories, and given a smattering of the examples he gives. (p. 4-5)

• Mind or subjective play: dreams, daydreams, fantasy, imagination, Dungeons and Dragons, playing with metaphors.
• Solitary play: hobbies, collections, listening to music, art projects, pets, reading, yoga, collecting and building cars, Civil War reenactments, bird watching, crosswords.
• Playful behaviors: playing tricks, playing around, playing up to someone, playing a part, putting something into play, playing fair, playing by the rules.
• Informal social play: joking, parties, travel, leisure, dancing, getting laid, potlucks, malls, babysitting, creative anachronism, intimacy, bars and taverns, amusement parks.
• Vicarious audience play: television, films, cartoons, spectator sports, theater, jazz, rock music, parades, comic books, Renaissance festivals, museums.
• Performance play: playing the piano, playing music, being a play actor, playing the fishes, playing the horses, play voices, playhouses.
• Celebrations and festivals: birthdays, Christmas, Easter, Mother’s Day, Halloween, gifting, banquets, balls, weddings, carnivals, balls, Mardi Gras.
• Contests (games and sports): athletics, gambling, casinos, lotteries, pool, golf, parlor games, drinking, the Olympics, cockfights, poker, chance, board games, card games.
• Risky or deep play: caving, hang gliding, kayaking, bungee jumping, skateboarding, windsurfing.

There are seven rhetorics of play. The rhetorics of fate, power, identity, and frivolity are described as the ancient rhetorics, having a much stronger standing in classical literature. The modern rhetorics are progress, self, imaginary, and the self. These all emerged with changing philosophical and psychological trends dating within the past 200 years. The rhetorics are described: (p. 10-11)

1. The rhetoric of play as progress. Progress has dominated studies of the play of children. This rhetoric poses play as a developmental arena wherein players learn and practice for adulthood. This was developed partially because of the characteristic imitation of adults by children. This is something that Sutton-Smith is interested in challenging.
2. The rhetoric of play as fate. This rhetoric is older than the rest, going back to mythologies in which human lives are controlled by destiny, gods, or luck.
3. The rhetoric of play as power. Power is at the heart of competitions, and this poses that play is the expression of conflict. It emphasizes that those who control the play are its heroes. This rhetoric is strongly opposed to modern theories around leisure and progress.
4. The rhetoric of play as identity. Sutton-Smith’s use of identity is tricky in this space. What is meant here is cultural identity. This emphasizes social and cultural roles and structures. There is a focus on communal identity rather than individual. The individually focused complement to this is the rhetoric of the self.
5. The rhetoric of play as the imaginary. The imaginary ties into creativity and flexibility. “This rhetoric is sustained by modern positive attitudes toward creativity and innovation.”
6. The rhetoric of the self. The rhetoric of the self is usually applied toward solitary activities, but can be characterized by other ideas such as fun, relaxation, and escape. The central focus is in the experience of the player. This rhetoric is arguably the most modern because of its appealingness to individuality and consumerism.
7. The rhetoric of play as frivolous. Frivolity is difficult to characterize: it applies to absurdity and the historical roles of tricksters and fools.

Rhetorics of Animal Progress

It is important to remember that scholarly study of play has existed around animal play in addition to human play. A comprehensive theory of play would need to account for the play of animals. Here, Sutton-Smith compares animal play with the rhetoric of progress. Progress distinguishes between child play and adult play. Child play is open and creative, while adult play is closed and recreative. Progress argues for development and learning (progress to adult life. This is compared to animal play, specifically as studied by Robert Fagen.

Fagen describes five categories of animal play (p. 21-24):

1. Isolated, brief jerky movements performed repeatedly without defense or counterattack by others: typical of rodents.
2. Noncontact solo play and the social play of moving bodies through space, running and jumping ina  variety of patterns; characteristic of hoofed mammals, some rodents, and some birds.
3. Social play, some with no contact, like chasing, and some with contact, like sparring and wrestling; characteristic of most primates and carnivores, many ungulates, pinnipeds, and marsupials, and some birds.
4. Complex social play, which involves games with objects and features of the landscape. This form of play is enacted by adults as well as young animals, whereas most play of the former kinds is enacted only by juveniles or by parents with their young; typical of social carnivores, primates, elephants, some whales, dolphins, and porpoises.
5. Mother-infant games, such as peekaboo, as well as object construction and play with pebbles, sticks, flowers, feathers, and bones, and play with snow, water, and trees.

Approaches to animal play as relate to other rhetorics:

• Skill training (progress): this is challenged because things like play fighting is an inverse of rehearsal for the real thing.
• Play fighting (power): under the rhetoric of power, play fighting would establish heirarchies, but juvenile play does not lead to future roles. (I would contest, though, because actual contests have a ceremonial element which resembles play fighting more. Play fighting can be a sort of practice or indoctrination to these ritual forms.)
• Bonding (identity): This is common to power- that play would cement social relationships, but young frequently separate and do not maintain social relationships established from play partners.
• Flexibility (imaginary): Animal play is repetitive. There is not evidence to suggest that the purpose of play is flexibility.
• Emotional experience (self): This seems the most compelling so far, the animals play for their own happiness. The net conclusion of all of this is that no one is making progress to demonstrate that adaptation is play’s main function. (p. 34)

Rhetorics of Power

Consideration of power is highly relevant to modern gaming culture. Power ties into cultural establishment and social building. The discussion of play as empowerment emphasizes compensation and wish fulfillment, and the dynamics of these contests are reminiscent of Geertz on cockfighting. The cockfighting is a much more complex example that delves into other elements of social discourse. Huizinga strongly encouraged the rhetoric of power, claiming that there is a parallel between play contests and real events: “His definition of play primarily as contest reflects the widespread male rhetoric that favors the exaltation of combative power instead of speaking comprehensively about play itself. Combat may be widespread but it is hardly a universal truth about all play forms.” (p. 80)

The rhetoric of power is inherently about rationality. It is ordered. Power misses the forms of disordered and irrational play, of which there are many, even around contests: chance, symbolic inversion, playfighting, play therapy, and so on. Power demands a certain order and is deeply upset by the chaos of chance or frivolity. One challenge to play as power comes from Spariosu, who observes that the western perception oscillates between order and chaos. The rhetoric of progress and rationality gives way to game theory in the mathematical sense, which emphasizes utility.

Rhetorics of Identity

Identity here is cultural identity (not individual). This also fits in with institutions. The study of collective play observes who is controlling the situation and what they are getting out of it. The rhetoric of identity lends itself to a play of cultural power. A great deal of identity play involves developing cultural identity in opposition to other cultures. However, identity formation does not need to be competitive, it can be cooperative. Sutton-Smith cites Bernie De Koven on well played games. “When the game is played only for the good of the larger community that plays it, then it can be well played. If not so subordinated, games may run away with their players and cause friction and conflict.” (p. 100)

Noncontestive identity play can involve situations within a cultural context. This can involve a breaking down of established boundaries, which leads to a reincorporation within a larger community. Examples of these are celebrations such as Mardi Gras, Christmas, Thanksgiving, and so on. All of these involve a breaking down of conventional social barriers, such as those of family and friends, the celebration incorporates a wider community and strengthens these secondary or weaker bonds. Festivals have a seemingly contradictory relationship, though. A festival might create a strong feeling of identity in the participants, but festivals can also encourage ambiguity within the participants.

Rhetorics of the Imaginary

Imagination is an elaborate category. Transformation is at the core of this rhetoric, and it is thus somewhat semiotic and representative in nature. Other terms for this rhetoric might be creativity or flexibility. It represents an interaction of the factual and the imaginary. A key scholar in this literature is Bakhtin. A useful citation point referencing Geertz: “This rhetoric seems not so much concerned with play as an intellectual contest, a competitive bout, or a parade; rather , play and games are presented as ways of thinking about culture or as texts to be interpreted.” (p. 128) This comparison is very resonant with Daniel Mackay’s work on role-playing games, which are aesthetic products in of themselves.

Imagination connects play to art. Art ties into performance and enactment. Sutton-Smith references the use of play in literature. Writing is a kind of play, but, so is reading. In writing fiction, play relates to play of simulation, in the sense of playing the fictional world in the author’s imagination.

Referencing Bateson (1972): play is not just play, but it is a message about itself, and it is of the world and not of the world. This perception of play is relevant for digital games and worlds, where the inside and outsides are complex. There are several senses of play in games: playing in a world, playing with things in the world, and so on. Sutton-Smith connects this sense of complexity and layers to Boccaccio’s Decameron. This story has many layers, and play exists within and between the layers.

Play is important in literature and everyday life, especially the media. Sutton-Smith explains: “Derrida is only one of a number of postmodern or poststructuralist writes who have expressed discontent with earlier ways of looking at the fundamental character of knowledge, and have suggested that life is much more generally play- or game-like than has generally been acknowledge in the earlier, more deterministic scenarios of a functional kind, such as are still found in abundance in the social sciences.” (p. 145) Additionally, William Stephenson has argued in The Play Theory of Mass Communication, that “all media constitute play forms, and that when we are watching or receiving them we are essentially at play (in the broader sense).” (p. 145) This provides a good theoretical point from which to argue for the commonality between literature and games.

The ambiguity in the rhetoric of the imaginary is of the frames involved. The ambiguity of frames resembles Goffman’s take on frame analysis. Framing and navigation between frames is thus a form of play. Situational identification, however, is thus inherently ambiguous. This observation sheds some light on why it is so difficult to form a comprehensive social model of situated activities.

Rhetorics of Self

The rhetoric of the self has to do with freedom primarily. It also focuses on individualism rather than community. Self has to do with fun, experience, psychology, and intrinsic motivation. Self connects to identity in the sense of personal identity. Sutton-Smith mentions Turkle (1995), who wrote of people’s identity and gender play in chatrooms. Even though this is social, the play has primarily to do with individual play and self discovery. This sort of activity resembles masked play as found in German festivals, and countless other situations. “There can be no doubt that virtual worlds are a new play form allowing adults to play almost as amorphously as children.” (p. 178)

The rhetoric of self borrows from the phenomenological tradition with its heavy emphasis on personal experience. Gadamer specifically investigated play within the context of phenomenology. This has much to do with subjectivity, and the relationship between perception and understanding the self. Sutton-Smith references Csikszentmihalyi on the psychology of flow and peak experience. Sutton-Smith also references Rubin, Fein, and Vandenberg on characteristics of children’s play. These are emblematic of the rhetoric of self. (p. 188-189)

1. Play is intrinsically motivated. (That is it is fun). This can be challenged though, as historically play has often been extrinsically motivated by social and cultural customs.
2. Play is characterized by attention to means rather than ends.
3. Play is guided by organism-dominated questions, rather than context-dominated questions.
4. Play behaviors are not instrumental. This argues that play is nonproductive and does not have serious consequences.
5. Freedom from externally imposed rules is necessary. This position is problematic because of the use of rules which may be imposed in organized games or improvised.
6. Players are actively engaged in their activity. This is at odds with daydreams and vicarious play.

Play of self is seen as a kind of performance. Play by itself, for itself. It is about enactment. Also involves language and framing, which is reminiscent of Goffman. This involves meaning construction and deconstruction.

Rhetorics of Frivolity

Frivolity as discussed is reminiscent of game players who subvert mechanics and break rules in effort to create nonsense (or irritate other players). It is antagonistically motivated, but also aimed at deconstructing and mocking cultural order, as is common with festival tomfoolery. The practice however lacks the cultural support given to the frivolous as discussed in this section. Even though the frivolous is antagonistic to cultural orders, it operates in specially accomodated contexts. Attention to the important role of the actively frivolous would be important to account for in design of game worlds.

Philip Agre is a rare breed. He is a strong advocate of embodiment and situated action, and is also an AI practitioner. Agre was enormously influential on Michael Mateas, among others. Agre is interested in developing an approach to AI that is critically aware, as well as reformulating conventional AI applications in a manner that might be described as situated or embodied. His view on diectic entities has an enormous potential in application to an activity-centric approach to simulated characters.

Agre gives a good overview of the task at hand. His interest is in a change of approach around AI, based in both philosophical to technical arguments. He advocates the central idea of a critical technical practice. This idea has been very influential on Mateas and Sengers, in particular.

Introduction

Agre’s goal is to shift from cognition to activity. AI has long been mentalistic, proclaiming to be a model of cognition and pure abstract thought. However, the practice of AI has tended to best work in application to specific practices and activities (note Hutchins). Theorem proving is not cognition in general, but it is a specific activity which can be computationally formulated and modeled. Focus on activity tends to model “situated, embodied agents.” (p. 4) The term “agent” applies to robots, insects, cats, and people, and has strength in ambiguity. Situation implies that an agent’s actions make sense within context of its particular situation. For an agent to be embodied, it must simply have a body: “It is in the world, among the world’s materials, and with other agents.”

There is a review and comparison of the planning based approach common to AI. Ordinary life has a sort of routine, and planning tends to make use of that routine. Agre poses that routines come from emergence. I would argue, though, that in social situations, aspects of routine may have been emergent, but they often become institutionalized. There are three questions that operate around activity, and each of these is important to consider for activity centered applications: (p. 7)

1. Why does activity appear to be organized?
   Planning view: activity is organized because of plans.
   Alternative: orderliness is emergent. Form of activity is influenced by representations, but not determined by them.
2. How do people engage in activity?
   Planning view: Activity is planned and contingency is marginal.
   Alternative: Activity is improvised, contingency is central. People continually redecide what to do.
3. How does the world influence activity?
   Planning view: The world is fundamentally hostile. Rational action requires attempts to anticipate difficulties and everyday life requires constant problem solving.
   Alternative: The world is fundamentally benign. Environment and culture provides support for cognition. Life is a fabric of familiar activity.

AI work is itself a practice, which has its own values: Getting computational systems to work. Systems that do not work lack value. This lends to the idea that only things that can be built can be believed. On the other hand, building is also a way of knowing. Agre argues that emphasis should be shifted from technical product, to practice. The use of models and ideas is torn between science and engineering, both practices which led to the development of AI. Science aims to explain with models, whereas engineering seeks to build. AI straddles these two drives. Similarly, works of art exist on that same border, they are both explanations (expressions), and constructed products.

Metaphor in practice

There are two parts to Agre’s thesis: (1) AI is organized around the metaphor of internality and externality. The mind has an inside and an outside and discrete inputs and outputs at the interface. (2) A better starting point uses a metaphor of interaction, and focus on activity. The important part behind this is to be critically aware of the metaphors used in discourse about the practice.

There is an important nugget here: The technical criticism is with the lack of self reflection. A model is an interpretation and a language, it is a way of seeing things that is inherently metaphorical in nature. There is a double vocabulary, one at the level of discussing the subject of the model, as well as discussing the model itself and its interaction with the software. This duality of discourse is reminiscent of Mateas. “Scientific inquiries based on technical modeling should be guided by a proper understanding of the nature of models. A model is, before anything else, and interpretation of the phenomena it represents. Between the model and the putative reality is a research community engaged in a discursive operation, namely, glossing some concrete circumstances in a vocabulary that can be assimilated to certain bits of mathematics. The discourse within which this process takes place are not transparent pictures of reality; nor are they simply approximations of reality. On the contrary, such discourses have elaborate structures and are thoroughly metaphorical in nature. These discourses are not simply ways of speaking; they also help organize mediated ways of seeing.” (p. 46)

Machinery and dynamics

The two metaphorical theories operational in AI research are mentalism and interactionism. Newell and Simon’s GPS seems interaction oriented at onset, but shifts focus onto abstract representation very quickly. GPS aims for disembodied conceptions of the world. It equates objects with their representations, which is a dangerous phenomenological pitfall.

Agre proposes discussing computation in terms of machinery and dynamics. His goal is to discourage focus on machinery and instead emphasize dynamics. However, as presented, machinery is a useful metaphor. It implies situation within an environment: a machine has a physical presence, which steps away from raw functionalism. The term machinery is also very Deleuzian, which may be either positive or negative. Dynamics instead centers on interaction. Agre encourages us to get rid of computational machinery, and instead invent new dynamic effects, rather than devices.

Abstraction and implementation

Abstraction and implementation is an interesting dyad: Functional definition versus physical construction. What is the relationship between these to representation? It is important to discern the levels of abstraction at work when constructing artifacts. Newell, Simon, and Shaw constructed GPS which is not a model of cognition, but a model of problem solving in a particular domain. There is a generative element in theory, common to the work of both Newell and Simon as well as Chomsky. Soar bills itself as a “unified theory of cognition,” but is still primarily concerned with abstraction. Minsky is a researcher who seems to take this into account: he claimed to represent a constellation of mini-theories which are in turn heavily situated. This approach emphasizes implementation rather than abstraction.

Dependency maintenance

There is a review of critical technical practice. Agre’s aim is not to break the traditions of AI or start over, but to become critically aware of representation and computational machinery. The subject matter in Agre’s examples fold back to everyday activities. The goal is to see if the planning paradigm tells “reasonable stories” about everyday activity. Effectively, this means to see if the planning view of cognition reasonably accounts for everyday interactions. This involves a comparison of activity and routine.

1. Different individuals use different routines for the same task.
2. Routines emerge from repetition and practice.
3. Novel activity is not routine.
4. Routines are upset by circumstances.

Planning and improvisation

Hayes-Roth and Hayes-Roth: real planning is opportunistic and incremental. A relevant question about real world planning is this: “How is it that human activity can take account of the boundless variety of large and small contingencies that affect our everyday undertakings while still exhibiting an overall orderliness and coherence and remaining generally routine? In other words, how can flexible adaptation to specific situations be reconciled with the routine organization of activity?”

Running arguments

The situation and activity oriented framework described that is an alternative to planning is “Running Arguments” or RA. The cycle of RA is determined by the following steps: (p. 175)

1. The world simulation updates itself.
2. The periphery computes new values for the central system’s inputs, which represent perceptual information.
3. The periphery compares the new input values with the old ones. Any newly IN input is declared a premise, any newly OUT input is declared no longer a premise.
4. The central system propagates dependency values and runs rules. Both the dependency system and rule system continue to run until they have both settled.
5. The periphery inspects the values of the central system’s outputs, which represent motor commands.
6. The periphery and world simulation together arrive at as set of proprioceptive propositions (judgments about the success or failure of the agent’s primitive actions) and a set of motor effects (the immediate physical consequences of the agent’s actions).
7. The world simulation updates itself again, and so on ad infinitum.

There is a set of principles regarding RA, discussed on (p. 179):

1. It is best to know what you’re doing. Executing plans is derived from the symbols of the plan, not from an understanding of the situation. I think this suggests a model which makes use of the situation within the symbols themselves. Flexibility relies on understanding of situation and its consequences.
2. You’re continually redeciding what to do. Much of the relevant input from the world is constantly changing, rather than remaining fixed.
3. All activity is mostly routine. Activity is most frequently something which has been done before.

Representation and indexicality

Actions are about the world. This is an important idea! Aboutness ties into intentionality. At the same time, Agre argues that “world models are the epitome of mentalism. On its face, the idea seems implausible: a model of the whole world inside your head. The technical difficulties that arise are obvious enough at an intuitive level.” (p. 225) Knowledge representation is either mentalistic or platonic, and both reinforce the planning approach. Mentalistic models imply the existence of a world model in some memory state somewhere. Platonic models refer to nouns and items in purest abstract, appealing to universal qualities. Models of this nature resemble things like the Cyc project.

For a variety of reasons, I am very defensive of models, but world models have to do with an area that is extremely problematic, specifically knowledge representation. Understanding of an agent or character’s general knowledge, as well as knowledge about other characters and the world, are extremely difficult to model coherently. Existing theories, especially those around mental models, tend to use propositional models, which do not seem appropriate. Agre’s alternative is indexical representation, which is relevant for situational circumstances, but does not seem appropriate for larger scale activity. Agre’s alternative approach derives from phenomenology, especially Heidegger and Merleau-Ponty.

Diectic representation

Diectic representation is a different kind of world representation. The examples given are immediately situational: “the-door-I-am-opening, the-stop-light-I-am-approaching, the-envelope-I-am-opening, and the-page-I-am-turning.” (p. 243) These are diectic entities, which are indexical and functional. They are also immediately given a perspective, as all of these contain some reference to the entity itself. I would argue that these diectic entities absolutely tie into models, but those models are situational ones. There is considerable challenge to the idea of diectic entities as symbols (note Vera & Simon), but their use is very different from that of conventional symbols.

The most important distinction between diectic entities and symbols is their situated and non-objective nature. “A diectic ontology, then, is not objective, because entities are constituted in relation to an agent’s habitual forms of activity. But neither is it subjective. These forms of activity are not arbitrary; they are organized by a culture and fit together with a cultural way of organizing the material world.” (p. 244) This phrasing anchors diectic representations in a cultural basis. Given this context, it would make sense for diectic representations to be variables in defining activities in a simulated world populated by agents.

Conclusion

Mentalism ties into psychic unity (note Shore). “Perhaps the most unfortunate cultural consequence of mentalism is its tendency to collapse all of the disparate inside phenomena into a single ‘mind’ vaguely coextensive with the brain.” (p. 307) Mentalism too yields to objective accounts of reasoning, which proclaims a kind of universality of goal oriented thought characterized by Western philosophy.