anarkhos

I

Folk Psychology (hereafter referred to as FP) would have us believe that unity of self is a precondition of rationality. What is meant by unity of self? Before we can address this question, we need to consider personal identity.

What makes for personal identity? The FP concept of personal identity would encompass two dimensions: embodiment and memory. According to this theory, among other things, a person is a continuous, single entity making its way in the world, with access to memories chronicling the biography of the mind and its associated physical body. To examine whether this account of personal identity correlates with reality, let us consider the terms, one at a time.

Folk Psychological Embodiment

FP embodiment is a bundle of cells bounded by an individual's outer layer of skin. The person is conceived to be located in the space occupied by the bundle of cells. It is from this perspective that the world is observed and all of reality is related. The border between the individual embodied and the external world, with the exception of various medical procedures and tragic accidents, is static: the exterior limits of the bundle of cells.

Folk Psychological Cognitive Processing

With respect to mental processing, the individual is conceived to be a single agent, with static embodiment, acting in accordance to a set of propositional attitudes. Rational individuals conduct their lives as single agents in accordance with a single, comprehensive, system of beliefs. To satisfy the conditions of unity and rationality, it is expected that the individual will go about the commerce of life pursuing a consistent integration of goals and desires. When this rationality breaks down and an individual's goals or desires are in mutual conflict, depending upon the magnitude of the internal conflict, the individual is thought to be flighty or irrational. In such a state, wavering occurs.

Wavering is that oscillation in emotion or belief FP associates with disunity. Wavering is not rational. As mental representation, for FP, is in the form of the propositional attitudes, it is not possible to hold, consciously, conflicting beliefs or desires. Thus, as one desire gains hold of consciousness, its contrary is displaced. As the contrary regains, the other is displaced. This oscillation is wavering.

Folk Psychological Memory

As for memory, memory is a collection of experiences, ordered chronologically, and belonging to a single unified self. Any memory, with the possible exception of the earliest and most recent is effortlessly ordered with respect to a chronological "betweenness" relation. All of memory forms a single thread. Even when it breaks down, and there a gaps, the gaps are gaps in a single thread. Thus, all remembered experiences of the same individual belong to the same thread of memory.

Folk Psychological Unity of Self

Thus, barring break down, FP conceives of personal identity as a unity of a static physical embodiment, a set of propositional attitudes, and a single (possibly broken) strand of memory chronologically relating the experiences of the total system. Does this conception of the personal identity, and its resulting demand for the unity of self, conform to experience? Let us consider Lisa.

II

An Alternative to Folk Psychological Embodiment

Lisa is a computer scientist studying philosophy at UCSD. When Lisa was a child, her parents saw to it that she was well instructed in piano. When Lisa isn't hacking AI programs in the Cognitive Science Building, studying philosophy in the library, arguing for revolution at the Che Cafe, or surfing at Black's Beach, she can be found at home playing one of her favorite pieces by Chopin.

As her fingers run across the keyboard of her baby grand piano, eyes closed, body gently swaying like a metronome, her mind is full of music. The piano and Lisa are one as she navigates the world of music. When Lisa plays, it is easy to discern her style. Her feelings mold the notes in just such a way that only Lisa seems able. In the space of music style, Lisa has a location, and it is with respect to this location that she relates herself to other musicians. Some are similar to Lisa. Most are different from her in a great many ways. In the world of music, Lisa's embodiment extends beyond the bundle of cells delimited by her skin and extends to encompass her piano and the acoustics of her room. Her identity is this embodiment conjoined with her location in music style space.

Sitting in the sweaty confines of the Cognitive Science Building's computer lab, surrounded by computer geeks, hands on the Sun Workstation's Jolt Cola stained keyboard, Lisa's mind is lost somewhere in LISP syntax space. Like a swimmer expertly applying a stroke through a warm pool of calm water, Lisa manipulates LISP expressions on her computer screen. The computer she is logged into is somewhere across campus. By means of the Sun Workstation and a network, she is interacting with a remote LISP listener. Lisa isn't thinking about this. She hardly notices her fingers moving as her mind edits the code on the screen. In the world of LISP programming, Lisa's embodiment extends beyond the bundle of cells delimited by her skin and extends to encompass the Sun Workstation, the network, the virtual machine emulating the editor on the remote LISP machine. Lisa has a unique programming style. She feels personally offended when someone criticizes this style for it is an expression of her self -- her LISP self -- the self she compares and separates from other LISP programmers. Her identity is this embodiment conjoined with her location in LISP coding style space.

An Alternative to Folk Psychological Memory

When Lisa surfs with her surfing buddies at Black's Beach, she confidently leaves behind her geeky programming self and her gentle pianist self. Here she's one of the guys: challenging gnarly waves; screaming in the surf; conquering waves with a style that her friends envy and admire.

Exhausted, at the end of a long day of surfing, Lisa and her surfing buddies huddle around a fire on the beach. All through the night they recount memorable waves conquered. The stories progress from one surfing day to another, seamlessly continuous, as if nothing happened in their lives between surf days. In Lisa's mind, recounting her sea-faring conquests, no mid-term examinations, or sweaty nights in the Cognitive Science Building, fit between her surfing memories in chronological "betweenness" relations. They are memories belonging to another Lisa, part of another collection of autobiographies. If someone says the right word, another world of memories might be pulled in and then Lisa would feel that uneasiness we all feel when inappropriate thoughts surface. "Computer Science belongs in the lab, not at the beach! This is a beach party. Give me a break..."

If, however, Lisa were to be asked what was happening in "her" life between two surfing days, no doubt there would be a moment of hesitation as she consciously struggled to integrate the memories of her other selves into a unified narrative. Lisa attempts to remember what major event occurred near a given surf-day and how it would tie into the other domains of her experience. What is this struggling which occurs when we integrate memories from a multiplicity of contexts? Surely, it is indicative of an attempt to synchronize separate threads corresponding to divergent selves, and not a single thread associated with a unified self. It would appear that the totality of memories corresponding to an individual are of a multi-strand quality, synchronized at uneven intervals by the occasional event which invades the experiences of contextually separate identities.

An Alternative to Folk Psychological Cognitive Processing

At the end of the quarter, Lisa throws a party. She invites all of her friends. Lisa didn't think much about what the mix would be like. Busy with finals, she had simply invited every friend she encountered. Many of Lisa's friends do not know each other. The day of the party she frantically darts about the grocery store trying to decide what to buy. Suddenly it occurs to her how diverse she is. She needs several cases of beer for her surfer friends, a fine white wine for the music clique, plenty of Jolt Cola for the computer geeks, and cheese and crackers for the philosophers. "What kind of party is this going to be?" she wonders, with rising dread.

Her door bell rings, and one friend after another pours into her tiny apartment. The surfers are rummaging through her vinyl collection, arguing over what Grateful Dead track to put on first, the philosophers are collecting at the other end of the room, arguing about Nietzsche. Some of her music friends have retreated into the bedroom to get away from the Dead while the computer geeks are sweating in the kitchen, scratching their messy hair, discussing the virtues of strong data typing. In the middle, Lisa finds herself experiencing an identity crisis.

Lisa knows that all of these people are her friends. She fears that if she hangs out with the geeks in the kitchen, the surfer dudes will think she too is a geek. She is a geek, but she's not only a geek. Her music friends, she fears, are on to her taste for the Dead. Will they lose respect in her? Will they fail to take her seriously as a musician? What do the philosophy friends think of the meat-heads by the sound system? Most importantly, if she takes on any one of her identities, will her other friends consider her two-faced? Haven't we all been Lisa at one time? If we all have, then why should this dilemma be characterized by two-facedness? The point is: we are all a disunity of selves! Folk Psychology torments us by imposing a false belief about what it is to be a person. We are all two-faced (no! multi-faced) and we should be proud of it! It is a great testament to the plasticity of our minds and its representation of personal identity. What might this representation look like? We consider this presently.

Identity Spaces

We begin with identity spaces. Identity spaces are domain specific character spaces. A domain is an area of human endeavor. Domains include: music, programming, poetry, philosophy, family relationships, love relationships, sports, etc. Domains encompass all of the specific activities in which a person may engage. A character space, within a domain, is a vector of degrees to which a person, engaged in a domain, partakes; where each degree corresponds to a method of behavior within a domain. Given any domain, a person's identity, with respect to the domain, is that collection of regions within which the specific behavioral characteristics of the person typically fall. These regions are prototypes of the person's identity within the domain specific character space. A person's total identity is the collection of all personal identity prototypes, across all domains.

This representation would be a natural fit to the Connectionist Model of Parallel Distributed Processing. What phenomena could we account for with this representation? The answer is that it buys us much more than the Folk Psychological model does.

Using this model, an individual may represent not only their own personal identity, but the identities they attribute to others. Others encountered within each domain will be associated with prototypes in the same domain specific character space. The similarity between the self and others within a given character space will indicate similarity by inter-prototype distances. Given a specific domain, those prototypes occupying regions of character space closest to the personal identity prototype will be those associated with individuals perceived to be most similar within the given domain. Depending upon the individual's degree of confidence within a domain, and the degree to which they are competitive, the distance between prototypes would be a good indicator of friendship potential. The overlapping of the prototypes of others with any of one's personal identity prototypes might well give rise to that feeling of kindredness we sometime feel (a feeling, interestingly characterized, by an inexplicable feeling of "closeness").

At times we all behave in ways which we wish to disassociate with our identities. When this happens we say things such as: "I wasn't myself that day!" or "I'm sorry, it wasn't me!" Using this model, such incidents would be just when our behavior failed to map into our personal identity prototype for the domain in question. In Connectionist terms, the neural network would not classify the pattern as "I", but as "someone else", hence "not I."

As for embodiment, the characteristics defining the domain specific character space would be in terms of behavior meaningful to the domain. Meaningful behavior would not be the intermediary steps to affecting the desired behavior, but the totality of the behavior. For example, in the case of playing the piano, a chord is not a hand position as it strikes a specific set of keys, but the entire causal chain, from decision, to hand formation, to striking, to the actuation of the internal mechanisms of the piano, to the emergence of a musical component arising from the piano. This would be one element, and in the abstract space of domain specific character space, it would have a single characteristic. Thus accounts this model for domain specific embodiment.

The model at hand permits the simultaneous representation of an individual within multiple domain specific character spaces. In the FP account, one must waver between conflicting feelings. Our representation does not have this problem. Let's consider a specific example.

My friend, Bridget, and I are competing in a spelling bee. We are tied and a ten thousand dollar prize hangs in the balance. The next word determines the outcome. Bridget wins and I am left to my conflicting emotions.

I am upset because I lost. I could really use that ten thousand dollars. With it I could pay off my debts and take that vacation in Greece for which I've been longing. On the other hand, Bridget is my friend. I love her very much. I know Bridget could use the money too. I'm very happy for Bridget.

According to the Folk Psychological model, my feelings are wavering. At one moment, I am angry about losing. At the next moment, I am happy for Bridget. Back and forth my feelings go. In time I'm expected to come to terms with the situation and take a detached view. "On the whole, the situation is good, for it could have been neither of us!"

According to our model, two domains are in conflict here. First, there is the friendship domain. In this domain it is characteristic of me to feel happy when my friend encounters fortuitous events. The other domain is the domain of self-interest or acquisition. I want that money. I just lost it. In this domain it is characteristic of me to feel angry when a sum I hoped to win is denied me. I am angry. Within the friendship domain, it is rational for me to be happy about Bridget's victory. Within the acquisition domain, it is rational for me to be angry about Bridget's victory. I am simultaneously happy and angry. I am not wavering.

According to the FP theory, I should feel guilty about my anger. Bridget is my friend. I should be happy about her victory. Under the FP theory, I must be one or the other. According to our new theory, there is nothing to feel guilty about. I am happy about Bridget winning. I am also true to myself, because I am angry about her winning.

If we reject FP, what follows? What follows, I believe, is far more humane. In a world where personal identity is accepted as a disunity, we would all feel more comfortable pursuing our many and varied talents. We would expect our emotions to be chords of feeling emanating from a multiplicity of personal identities. No longer would we feel guilty about being true to ourselves. Terms such as "torn" and "confused" might give way to "divergent" and "various". We would cease tormenting ourselves over our inconsistencies and embrace our totality of selves as a bundle of capacities, deserving of full development.

With respect to character flaws, they would cease to indicate a flaw in a given person generally, and instead would indicate a failure within a specific domain. Despite our culture's obsession with pigeon-holing individuals as generally untrustworthy for specific failures (consider our obsession with the personal live's of our political leaders), we would be able to consciously accept what we have all secretly felt for so long: that an individual can be of questionable character in one domain while being of great value in another.

Clearly, this model accounts more accurately for personal identity, external behavior, and the internal experience of being than does the FP model's accompanying demand for unity in feeling and character. Any system based upon the propositional attitudes will fail to account for the simultaneous multiplicity of feelings and conflicting behaviors of the human person. For this reason, and others not covered in this paper, we should reject it.

I leave the reader to ponder the explanatory benefits of breakdowns in this system of representation. One direction which begs exploration is Multiple Personality Disorder (hereafter referred to as MPD). Where FP might attempt to account for MPD by entertaining a fragmentation of the self, perhaps our alternative representation could account for MPD in terms of a break-down in multiplicity. Specifically, upon the activation of a given prototype within a domain specific character space, MPD would be the phenomenon of the inhibition of all personal identity prototypes within other domain specific character spaces. We are all familiar with stories of specific talents associated with the individual personalities of such persons, the manner in which particular personalities are called in to "handle" particular situation types, and the theory that these personalities were constructed at an early age to separate the "self" from unspeakable torment. All of these phenomena make sense within this model. Not so, it seems, for FP.