Return to People

Faisal Kadri’s Paper Proposal

Understanding and Learning to Reconcile Differences between Disciplines through Constructing an Artificial Personality

Constructing a model of reality is the essence of understanding, and learning is the recursive process of validation and updating of a conceptual model. Our individual and unique models of reality are what make our outlooks so different and sometimes incompatible. Constructing an artificial model of personality offers an opportunity to reconcile different outlooks of disciplines that may seem irreconcilable at first sight. An example is the difference between the cognitive linguistics of Noam Chomsky and B. F. Skinner’s approach to behaviourism.

Chomsky is a linguist who believes that the medium of thought is natural language, he asserted that the structure of sentences is overwhelmingly similar to the structure of thought and that language is the mirror of the mind. So, inherited human disposition to learn language separates (cognitive) humans from (motivational) animals, and the structure of language follows formal logic rules, while animals and their behaviour have “nothing to teach us about our thought processes and languages.” This understanding set the stage for a collision course with the leading animal behaviourist B.F. Skinner in a well publicized critique of the latter’s work.

Chomsky argued that laboratory concepts, such as stimulus, response, and response strength, are inadequate when applied to human behaviour. Skinner argued that there is no reason to assume that animals under laboratory conditions will behave differently from being outside and suggested that animal behaviour extends to humans. Chomsky suggested that language structure leads to “near infinite” number of combinations and it would be impossible to learn them through rote memorization. The near infinite linguistic variety contrasts sharply with a small number of categories of animal motivations. To a cyberneticist this is revealing; while the jump in variety from near infinite to a single thread of behaviour seems impossible, the existence of intermediate non-linguistic stages of reduction is a realistic assumption, and the different varieties point to the likelihood that the arguments of linguists and behaviourists are not mutually exclusive. The artificial personality is constructed on a structure that limits the diversity of motivational contexts, thus reflecting a B.F. skinner scale of diversity and providing a transitional diversity for a Chomsky near infinite scale of linguistic diversity.

Another example is the difference between the higher diversity of the emotions with other motivations. The emotions are displays of feelings; feelings are the subjective representations of the emotions, yet both are separate subjects of academic research and have their own varieties enumerated by scholars, mostly without reference to a structure that can limit their growth. One noted exception is the Plutchik model, where different emotions and feelings are located on two dimensional coordinates over a wheel with 8 spokes, representing 8 primitive emotions which are common with animals and form the basis for higher diversity of emotions. Plutchik placed the primitive emotions in the centre of the wheel and the higher diversity radiating outwards. The location on the wheel describes how the emotions are related uniquely to each other and can be understood as projections from the dimensions of the artificial personality.

In both cases we have a situation of higher diversity that is bound by structure and a lower diversity with no apparent structure, the artificial personality reconcile the diversities by adding structure to the lower diversity.

Observing the differences and similarities between diversities is a process of recursive learning and construction of a structured model of reality.


Skip to comment form

  1. What do you mean by “reality”?

    I have some comments which are not really central to your argument, but perhaps might be of interest, anyhow. I hoep so.

    I once did some work on the relationship between Finnish folk architecture, and Finnish language. Finnish is not a European language and behaves in surprising ways—to an English speaker. Having shown a parallel the question was whether the language determined the architecture, or, possibly, vice versa. I opted for the language determining the architecture: after all my reason for learning Finnish was that I believed that language forced me to think in particular ways (teenage angst). Since that time (I was in my early 20s when I did this work), I have come to the conclusion that the relationship is analogical, that there may be something in what I’ll call “Finnish mind” which that is expressed and revealed in both these artefacts.

    So while I agree with the Chomsky evaluation of Skinner (I’m not sure Skinner thought that his work told us about thinking, other than the thinking that works in the brain making the explanation of his black box mechanism), I’m not sure that I can go along with his rather power-based view.

  2. I am interested in the applications of artificial personalities. What do you see as the purpose of this AP?
    Faisal i know you have completed much work on simulated models and was wandering if an artificial personality could be constructed to act as a secretary. A secretary that can respond to callers and transfer this message to the end user? But the secretary would need to have the ability to “learn” from events to improve the filtering and response repertoire. Ill chat to you at the conference.
    The idea of language being linked to the architecture is not that uncommon and i would agree with you Ranulph. The work of Alasdair MacIntyre and his paper Relativism, Power, and Philosophy sets the scene of a person who inhabits two dissimilar linguistic communities and was socialized in both. This paper gives interesting examples of how conflict arises without intentional means when one tries to apply the thinking of one group to the other. The worldview of one is often part set in the language of the person.
    He talks about a boundary situation in which the two groups or cultures have different methods for carrying out their activities, which includes their ethics, laws, rationality, beliefs and values. How did you cross that boundary (i am assuming that you are not Finnish?) to immerse yourself into a different linguistic community?

  3. Ranulph,
    I see reality as the ultimate source of information, which may not be real in the usual sense. For an architect reality is the desire of his/her client which may seem more solid than the property of concrete or the price of material.
    I don’t know much about Finnish, I guess the structure of language may influence the positioning and functionality of architectural units. A thought occurred to to me regarding languages with truly logical differences and possible influence over structure, in particular in South American Aimara language which sees the future behind the observer and the flow of time in opposite direction to the rest, I wonder if there is any feature in their architecture which reflect this concept..

  4. Philip,
    Artificial personalities become very interesting when modeling reality, an artificial secretary with the capability of stereotyping the personalities of its users is one application, a selection of compatible dialog accordingly is a benefit.
    To me, the most interesting aspect is that such application is an implementation of second order cybernetics. The mere usefulness of an artificial secretary can be achieved with regular artificial intelligence but constructing and modeling reality is second order.

Leave a Reply