An ontology is a formal specification of a domain, that gives the ability to express assertions, make inferences, and reason within that domain.
The PCC ontology (pcc.owl) provides a set of classes and relations covering the domain spanning perception, cognition and communication. This allows us to describe and reason within the domain of what we believe exists, what we experience, what we think and what we communicate. We can analyse such things as perspectives, topics, analogies, etc.
PCC is built upon a conceptual foundation of information theoretic and system theoretic ontologies (SMN-IST Ontology). Hence it is consistent with the fundamental principles of information, information processes, systems, interactions, self-organisation and dynamically evolving complexity.
It is also consistent with SMN-technology, hence any expression within PCC can be animated as an SMN system. Furthermore it is consistent with SMN-theory hence many profound and subtle scientific and metaphysical issues can be explored.
The purpose of the PCC ontology and related patterns is to create a general purpose framework that is useful for people to accurately describe and reason within the domain of empirical knowledge. This current work is an initial exploration into how to create and use such ontologies – it will evolve as more is learnt from the process.
It can also be used as the basis for a non-conditioned conceptual structure with all word labels on the nodes replaced by random numbers, thus leaving only the conceptual structure. Then word labels can be explicitly associated with the nodes to give them conditioned meanings within different contexts. See Roots of Meaning for more on this.
In this introduction we cover the basics of the PCC ontology, its foundational ontologies and various patterns (ways of using PCC).
To fully understand the ontology and patterns described here requires studying the ontologies themselves, however as a form of introduction, below are some diagrams taken from the ontologies; they give an overview. The diagrams have been simplified, hence some relations have been omitted, such as inverse relations and class relations (class hierarchies are shown separately in some cases). This allows the overall structure of the ontologies to be more easily discerned from the diagrams.
At the core of the PCC ontology is the concept of a 'phenomenon' and of particular phenomena and their relations. A phenomenon is anything that can be experienced, conceptualised or represented. An “occasion of experience” consists of several phenomena, an object (that which is experienced), an observable (the experience), a concept (understanding) and a representation (symbol). Note that a representation is not just a linguistic statement or image but could also be a memory or any form of symbolic encoding and representation. Another type of phenomena (shown later) is a state, which is publicly observable. Below is a diagram of the relations between the main phenomena.
Below is the PCC class hierarchy.
Note that entity and relation are both subclasses of phenomenon.
Below, in the two central columns, we see some subclasses of phenomenon and how they relate:
In the central two columns we see phenomena (entities and relations). Looking up the four levels we see the four types of phenomena, (state, observable, concept and representation). Note that all phenomena are able to be experienced as an observable.
The outer two columns show systems and system interactions that can be used to simulate the phenomena. PCC provides a context to describe conceptualisations but because it is built atop a system theoretic ontological foundation that is connected with SMN, this allows the entities to be simulated as virtual systems (more later).
We begin to see above how PCC provides a context in which we can create a universe of discourse about things that we believe exist, that we experience, that we think and that we communicate.
Phenomena are experienced as observables.
Observables are conceptualised by concepts.
Concepts are represented by representations.
Concepts have variants that have contextual relations with other concepts. Hence a general concept can mean one thing in one context and something else in another context.
Concepts are analogous to concepts via an analogousRelation.
Analogies allow us to create a mapping between concepts.
Entities have relations, which connect them with other entities (relation-to-entity relations not shown for simplicity). This allows networks of entities and relations to be made.
Every entity can be experiencedAs an observable entity, and every relation can be experienced as an observedRelation relation.
Anything that is experiencedAs and observable, then conceptualised and represented can be brought into the universe of discourse and woven into the network of phenomena (states, observables, concepts and representations).
Below we see another perspective on the PCC ontology...
This diagram shows how occasions are defined. There is some consciousness that has some occasion of experience, this occasion may be related in some way with other occasions via an occasionRelation. The consciousness is simulated by an atomicSystemEnitity or complexSystemEnitity. The occasion entity is simulated by an occasionSystem, which contains phenomenonEntities that simulate phenomena within the domain of discourse.
PCC is developed atop a metaphysical foundation derived from the system.owl ontology, which depends on the information.owl ontology, which are components of the SMN-IST Ontology. These define the basic concepts required for a system to exist and to experience phenomena. They are also consistent with the smn.owl ontology (see description) and the smn-modelling.owl ontology (see tutorial), hence they can be simulated using SMN.
Below is a brief overview of the foundational ontologies, then we see some example uses of PCC.
Above we see information and an information process. A schema structures an information process, which then transforms data. Data can be formed into complex data structures.
Above we see that a system has a state, which is implemented by data. An information process transforms the data and thereby the state changes and the system is animated. Systems have interactions with other systems and thereby participate in complex networks of system interactions. We can also explicitly represent complex systems and their interactions, however these are not directly modelled in SMN (only atomic systems and interactions are directly modelled in SMN).
Following are some 'patterns' which are examples of ways of using the PCC ontology. There are many basic 'patterns' (types of expressions that are possible within the universe of discourse), which help describe and reason with things such as topics, contexts, analogies and so on.
The PCC ontology provides a universe of discourse within which various expressions can be made. Expressions are composed of instances of PCC classes or their sub classes, which you can extend.
Above we see a diagram of an expression called “stateTopic”.
There are four components to a topic; an object (any phenomenon), an observable, a concept and a representation (all of these are phenomena). Any phenomenon can serve as an object. I.e. a topic can be about state objects, observables, concepts or representations (either entities or relations).
This gives us four basic types of topic in one dimension and two types (entity or relation) in another dimension. Hence there are 8 fundamental types of topic.
Each topic has an object, observable, concept and representation.
These are not necessarily unique, e.g. two topics may share common elements except they have different representations. For example, a particular communication event might involve person A experiencing something and producing some statement (representation). Person B then experiences that representation and produces some statement in response to it. This describes a simple conversation where each statement is an attempt to express one's concept of the previous statement.
One way to use the topic pattern is, if you wish to create some topics, just import topic.owl into your project, then do a deep copy of the topic instance. Then shift the copies to where you want them and customise them. Then the topic.owl import can be removed, leaving only the copied topics. Any pattern can be used in this way.
Below we see an example of two interconnected topics from twoTopics.owl.
For an example use of this pattern see twoPerspectives, which is an initial exploration into the use of the PCC ontology as a conceptual framework within which to conduct analyses of issues and phenomena that involve perception, cognition and communication. It explores the underlying causes of two general perspectives, one a naïve realist perspective and the other a non-naive realist perspective.
Above we see a diagram of an expression of contexts from context.owl. The general concept has two different nuances in two different contexts. Any concept can serve the role of general, variant or context.
Above we see a diagram of an expression of an analogy from analogy.owl.
There are four concepts relating to two people and two particles, which are formed into two groups (sub-expressions) on the left and right by the interconnecting relations.
These sub-expressions would normally form two stacks like in the two topics example above, but here the objects, observables and representations have been left out for simplicity.
An analogy is asserted which creates a mapping between the sub-expressions. Hence the attraction between two people is asserted to be analogous to the attraction between two oppositely charged particles.
This is a simplistic example, but it illustrates the basic use of the pattern. It could have any number of mappings between any number of different sub-expressions.
If you wish to use this ontology, just import pcc.owl into a new Protégé project (see this tutorial, on how to use protégé). Then optionally extend the class hierarchy and instantiate instances then connect up the instances into the desired expression. Detailed tutorial coming soon...
In general we can create conceptual networks describing related expressions about empirical objects and events. These can be connected via representations with the domain of communications and cultural phenomena and via observables and states with empirical data, or via systems with simulations.
There are many possibilities for exploring domains such as logical argument, dialogue, enquiry, group debate, dealing with uncertainty, decision making, information gathering, merging perspectives, translating between domains, etc... There are also core principles that apply across all domains, which can be discovered and refined.
A general map for further extensions could be inferred from the diagram shown below, which was originally made for a rather different purpose.
These diagrams have illustrated some features of an ontology that has been developed to help describe and reason about what we believe exists, what we experience, what we think, what we express and all combinations of these things such as topics, contexts, analogies, dialogues, debates and so on. The aim is to develop an ontology and methodology that allows us to describe and reason with empirical knowledge and complex knowledge of all kinds.
The aim is to create a computational environment that assists people in the task of rational discourse on empirical subjects. A scientific calculator application can assist us in calculations, and a compiler can assist us in creating syntactically correct code. In a similar manner this application seeks to assist with creating accurate descriptions of empirical topics, enquiring into empirical topics, testing empirical theories, reasoning from empirical knowledge, debating issues, sharing 'real-world' knowledge and so on.
If the environment is simple and intuitive enough it could be developed wikipedia style and grow into a vast repository of knowledge about the world in general that could be used for inference and semantic applications.
This is just a start but it is growing...