Conceptual Basis

1. Overview

The Cognitive Reality Ontology (CRO) is an upper ontology designed on the basis that what we know about the world is available through mostly through existing information plus small incremental additions from sense data produced by the knower. Informally, we might say that information is (nearly) all we can count on. As such, CRO accounts for all categories of entity that may be represented or referred to by information, plus categories covering information entities themselves. Information is assumed to be created by agents within a social reality. Social reality is understood in its widest sense, i.e. as the locus of all human and related agent activities, including the domains of business, engineering, scientific, legal, government and financial.

CRO thus encompasses anything that may be described, defined, declared, ordered, referenced or otherwise stated in a social reality. Its scope is consequently greater than purely materialist or temporal-spatial ontologies, covering not only material entities and materially-dependent processes, but agents, information, models and abstracta.

Agents are considered in CRO as capable information-processing entities, physically inhering in material bodies, either biological or automaton. Since CRO is concerned with social reality, agents are treated as independent physical entities, contrary to the material-first approach in ontologies such as BFO.

CRO also treats logical information items as independently existing entities. Although information physically exists only in its physical representation, i.e. media such as paper or cloud databases, agents within social systems treat information as existing in terms of its content (i.e. meaning) and don’t generally care about its material bearers, other than for purposes such as forensic investigation, cloud pricing etc. The meaning of information only 'exists' subjectively in the minds of those who apprehend it (humans and arguably AIs). It is nevertheless this form of existence that we treat as if it were material; accordingly, when someone asks for a certain legal document, they are looking for the document content, with the particular physical bearer (e.g. a markdown file on a file system) simply being the most convenient means available at the time.

CRO understands a social reality as a container of physical entities, information and agents, and as inhering in some world, i.e. a spatio-temporal universe with history. This may be our veridical physical world, or a virtual world, for example created by a simulation, known only via the information we have about it. Since real-world simulations are of increasing importance in public health, economics , climate science, and professional training (medicine, aviation, etc.), it is essential to have a solid solution for reasoning about such alternate realities.

Concretely, CRO supports the building of information models that allow us to treat information about (say) a historical epidemic in the real world and a simulated epidemic (i.e. an alternate reality) in the same way. Inferential reasoning is performed against the information within a given reality.

The same approach enables systems to treat fictional worlds such as that of Sherlock Holmes as an alternate reality whose ontology is the same as our real world. Even fictional realities containing supernatural phenomena (Odysseus, Tolkein, CS Lewis etc.) can be managed with slightly extended form of CRO.

CRO treats abstracta, such as mathematics and geometry, as having virtual existence only, outside of time and space, and consequently any world. True abstracta such as calculus are therefore the same for all worlds, and only find material existence when physically represented on media within some world. As for information, abstract entities are understood to virtually exist when mentally instantiated by agents within some world.

CRO supports the creating of sophisticated information systems, but is not on its own sufficient to define the complete semantics. Most semantics are defined by models which rest on the ontology. As such, model building, models and languages of models are of central importance, and are therefore accommodated in CRO as first order categories.

A final core tenet of CRO is that since most information available at any moment for some professional or other purpose already exists, and was created by third parties, the provenance and quality of information are explicitly described in the CRO and derived information models.

2. Justification

The Cognitive Reality Ontology (CRO) is an upper ontology whose design is based on the assumption of pervasive epistemic dependence of modern civilisation. We take it as a basic condition of cognitive existence (i.e. the existence and mental function of agents) that the vast majority of knowledge available to any agent in a modern, specialised society, including one’s own areas of expertise, comes from existing information sourced from third parties, such as media organisations, scientific institutions and other expert sources. Following cite:[Hardwig_1985], McLuhan, Alvin Goldman and others, we assume most agents to be heavily epistemically dependent on sources of knowledge existing prior to any direct observations they may make.

scientist-library
Figure 1. Epistemic dependence

We distinguish two kinds of 'knowledge' here:

  • categorial knowledge, i.e. knowledge of universals, classes or categories - often known as subject matter knowledge or general knowledge;

  • knowledge of individuals, i.e. knowledge in the narrow philosophical sense of a particular person, storm or criminal trial; may be known as a 'case' or 'subject'.

An example of the first is a lawyer’s knowledge of US corporate law; and of the latter, the same lawyer’s knowledge of the case Dodge versus Ford Motor Co., 1919, or a new case she has just taken on.

Other than for specialised researchers, almost all categorial knowledge is pre-existing for most agents. Knowledge of individuals usually is as well, due to the general way that work concerning healthcare, equipment maintenance, law, finance etc, proceeds based on record keeping over time.

Examples of epistemic dependence include the following:

  • a private individual’s knowledge of current affairs is almost exclusively due to media organisations and/or information shared on social media;

  • a private individual’s knowledge of academic subjects is almost exclusively provided by pedagogic sources, including text books, lectures etc;

  • a physician’s knowledge of the condition of diabetes mellitus is based almost entirely on prior medical research and publications;

  • a physician’s knowledge of a new patient is based on the existing health records, and patient-provided information, with new information being created by the contemporary investigation or treatment undertaken by the physician;

  • similarly, prior to becoming involved, any agent’s knowledge of a car, public asset, mortgage, or crime investigation is provided by existing maintenance or case records.

Human and institutional agents do of course still generate and use sense data, mainly via reliable instruments, although even in modern medicine, there is no machine that can replicate an experienced physician’s near instant assessment of COPD (chronic obstructive pulmonary disease) from across the room. New sense data are understood as sources of incremental addition to any body of existing knowledge, rather than the main source of truth about any phenomenon. Exceptions include leading edge research on genuinely novel phenomena, such as a new category of virus, or gravitational waves.

Philosophically, CRO makes the assumption that we are forced to mostly live in a form of Plato’s cave, with only occasional sojourns outside. Given that, we endeavor to make the light flickering on the walls a representation of the highest possible fidelity and reliability, of what lies outside.