| the scientific landscape |
| representation (models, semantics) |
| logic & reasoning | complexity & metrics | category of relations | ||||||||||
|
| meaning
| linguistic glossary |
| components & datatypes | |||||||||||
| truth & reality | ||||||||||||||||
|
Location: http://www.cs.mun.ca/~ulf/gloss/cats.html.
By Ulf Schünemann since 2002.
Please mail any comments.
| ||||||||||||||||
«The first step in designing a database, a knowledge base, or an object- oriented system is to select an appropriate collection of ontological categories. In database theory, the categories are usually called domains; in AI, they are called types; in O-O systems they are called classes; and in predicate calculus they are called sorts. Whatever they are called, the selection of categories determines everything that can be represented in a computer application or an entire family of applications. Any incompleteness, distortions, or restrictions on the framework of categories must inevitably limit the flexibility and generality of every program and database that uses those categories» [TOC 670].
«The ordinary language expression `there are' is ambiguous,
as it designates two different concepts:
the logical concept something
and the ontological concept of existence.
Logic takes care of the former, analyzing it as the existential quantifier,
which it might be better to rechristen particularizer,
or indefinite quantifier ....
Surely most contemporary philosphers hold that
formalizes both the logical concept "some"and the ontological concept of existence.
Consider the statement "Some sirens are beautiful", which can be symbolized
"(
x)(Sx & Bx)". So far so good.
The trouble start when the formula is read "There are beautiful sirens".
The existential interpretation is misleading because it suggests
belief in the real existence of sirens, while all we inteded to say was
"Some of the sirens existing in Greek mythology are beautiful"» [MB3 155].
| Construct [MB3 117] | Concept «Concepts, such as the notion of a thing, are the building blocks of propositions - such as 'All things change'» [MB3 116] | Individual | |||||
| Set | |||||||
Relation
| Proposition = value of propositional function/predicate/attribute
| Propositions are «constituents of contexts - such as the set of all propositions concerning dogs - and of theories» [MB3 116] Context (closed context = theory)
| | |||||
[MB3 54] «Finally a word or two about substantial existence as distinct from
existence in general. The existence of simple individuals can be ascertained
only by scanning the entire set S [of concrete objects].
If a belongs to S and is not the null individual
we say that a exists. That is, we explicate "a exists"
as "a != U & aS".
This concept of existence can be called bare substantial existence.
It suffices in the theory of substance, where the only individuating property
is composition. ... Note that this is a notion of ontic or physical existence
to be distinguished from that of conceptual existence. In formal science to
say that an individual a exists it suffices to assume or to prove
that a belongs to some set that has been satisfactorily characterized.
In ontology we have no use for arbitrary sets except perhaps as auxiliaries
devoid of ontological import: here to exist (substantially, physically) is
to have a number of substantial properties, among them that of associating
with other nonceptual objects.»
| See the | ontological status of types/universals |
| Particular | Location | SPACE | a region of absolute space | ||||||||||
| TIME | an interval of absolute time | ||||||||||||
| Object | Concrete Object | CONTINUANT | have a location in space, which may vary over time;
their parts are continuants, not occurants.
(eg, an apple) | ||||||||||
| OCCURANT | are "generated" by the behavior of continuants over time;
they rigidly depend on the continuants which take part in them;
have a unique temporal location,
spatial location is bound to that of the partaking continuants;
their parts are occurants, not continuants.
(eg, the fall of an apple) | ||||||||||||
| ABSTRACT OBJECT | have no location in space nor time
(eg, Pythagora's theorem) | ||||||||||||
| Universal | Property | CATEGORY | [+rigid, -IC], carries no identity criteria (IC)
(eg, location, object, event[dependent]) | ||||||||||
| TYPE | [+rigid, +IC]
(eg, person[+independent])
| ||||||||||||
|
But where are the abstract/conceptual objects in the GOL ontology? - Note that the urelement:set dichotomy is not meaningful for them: «Whether an object is an individual or a set is of interest to ontology: if the former then it may be either physical or conceptual, if a set it can only be a concept. But whether a conceptual object (a construct) is an individual or a set is metaphysically and even mathematically irrelevant. This is for the following reasons. First, what is a (conceptual) individual in one theoretical context may become a set in another and conversely: it all depends on the fineness of our analysis. Second, some set theories draw hardly any distinction between sets and individuals» [MB3 52]. «To us the basic dichotomy in any set of objects is not that between individuals and sets but that between physical objects and conceptual ones.» [>] |
| Entity | Urelement | Individual (instance of 0..1 universals)
| Chronoid (instance of Time universal)
| Topoid (instance of Space universal)
| Substance (bears >0 moments [inherence relation])
| Moment (existentially dependent on substances [inherence relation])
| Quality (one-place moment)
| Relational Moment (instance of relational universal)
| Universal (with >0 individuals as `instances', and 1 extension-set)
| Quality Universal (with qualities as `instances')
| Relational Universal (with relational moments as `instances')
| ... Time (chronoids as instances),
Space (topoids as instances) ...
| Set (with entities as `members')
| Extension (of >0 universals)
| Extensional Relation (extension of >0 relation universals)
| ...
| ...
| | |||||||
Frame- and environment-dependent mutual properties.
«A mutual or relational property of an entity may or may not depend causally
upon some other individual. For example velocity is a mutual property
since it depends on both the moving entity and the reference frame,
but it is not caused by the latter. The same holds for distance, duration,
frequency, mass, temperature ...: they are frame-dependent
but not causally dependent upon the reference frame, which is assumed to be passive.
On the other hand certain mutual properties are environment-dependent.
This is the case with the force on a body, ... solubility, ...
the performance of a student ...: in all these cases the environment
excerts an influence on the entity of interest. All the phenomenal properties,
such as color and perceived loudness, are mutual properties of this kind,
i.e. they depend not only upon the object-in-the-environment
but also upon the subject or perceiver» [MB3 66].
«[I]t might be thought that,
since all relations can be reduced to binary relations (Quine, 1954),
the n-arity of a predicate is unimportant. This may well be so,
but such a logical reduction need not be construed as mirroring anything in reality.
Our planet does not cease to revolve around the Sun between Mercury and Saturn
at the moment ternary relations were shown to be reducible» [MB3 69].
Unarized dichotomic/Boolean properties.
«Fortunately we can introduce a remarkable uniformity,
and thus pave the way for the search for structure,
by [replacing]
every mutual property by a bunch of intrinsic properties. ...
Consider the property of falling. ...
In science, this is analyzed as a binary predicate F
such that `Fxy' is interpreted as "x falls on y".
We can now freeze the second argument, i.e. take it forgranted
that whatever falls does so on a fixed body b such as our planet.
That is, we can form the pseudo-unary predicate Fb
such that
Cf. Phenomenology:
«Phenomenology distinguishes sharply between perceptual
properties on the one hand, and abstract properties on the other.
Consider two white billiard balls, called A and B. The white colour
of A, which one can see with one's eyes, is said to be located in
space where A is. The white colour of B, similarly, is taken to be
located where B is. Furthermore, it is maintained that the colour
of A is not identical with the colour of B, since they are located at
two different places. The same shade of colour, according to this
analysis, divides into as many 'colour instances' of that shade as
there are individual things with this colour shade.
Concrete objects have no (unary) formal/conceptual properties
(hence mathematics cannot illucidate them):
«[C]oncrete objects (things) have no intrinsic [ie. unary]
conceptual properties, in particular no mathematical features. ...
What is true is that some of our ideas about the world,
when detached from their factual reference, can be dealt with by mathematics.
... not the world but some of our ideas about the world are mathematical.» [MB3 118f].
Conceptual objects represent concrete objects.
«The relation between the predicate and the corresponding property
is that of representation: the former represents the latter
... [N]ot things but our models of them have mathematical properties,
and this because we conceptualize substantial properties as functions.
This mode of representation is so deeply ingrained in our habits of thought
that we often mistake the deputy for his constituency» [MB3 106].
The Stoics had a list of four categories:
substrate, qualified, disposed, relatively disposed
[x]
Event.
«An event is sometimes defined as a change (for example, the loss
or acquisition of a property by something) ...
However, many theories of events include states that consist in things' having
(or retaining) properties (e.g. the lawn's staying wet) as well as changes that
consist in their acquiring or losing them (e.g. the lawn's becoming dry)»
[x].
«Every collection of events is a set of changes in some thing or other.
We have found no use for the fiction that the are changes
that fail to consist in modifications of the states of some thing or other.
Anyone claiming that there are such changes in reality should exhibit
empirical evidence to this effect and proceed to build a theory of such
thingless (or immaterial) changes.
If we change the value of the parameter, say to
«Every one of [the unarized predicates] may be taken to represent
an intrinsic property and moreover an individual [ie. dichotomic/Boolean]
one, for it is either possessed or not possessed by the corresponding
substantial individual» [MB3 71].
4. Resultant and Emergent Properties
«Let Pp(x) be a property of an entity
x
S
with composition C(x) [SUPER] {x}.
The P is a resultant or hereditary property of x
iff P is aproperty of some components y
C(x)
of x other than x;
othewise P is an emergent or gestalt property of x»
[MB3 97].
Interaction of Dichotomies
1. Properties can be particular/individual, not only universal
One may talk of, e.g., "the paleness of Socrates", as opposed to the paleness of everybody else.
However, ontological economy may speak to eliminating such individual properties [x].
Understood as an abstraction from individual properties,
the same quality may be possessed by more than one object
(in a very different way from the common possession of a subtantial thing, like a yacht)
[x].
However, all of these instances are instances of the same colour
shade. There exists, therefore, according to phenomenology,
also the abstract colour shade of which the instances are
instances. Let us call this abstract colour the 'universal
whiteness'. Phenomenology asserts that there is not only a direct
perception of instances of whiteness, but also a sort of direct
perception of the universal whiteness. This perception is called
'eidetic intuition'. By means of eidetic intuition we have knowledge
of the essential features of the world. Phenomenologists call
such universals essences.» [x]
Individual states in object-oriented modeling
have been proposed by Bock [Conrad Bock: A More Object-oriented State Machine; JOOP Jan 2000].
He describes an object-oriented model of Persons
whose health state is modeled as a state machine
were states Well and Sick can alternate.
In the Sick state only, the Person is taken care of by a Doctor.
On the instance, an individual Person like John
has a corresponding individual instance of the state model
with its individual states JohnWell and JohnSick/
JohnWell and JohnSick are considered state types (classes)
which are specializations of Well and Sick.
An instance of state JohnSick
is, e.g., the 23rd time when John was in state JohnSick.
In that state instance JohnSick#23,
John was/is taken care of by Doctor Susan.
health .--------------------.
model | .------v | care
Person---------| Well Sick<---|--------Doctor
^ | A ^------' A | ^
: `---|----------|-----' :
: | | :
: «instance» | | : «instance»
: | | :
: health .---|----------|-----. :
: model | | .------v | | :
John-----------| JohnWell JohnSick | :
| ^------' ^ | :
`----------------:---' :
: :
«instance» : :
: care :
JohnSick#23<-------Susan
2. Properties/universals can be substantial, or formal
navigation bar: Universals (properties, attributes, features, forms, types)
object
= universals (shared)
+ X (unshared)
concrete obj.: properties
<-> concept: attributes
individual: properties
<-> particular: features
properties
<-> classes
ontological status of universals
attributes = properties?
«All object have properties. If the objects are conceptual or formal,
their properties will be called formal properties, or attributes
or predicates for short. If the objects are substantial individuals,
their properties will be called substantial properties, or
properties for short» [MB3 58].
P is a substantial property, PProp,
iff some substantial individuals possess P, ie.,
(Ex)(x
S & x possesses P)
[MB3 71].
Substantial properties are in re, while
conceptual universals (predicates) are entita rationis
(post rem is represeting substantial universals preexsting knowledge,
or ante rem if anticipating experience).
-> ontological status of universals
Prop
is a (substantial) universal in a set
T
S of entities
iff P's scope S(P) = T.
(cf. scope)
Attr
is a (conceptual) universal in a set
T
C of constructs
iff A's extension E(A) = T.
[MB3 105]
«[T]he so called formal properties of things, such as number and shape,
appear to be full fledged conceptual universals inherent in concrete entities
and thus defy the substantial/conceptual dichotomy ...
Thus numerosity is a property of any collection whatever the nature of its components.
True, but then a set is a concept. Surely dogs are normally four legged,
this being a substantial property of theirs.
The corresponding mathematical property is the cardinality of the set of legs of a dog.
On other words four leggedness is a substantial property to be distinguished
from the mathematical property of the set of legs of a quadruped» [MB3 106].
«Intrinsic [ie. unary] properties are either of [physical] things or of [abstract] constructs.
One the other hand a mutal [ie. relations] (e.g. binary) property can link things
with constructs. An example of a mutual property of this kind
is that of representation, such as it occurs in the statement
`Proposition p represents thing b'» [MB3 118].
-> continue quote:
A concept (with attributes) represents a substantial individual (with properties).
-> substantial properties must be representable by propositional functions
3. Aggregates =/= Sets
navigation bar: Wholes and Parthood
wholes & system composition levels
part-whole relations
different kinds of parthood and compositeness
aggregates (concrete)
=/= sets (conceptual)
ontological status of wholes vs. parts
holism, atomism (allied with reductionism), systemism
[MB3 52, where we replace Bunge's sign for the universe to `U']
«While the null individual, a concept, is (indentical with) [],
the world (or reality) is said to be denoted by the symbol `U'.
In our theories of substance the world is an individual but not just one more
individual: it is the entity that contains as components all other entities.
But this is as far as our theories of substance go: they give no details about
the structure of U. On the other hand they do assign definite structures
(semilattice in one case, lattice in the other) to the furniture of the world,
i.e. S.
We emphasize that S, the set of all entities, is not the same
as the aggregate or whole [S] composed of all physical objects. The
difference between the concept S and the entity denoted by [S] = U
illustrates the construct/thing dichotomy. To us the basic dichotomy in any
set of objects is not that between individuals and sets but that between
physical objects and conceptual ones. That is, we assume that every class
O of objects is split into a class C of constructs and a class
T S of substantial individuals ...»
Mixed Categories
Classical Categories
Aristotle divided everything there is into
10 ``logical categories'' (supreme genera)
based on an extension of his substance:accident dichotomy [AUOOP].
mode of being subdivision category
substance
exists in itself and not in another
SUBSTANCE
accidents
exist not in itself but in something else, i.e., ultimately they belong to or inhere in some substance
an object's
intrinsic
accidents
absolutely intrinsic
inheres in one oject alone
QUANTITY
QUALITY
(eg. courage) - something which can be possessed by, and attributed to, something [x].
C.f. primary qualities (e.g., shape) vs.
secondary properties (e.g., color) [x].
relatively intrinsic
inheres in the object but refers to another object
RELATION (internal/external)
-> subclassification of relations
extrinsic
accidents
causal looked at the object as a cause ACTIVITY
in the object as an effect PASSIVITY
measured by time WHEN
by place WHERE
according to the disposition of internal parts (relative place) POSITION
other
(e.g., decorations, clothes, ownership ...)
HABIT
Dölling's categories
«Taking the ontological distinctions expressed in natural languages
seriously, my starting point is that we conceptually disciminate between at
least the following categories of entities:
things like dogs or cups, personst like men or pupils, stuffs like gold or
water, configurations like piles or heaps, groups like teams and families,
institutions like offices or schools as well as kinds like kinds of things
or kinds of stuff.» [OntDom]
«[E]ach of the domains .. includes two sub-domains: the domain of atomic
entities and the domain of plural entities of the category in question.»
[OntDom]
entity object instance of kind
physical object
stuff constitutes thing
aggregate
thing (of stuff), several may constitute configurations
non-person
person constitute group, associated with institution
configuration (of things)
social object (of persons)
group
institution
kind (of objects)
Grossmann's categories and STATE OF AFFAIRS
Mainly from [OntRed 177-190]
«If I were tempted to make deep pronouncements in a calm manner,
I would say that it is an ontological feature of our world
that individual things are characterized primarily by the properties they have,
while numbers and classes are primarily (or, perhaps, only)
characterized by the relations numbers have to other numbers
and classes have to other classes.» [OntRed 54f]
ENTITY
STATE OF AFFAIRS
"discovered" 1902/10 by Meinong
as the intentions of judgements
and assumptions [x]
=/= proposition which were introduced as that
which exists for every true and false sentence,
so that logical connectives always have some existent
to relate [OntRed 42]
Grossmann has no use for propositions.
since his connectives (like the relations of intentionality,
constituent, and, obviously,
exemplification [of state-of-affairs and of entity])
can also relate non-obtaining = non-existing states-of-affairs!
(obtaing = existing states-of-affairs are a.k.a. facts)
Quantified vs. non-quantified (atomic vs. molecular (contains negation and/or connectives))
Equal if containing the same constituents in same order and number
THING
INDIVIDUAL mainly: temporal and spatio-temporal
STRUCTURE in particular: spatial and/or temporal
CLASS
RELATION attached to (two) entities.
PROPERTY The main bearers of properties are individuals and (some) structures.
But also a property (eg, red) can have a property (eg, being-a-color)
QUANTIFIER numbers, "all", "some", "none", "the"
EXISTENCE attached to "existents" = all things, and obtaining states-of-affairs
(existence is not a "property" with which existents are related by "exemplification")
NEGATION attached to non-obtaining state-of-affairs
(there is no relation of negative exemplification)
Relations, Roles, Contexts
Individual / Role / Context
A somwhat similar distinction between the individual object,
its role towards others, and that relating context
in which objects have roles towards one another
seems to be expressed in Peirce's Firstness, Secondness, Thirdness (1891)
[from TOC 677f]:
«"First is the conception of being or existing independent of anything else.
Second is the conception of being relative to, the conception of reaction with, something else.
Third is the conception of mediation, whereby a first and a second are
brought into relation"
An act of stabbing is a Firstness that can be recognized by objective
criteria at the instant it happens. No other events or mental attitudes
need to be considered in identifying an act of stabbing. Killing is a
Secondness that depends on whether the victiim dies. ... [T]he act of
stabbing does not become a killing unitl after the event of dying. Murder
is a Thirdness that depends on the motives of the agent. ...»
physical things information things sign
Firstness entity form representamen
Secondness role proposition object
Thirdness circumstances theory interpretant
States and Events
Surprisingly enough, systems theory has so far failed to give an
exact and general enough analysis of the concept of the state of a system.
To begin with the usual accounts (e.g. Zadeh and Desoer, 1963; Mesaraovi'c
and Takahara, 1975; Pabdulo and Arbib, 1974)
do not apply either to continuous systems such as fields or to quantum-
mechanical systems. Moreover the concept of a reference frame has no place
in those accounts, probably because it is not needed in automate theory,
electrical network theory, and a few other theories. Yet the concept is
central to many other theories. ... For example, it occurs in the
explanation of the working of an electrical motor - not to speak of the
underlying theories of mechanics and electrodynamics. The notion of a
reference frame is in fact so central in physics that the states of any
real physical system are relavtive to some frame or other ...
A second concept one misses in systems theory is that of a law. ...
[T]he general concept of a law - the philosophical concept as distinct from
a particular law statement such as Ohm's - ought to play some role in
systems theory if only for a mathematical reason. Indeed, but for the laws
- which place restrictions upon the ranges of the components of the
function F occurring in any functional schema of a thing - we should
accept the usual characterization of the collections of possible states (or
state spaces) as vector spaces or even as inner product or metric spaces.
The existence of laws ruins this characterization. ...
Nor have metaphysicians exactified the notion of state ...
Likewise the various systems of inductive logic employing the concepts of
state and of state description fail to analyze them in a manner consistent
with the concepts of state occurring in science.»
Events can be regarded as identical
iff they have exactly the same causes and effects (Davidson's first answer),
or iff they occupy exactly the same places at the same times (Quine and Davidson's second answer),
or iff they consist in the same objects' having the same properties at the same times (Jaegwon Kim).
[x].
Nevertheless once in a while philosophers and even scientists
are found claiming that there are such thingless changes. ...
One of them is the information process, sometimes regarded as not being
based on the transport of matter or the propagation of fields.
The reason for this misconception is that statistical information theory
is a black box or phenomenonological theory so general that it
disregards the precise kind of signal carrying the
information, the information transmission mechanism, and the kind(s)
and quantities of energy involved. For this reason information theory
is a nonphysical theory: it is instead a theory qualifying as a piece
of scientific metaphysics, since it deals in an extremely general way
with a genus of conrete things. But, of course, information is a property
of certain physical (or chemical or biological) processes, namely signals,
which are processes of a kind. No signal, no information transmission.
And signals, let us repeat, are chains of events occuring in concrete things:
the transmission of information consists in events propagating across space
and carrying energy.
Remove such real processes and only parapsychological anecdotes remain»
[MB3 271f, underlining added].
Ulf Schünemann 250402, 120502