Conceptual description of individuals

Conceptual and relational concepts

The description of the subject area begins with the selection of objects, and we have to do this without having any classification, which we could use as a template for distinguishing them. We only have our own idea of ​​the world (and the subject area), our own conceptual grid, in which typical instructions are stitched: this is a dog, this is a machine, this is a contract. Moreover, it is noteworthy that in order to bring an individual under the concepts (“dog”, “machine”, “contract”) we do not need to analyze, fix attributes - concepts appear in our thinking in the form of Gestalt-images, with which the comparison of the environment individual. The last remark suggests that any description of the subject area should begin with the formation of a list of concepts, with the help of which we will be able to select objects without paying attention to their attribute description.

In the previous text, Classes, sets, groups, systems, it was proposed to distinguish between two types of concepts: conceptual and specializing . The conceptual concept characterizes an individual from his appearance / birth to extinction / death (A dog bug for life), and specializing only occasionally in individual segments of the individual’s existence (for several years the Bug was a sled dog). It can be said that the conceptual concept indicates what the individual is in essence, and not situationally. The American philosopher Lynn Radder Baker, denoting the conceptual concept as a “primary view,” described its specificity as follows: “You can check whether a view is primary or not primary by asking the question: if this (for example) cat would not be a cat, would it continue to exist ? The answer is negative; the cat is the primary species. And if we asked, would a student exist if she were not a student? The answer is yes; the student is not the primary view ”[Baker, 2011].

The formal feature of a conceptual concept, unlike a specializing one, is the immanence of the first. We fix the individual's falling under the conceptual concept outside and prior to any relation with other objects, while in order to single out the specialization it is necessary to indicate the relationship of the individual with another object. To answer the question “what is a husky?”, It’s enough for us to point a finger at a husky or, if it is not there, begin to enumerate typical breed attributes. But in order to clarify the meaning of the term “sled”, it will be necessary to point out the sleds and explain the concept of “riding in a harness”, and only then conclude that the Bug is “sledding” because it relates to “sledges” and “driving”. Or, for example, a person falls under the concept of "student" not just by itself, but only because of the presence of the person-university relationship. But this man is a man immanently - regardless of any relationship with other objects. Therefore, “like” and “man” are conceptual concepts, and “sled dog” and “student” are specializing. Although after these explanations, the specializing concept would be more precisely called “relational”, since it is fixed only and exclusively through the relations of objects. Once again I would like to emphasize that the distinction between conceptual and relational concepts is quite formal and absolute - it is not leveled by the choice of any, including subjective, points of view. There is hardly such an analyst who can explain what a “sled dog” is without reference to the concept of “riding” (not related to a dog in any way) or who such a “student” is without mentioning the connection with the concept of “higher education institution”.

Further in the text we will talk only about conceptual concepts.
')

Hierarchy of conceptual concepts

In the previous publication, concepts from the chain “animal - dog - like” were given as conceptual. The bug falls under them just after its birth: it is an animal, a dog and a husky. And in order to assert this, one does not need to refer to any other concepts, to explain that, in addition to animals, there are also plants, that there are also cats and shepherd dogs.

The first thing that catches the eye, or rather, comes to mind when analyzing the cited number of conceptual concepts, is that we have an unequivocal hierarchy - coordination according to the degree of generalization-refinement. It can also be noted that the concepts in this hierarchy are not equally conceptual, that is, they do not equally describe what the Beetle is in and of itself. It is clear that if, pointing to the Bug, to ask "who is it?", Then the answer, of course, will be "dog." Perhaps dog lovers will further clarify the breed - “husky”, and no one will remember the generic term “animal”, although everyone knows him. Consequently, in the hierarchy of conceptual concepts, under which an individual falls, one can single out one thing - the basic, most accurately expressing the essence, content, and nature of the individual. This most conceptual of conceptual concepts will be called a “concept”.

The main idea that I would like to focus on when separating concepts from the hierarchy of conceptual concepts is to understand that we initially perceive the world and think it at the basic level of concepts and only in special cases we rise to generalizations and detail. Here's how this idea is presented by J. Lakoff: “It is at this level of experience that we clearly distinguish tigers from elephants, chairs from tables, roses from daffodils, asparagus from broccoli, copper from zinc, etc. One level down - and everything becomes much more complicated . It is much more difficult to distinguish one type of giraffe from another than a giraffe from an elephant. Our ability to gestalt at a basic level is not adapted to make clear distinctions easily at such lower levels ”[Lakoff, 2004]. And here, analyzing the practice of building models of subject areas, we must ask ourselves: do we not recklessly act, starting the construction of an ontological tree from above, with the most common concepts? Do we not adjust the structure of the conceptual level to an artificially invented scheme? Wouldn't it be more appropriate to start building subject domain models with highlighting basic concepts, developing it up and down as needed?

For now, let us leave these questions unanswered (although it is obvious) and turn to the structure of the hierarchy of conceptual concepts. Before us, it (although, of course, without yet naming it so) was analyzed by E. Rosh [Rosch, 1976], who singled out three levels — supreme, basic, and subordinate (in full accordance with our example).

Highest level	Animal	Fruit	Bird
Baseline	DOG	AN APPLE	SPARROW
Subordinate level	husky	Anise	Field sparrow

Perhaps, for a general philosophical view of the relationship between concepts, such a simple distinction is quite sufficient. However, when analyzing typical subject areas, it turns out that in order to build more detailed models, it is necessary to select a larger number — up to seven — of the hierarchy of conceptual concepts. Below is a description of the levels by which conventional names are assigned: category, type, concept, genus, species, variety . Let's start with a basic level.

Concept

• Answers the question: what is it?

The concept is, in essence, what we express in one word when pointing to something. We can draw a concept (if a spatial object falls under it) “in general”, with a silhouette - and those who are looking will not have any problems with understanding what this is about: this is a table, this is a dog, this is a book, this is a car. In our activities, when dealing with individuals, we designate them with the names of concepts: I sat down at the table , the dog barks, I liked the book , I bought a car . In the general case, one individual can fall under only one concept: a table cannot be a chair, a dog a cat, a book a car. However, there are exceptions when one individual falls under two concepts, for example, an amphibious vehicle, a sofa bed, a bayonet, etc. But in these cases the condition is surely met: the individual cannot fall under two concepts at the same time, participate at once as two concepts in one event, for example, an amphibian cannot be both a ship and a car at the same time, and a sofa bed can at any moment be either a sofa or a bed. It is clear that in order to simulate such situations, the language of ontology description must necessarily be temporal, that is, have a linkage of fact-statements to time.

Category

• Answers the question: what is it? What is the essence of this?

A category is the upper threshold of associative perception. A category is understood as a general idea, the essence of concepts, regardless of the way they are used, the scope of activity. The car belongs to the category of "mechanisms", and the dog - to the category of "living organisms." Further generalization of the concepts “mechanism” and “living organism” will lead to a complete loss of content, to leveling of any semantic differences, to an empty indication - “thing”, “object”, “something”. Usually, the concept - and therefore the individual falling under the concept - falls into one category, except when the concept itself has a double (or more) purpose, such as a cane-blade, etc. But, as is the case with concepts, one individual can not be assigned to different categories at the same time: locally in time, in a single event, it can only fall under one category. No categories of categories by analogy with classes of classes (subclasses, superclasses) are implied. Concepts falling under a category are divided into types.

Type of

• Answers the question: what's the point? what do with it? What kind of activity?

The type indicates the special purpose of the concept within the category, the way of functioning, the purpose of its use in specific activities. Types of mechanisms: vehicle, device, lifting means, etc. Types of living organisms: animals, plants, fungi. It should be emphasized that the type is not a subclass of the category (although it is quite convenient to represent it that way), but the type of concept in the category. That is, you should say this: there is the concept of "car", it belongs to the category of "mechanisms" and its type, purpose (in this category) "vehicle".

Rod

• Answers the question: why, how specifically to use? what tasks are solved?

Rod fixes the function, adaptation, specialization of the concept: how is the car used? for what tasks? What is a dog for? - passenger car, hunting dog [note: here the notion “hunting” is precisely conceptual, not relational, since it characterizes not the specialized use of a particular Bug at a certain stage of its life, but the typical use of a group of breeds is any like, even if it never was on the hunt, conceptually would be a hunting dog]. The genus is traditionally called a phrase consisting of the name of the concept and the explanatory definition: red wine, lease agreement, office table.

View

• Answers the question: what?

View - this is a specific implementation of the kind: what is your car? which dog? - “Zhiguli”, like. A species is usually denoted by its own unique name, which does not directly indicate any specificity. The type of product is often indicated by the name of the manufacturer.

Variety

• Answers the question: which one? what implementation?

A variety is a concrete implementation, a special adaptation of the species (“Zhiguli 2101”, Karelian-Finnish Laika). The name is based on the name of the species with a clarifying explanation.

Individual

The individual is a single concrete thing that has to some extent the space-time localization: my car, my dog, last summer. Fixation of the individual, his distinction from other individuals is realized by the statement that he falls under one or another concept. One individual may fall under several concepts, provided that the conceptual functions of the individual do not coincide in time (the car-plane either travels on the road or flies through the air).

Different subjects can specify different concepts for one individual (well, if someone, for example, sees in the microscope only a tool for hammering nails). In order to fix an individual's concept, there is no need to attribute attributes or relationships to it - it is a concept by the very fact of approval by the subject of an individual falling under the concept.

Let us dwell on the last thought in more detail. For example, a certain individual (something in my garage) is a car (fixed as falling under the concept of “car”) not because it corresponds to a certain list of attributes (models) of the concept, but only because I decided so, ignoring the fact that this individual has only one wheel. In the system, it will appear as a concept "car" with one attitude "wheel - part - car". On the one hand, this corresponds to the natural ontology: in order to fix the fact that I have a car in my garage, I do not need to list the values ​​of all its possible predicates. On the other hand, creating a record about a new individual in the knowledge base first of all requires specifying a concept, and not listing the values ​​of its predicates — regardless of whether a person will make this record or input will be implemented by software agents that recognize individuals, say, using neural networks. .

The table shows examples of the conceptual description of the individuals “my car”, “bottle of wine on the table”, “application on my computer”, “my Beetle”.

CATEGORY what does it relate to? What is the essence of this?	mechanism	food product	program	an organism
TYPE OF What's the point? what to do with him?	vehicle	drink	application program	the animals
CONCEPT what is it?	CAR	WINE	TEXT PROCESSOR	DOG
ROD for what? what kind of? how to use? what tasks are solved?	a car	Red wine	office editor	hunting
VIEW which one?	Lada	Finca Traversa	MS Word	husky
VARIETY Which one of? what implementation?	"Lada 2101"	Finca Traversa Merlo	MS Word 2007	Karelian-Finnish Laika
INDIVIDUAL	my car	bottle on the table	computer application	my beetle

Clarifications

It is necessary to make several clarifications and comments about the above conceptual hierarchy:

1. It should be understood that the description through the hierarchy of "category - type - concept - gender - species - species - individual" is not an individual classification. The given hierarchy fixes the levels of perception, the levels of attitude to the individual, sets its conceptual integrity.
2. Since widespread terms used in other conceptual systems are used to designate hierarchy levels, terminological beats are inevitable, questions, claims, they say, the type is about something else, and the category should be understood differently, etc. The answer to this is simple: the conceptual scheme itself is important, and not the private naming of levels, which everyone can change to acceptable in a different terminological grid.
3. There will necessarily be situations in which it will either be impossible for an individual to distinguish all levels of classification (say, there is no variety), or, more likely, the introduction of intermediate levels will be required. Regarding the first case, we can only say: well, no, no. And for the implementation of the second situation will have to include fantasy. The fact that no classification can be exhaustive and universal is a common place.

Literature

1. Baker L.R. The ontological significance of artifacts // Ontologies of artifacts: the interaction of "natural" and "artificial" components of the vital world. Ed. O.E. Stolyarova. M., 2012.
2. Lakoff Dzh. Zhenshchiny, ogon 'i opasnye veshchi. Chto kategorii yazyka govoryat nam o myshlenii / Per. s angl. IB Shatunovskogo. M .: Yazyki slavyanskoj kul'tury, 2004. 792 s.
3. Rosch E. et al. Basic Objects in Natural Categories // Cognitive Psychology. 1976. No. 8. P. 382–436.