At the session of the second synthesis Terry Bollinger mentioned that Euclid's Elements begins with the definition of a point. A definition that is delightful: "A point is that which has no parts."
I would like to emphasize that 2300 years later, in Hilbert's axiomatic theory for the same entities in the same Euclidean space, the situation with the definition for a point has changed significantly, namely.
For Hilbert, a point is a primary (aka primitive) concept, along with two others: an infinite straight line and a plane.
For these three (there are no other primary) types of objects, 6 primary relations are introduced - all but one are binary, betweenness is ternary.
So what is a point for Hilbert?
This is any class of objects if it is possible to provide two more classes of objects for it, specify the geometry relations on all three, and check that the axioms of geometry are satisfied. The class of objects for which the axioms are satisfied as for a point can be considered points. I think Hilbert once joked "even beer mugs". This example illustrates how we apply our axiomatic theories: specify real or ideal objects for primary classes (sorts), specify primary relations on them, and check that the axioms are satisfied. If the axioms are satisfied (for example, theories of undirected graphs), go ahead, the theory is applicable in all its glory.
https://www.linkedin.com/pulse/concept-point-from-euclid-hilbert-alex-shkotin-fsg2e/
Alex
--
All contributions to this forum are covered by an open-source license.
For information about the wiki, the license, and how to subscribe or
unsubscribe to the forum, see http://ontologforum.org/info
---
You received this message because you are subscribed to the Google Groups "ontolog-forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ontolog-foru...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ontolog-forum/4b07eff3-0bb5-48b3-b8a9-5d4bc18a7f07n%40googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ontolog-forum/CAAN3-5epQgi%3DXAj4%3DDaLnExS7jtR_VF%3DZEC5KkQ8waLj6xR3pA%40mail.gmail.com.
John,
IMHO "scientists and engineers" adopts not just definitions but one or another theory to apply it to solve one or another practical problem. Idea of "set of definitions" is much better, but theory is not only a set of agreed definitions (it's a bad idea to take definitions from different theories). Theory has a set of derivation rules. And this set is not just modus ponens, there are many others. In FOL we have approximately six rules we use to construct formal proof. In practice, the number of rules is much larger (see).
There is no task "to choose one definition or set of definitions that can or should be adopted as an international standard for all purposes."
It may be a task (as I mentioned in our last session) to formalize Hilbert's theory of Euclid's geometry.
"open-ended variety of different definitions" is a proclamation of a kind of bad infinity. What we have is approximately 10 different axiomatic theories for Euclid's geometry. And this is true that having a problem in mind we choose a theory which best suits our goal and how to solve this problem. Euler invented basic undirected graphs theory trying to solve the problem of Konigsberg bridges.
But today's point is not inventing a new theory. We have a lot. But systematize and formalize existing. Have a look at OBO Foundry ontologies to see how many formulas we have there. What are the theories behind?
Recommendation: Keep in mind which one theoretical knowledge you use to solve a problem. Remember that formal definitions are not for you, it's for computer algorithms to make proper knowledge processing to help you to solve a problem.
Definition is a unit of theory, but theory is more than a system of agreed definitions.
Alex
--
All contributions to this forum are covered by an open-source license.
For information about the wiki, the license, and how to subscribe or
unsubscribe to the forum, see http://ontologforum.org/info
---
You received this message because you are subscribed to the Google Groups "ontolog-forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ontolog-foru...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ontolog-forum/4befeb14849d4fc6a4be2b80d0c556cf%402377029e14dd4cfd87fb1ff2f6d224a9.
John,
It's just a proposal to discuss what is a point in different theories. I took Hilbert's theory as it is well known. It is definitely interesting to take 9 others into account. But this looks like a small research.
If you have a reference to a theory where a point is a locus, please give a reference.
And I think Hilbert does not use the relationship "part_of" in his theory at all. This is a very interesting topic that "part_of" is always an abstraction from real relationships. In Hilbert's theory we have point lies on line, and line lies on plane.
Do we need a definition for a part_of relationship?
General principle: Abstraction is a fundamental mental action. Even what we see with good eyes is an abstraction from real matter. We use abstraction further when we choose a theoretical knowledge to use. We use theoretical knowledge to describe and then solve a problem. Be careful with abstraction. Remember that there may be a child in the water.
Alex
--
All contributions to this forum are covered by an open-source license.
For information about the wiki, the license, and how to subscribe or
unsubscribe to the forum, see http://ontologforum.org/info
---
You received this message because you are subscribed to the Google Groups "ontolog-forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ontolog-foru...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ontolog-forum/cd46dda913564912a2b8e79e065966d7%400d895feabeaf41b48e6ba2decf9fa099.
John,
Topics of "general framework for ontology" and "universal ontology" are interesting and classical. I am not sure I am ready to say something new.
Let me continue a little about the research I have mentioned before.
In the Russian version of the article Hilbert's axioms - Wikipedia there is a section about other axiomatic systems by the way for the same kind of space - euclidean and the same kind of objects and relationships in it.
If we truslate this section by Google-trans we get this [1].
In section "More modern axiomatics:" we have only 5 axiomatic theories to look at point appearance. This is a small research, I am sure.
As we see from "Weyl axiomatics - operates with undefined concepts of a point and a free vector." Here point is a primary (aka primitive) concept.
Axiomatizing theory we find out primary (aka primitive, undefined) concepts and relationships and axiomatize them. But we define all other concepts and relationships.
Applying theory to an area of interest we must show there our primary concepts and relationships, like this: the stars and planets on the sky are points. Beams of lights are moving along the straight lines, and orbits of planets move over the planes. Well, we can apply Hilbert's axiomatic theory to describe and even solve some astronomical problems. At least their geometrical part.
If we speak more precisely, we must say that we have an axiom: every physical object at any moment of time occupies some place in physical space. Thus, we have a mapping of a physical object into a geometric figure located somewhere in space. This is a mapping of "location". Perhaps you call it a locus.
Alex
[1]
Other axiom systems[ edit | edit code ]
Creators of pre-Hilbert systems:
Related to Gilbert's:
W. F. Kagan (1902)
O. Veblen (1904)
More modern axiomatics:
Birkhoff's axiomatics - contains the " ruler axiom " and the " protractor axiom ". Its variants are used in most American school textbooks, and Pogorelov's axiomatics is close to it .
Weyl axiomatics - operates with undefined concepts of a point and a free vector . A line and a plane are defined as sets of points.
To view this discussion visit https://groups.google.com/d/msgid/ontolog-forum/33e546f30b924c2eaddab6fac649e943%404bef0007639145808a78606cc37cef2d.
To view this discussion visit https://groups.google.com/d/msgid/ontolog-forum/CAFxxROSqd3LnqtGR9G6_xdkb3drfeXptmPVY1SWJGL43QHfYKQ%40mail.gmail.com.
To view this discussion visit https://groups.google.com/d/msgid/ontolog-forum/CAAN3-5d1kxu9PuY0jH6c80hD6mzHfC3vmcvqV57yTyfzZt_Csg%40mail.gmail.com.
John,
John
Exactly! While all theories are mostly beautiful, we need to know how to apply them. And this applicability reasoning is crucial. To open up a little Box's point. It may be an issue that to apply this particular theory to this particular problem is the wrong idea, and in other cases, is not useful. As Landau wrote, we do not apply quantum mechanics where Newtonian is applicable.
One task is to formalize theory itself like the theory of undirected graphs [1]. Another task to formalize one or another problem solving, using this theory [2].
But the task #1 is to recognize somehow (by abstraction, reduction, generalization etc.) that this particular theory is applicable to solve this particular practical task properly, taking into account "error bounds and estimates" [3].
Theory itself consists sometimes of a large number of subtheories for different kinds of objects and processes, i.e. situations.
In this case the structure of a theory itself is not trivial.
Alex
[1] (PDF) Theory framework - knowledge hub message #1
[3] https://en.wikipedia.org/wiki/Seven_Bridges_of_K%C3%B6nigsberg--
All contributions to this forum are covered by an open-source license.
For information about the wiki, the license, and how to subscribe or
unsubscribe to the forum, see http://ontologforum.org/info
---
You received this message because you are subscribed to the Google Groups "ontolog-forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ontolog-foru...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ontolog-forum/4af379caab444dc397b1e7dac12da955%40314f781b36424228afe29a43e724d3b3.