Setting up Theories

[StanyBecker]

Be welcome on my article about "setting up theories".
Beneath, you will find the article itself ("SettingUpTheories - article:").
After that, wou can find some informaton about myself ("About me:").
There, you will also find information about the why of these articles.
You can contact me by sending a reply or via "becker-vidal@skynet.be" ...



SettingUpTheories - article:

TITLE
An alternative way to build scientific theories



AUTHORS
Stany Becker



ABSTRACT
In this article, we have a look at the way scientific theories are built.
These theories describe scientifically a `part of reality´.
This `part of reality´ expresses itself by results from observations and measurements.

At this time, the usual way is to start from observations and measurements giving some results.
These results describe the `part of reality´ looked at.
Then we build a theory explaining these results.
Using that theory, we should be able to predict future results of other observations and measurements concerning the `part of reality´.

If the theory doesn’t predict the results (anymore), we have to adjust it (ad–hoc) or build a new one (that does).


The way proposed in this article is to start from a theory (the other way around).
As base for this theory, we use some `founding proposition´ as theoretical foundation.
From that `founding proposition´, we then build the theory.

The results of the observations and measurements should then automatically fit that theory.

If results don’t fit the theory (anymore), we first consider the way the theory was built from the `founding proposition´.
Only as a (very) last resort, we should to take a look at our `founding proposition´.
Maybe it is not fundamental enough or it is badly formed.


Based on the described way of building (using a `founding proposition´), we already built three theories:
1. “Mathematical description of reality”, based on “every description is built using codes” as `founding proposition´.

2. “Process management”, based on “all informatics–routines can be resumed in a (logical) proposition” as `founding proposition´.

3. “Data management”, based on “all data–management is done via assignments” as `founding proposition´.

These articles will be made available soon.
(Both “Data management” and “Process management” have lead to a tested and working application.)

On all three mentioned articles, there is a possibility for further research.



INTRODUCTION
How can we build scientific theories independently from a `part of reality´ (results from observations and measurements) that describes that `part of reality´?


The branch of this article is mathematics (“theory study”: foundation theory).


When starting from a `part of reality´ (results from observations and measurements), we may get into problems when the results from observations and measurements don’t fit the theory anymore.
This can be so because the theory was built ‘ad–hoc’ to explain the results known till then.

When on the other hand, we start from some `founding proposition´, we can build a theory to explain the found results from observations and measurements till then.
Now, if we find results that don’t fit the theory, we may have to adjust the way the theory was built, but not the `founding proposition´.

Only as a (very) last resort, we should to take a look at the `founding proposition´.
Maybe it is not fundamental enough or it is badly formed.
If the `founding proposition´ should have to be changed, we would have to rebuild everything based on it.
So, this better should never happen …


To summarize, in this article, we look at a way to build a theory independently from a `part of reality´ (results from observations and measurements) but that does describe the aimed `part of reality´.



MATERIALS AND METHODS
The question is “How can we build scientific theories independently from a `part of reality´ (results from observations and measurements) that describes that `part of reality´?”.


If we cannot use `the results from observations and measurements´ as a base, what should we use?
Some study and deduction provided the answer to that question.



RESULTS
After this study and deduction, we came to the conclusion that we can use a so–called `founding proposition´ as base of our theories.
A `founding proposition´ is a logical proposition that must be valid in the field of the research.

From that `founding proposition´, the theory is developped, keeping the results from observations and measurements in mind.



TABLES AND GRAPHS
void



DISCUSSION
Having found out that a `founding proposition´ is a good base, we look at some examples: “Mathematical description of reality_article”, “Data management_article” and “Process management_article”.
In this, examples “Data management_article” and “Process management_article” have lead to an application: `AssSet´ and `ReqLan´ respectively.
(These will not be further discussed here …)

Of course, there are other examples.
Point is, you should find a good `founding proposition´ applying this way of working.


Let’s continue with our examples “Mathematical description of reality”, “Data management” and “Process management”:
For “Mathematical description of reality”, we take as `founding proposition´: “all descriptions are built using codes”.
elaboration: `Mathematical description of reality´ (coming soon).

For “Process management”, we take as `founding proposition´: “all informatics–routines can be resumed in a (logical) proposition”.
elaboration: `Process management´ (coming soon).

For “Data management”, we take as `founding proposition´: “all data–management is done via assignments”.
elaboration: `Data management´ (coming soon).

Notice that the `founding proposition´ of “Mathematical description of reality” is more fundamental then the ones for “Data management” and “Process management”.
The two latter can be deduced from the former in the way that we can say that assignments and propositions are built using codes.

Some suggestions for further research:
• How can a `founding proposition´ that is fundamental enough and also well formed be found?
What should be the input and how should we proceed from there to find a `founding proposition´ that is fundamental enough and also well formed?
(Describe scientifically the way the `founding proposition´ was found for “Description of `part of reality´”, “Data management” and “Process management”, making sure that it is also applicable to other `part of reality´s.)

Other fields or further research include
• The elaboration of Mathematical description of reality can lead to further (mathematical) research as “Multiplication of vectors” and based on that: “inverting polynomials” and “solving polynomials equations”.
The result of such a kind of research could then be used for the Mathematical description of reality.

• The elaboration of Process management already has lead to an application `ReqLan´.
Further research can be done on the way how a so-called `founding proposition´ (base for a `ReqLan´–routine) could be built.

• The elaboration of Data management already has lead to an application `AssSet´.
Further research can be done in the field of so–called metadata.
More specific, research could be done in the field of a “value determining routine” as general and unique specification using its `founding proposition´ (from “Process management” – available soon).



As already said, in this article we propose another way to build scientific theories.
We use a so–called `founding proposition´ as base instead of results from observations and measurements.

This is a different point of view!
It is like introducing a new so–called paradigm.
It doesn’t add anything “new”, it only changes the point of view: here from `start from observations´ to `start from a founding proposition´.

You can compare this with what happened describing the orbit of planets across the sky:
At first, those orbits were described using circles.
But soon, it was found that circles don’t allow an accurate description, so observers started adding so–called epicycles, epicycles of epicycles etc.
That went on untill somebody changed the point of view from `only use circles´ to `also allow ellipses´, and suddenly the descriptions became easy.
Notice that no new knowledge was added!

What is proposed here is something similar: we just change the point of view from `start from observations´ to `start from a founding proposition´.
Here, we also don’t add any new knowledge.


To resume, we can say that we also can start from some `founding proposition´ and then built from there a theory that describes the aimed part of `part of reality´ (the results from observations and measurements).
Working that way, we will always have a sound foundation for our theories.



ACKNOWLEDGMENTS
void



REFERENCES (LITERATURE CITED)
void



About me:

My name is Stany Becker.

I have a degree in mathematics and as such, I am interested in study of space and in "theoretical studies of knowledge".

Because I am new in the subject of publishing articles however, my first article is about a methematical construction to uniformly compute sums. ("Sum via Integral")

After that, I will publish my view on the way theories could be set up. ("SettingUpTheories")

That way is then used in the next articles to point out an alternative way to contemplate reality and to manage data and processes.

Remark:
The articles having equations (like "Mathematical description of reality" and "Description of reality") will be postponed until I have found a way to properly publish equations (like in “pdf”-format).
I will test this on the articles in thread “Sum via Integral”.

If you have any ideas about this, please let me know …

 

[StanyBecker]

Be welcome on my article about "setting up theories".

Be welcome on my article about "data management".
Beneath, you will find the article itself ("Data management - article:").
After that, wou can find some informaton about myself ("About me:").
There, you will also find information about the why of these articles.
You can contact me by sending a reply or via "becker-vidal@skynet.be" ...



Data management - article:

TITLE
A uniform and generally usueable way to store and manage data.



AUTHORS
Stany Becker



ABSTRACT
In this article, we propose a uniform and generally usueable way of storing and managing data.
In this article, we give a way to describe reality starting from a `founding proposition´. (see SettingUpTheories, text `Data Management´ - first occurence)
As `founding proposition´, we take “all data-management is done via assignments”.
Based on that ‘founding propositon’, we see we can store the data using the assignments of those data.
We store these assignments in binary trees instead of in coupled tables with rows and colmuns using some so-called datamodel.



INTRODUCTION
How can we get an easy, uniform and generally useable way to store and manage data?

computer sciences – data sources (databases)


Finding an easy, uniform and generally useable way to store and manage data would be quite helpfull.
This is because storing and managing data now-a-days is a complex and expensive business.
This is mainly due to the coupling of tables, coming from a (usually difficult to build) datamodel and all the problems of normalising that model.
The presence of a datamodel also makes it rather hard to store and manage data.
So, not needing a datamodel to store data would make it a lot easier.


So, this article is about exploring a method to get an easy way to store and manage data.



MATERIALS AND METHODS
The question is “How can we get an easy, uniform and generally useable way to store and manage data?”.

In this article, another way is proposed to store and manage data:
We propose the use of binary trees to store assignments instead of storing data in tables with rows and columns to do this.

Because of the ‘founding propositon’ “all data-management is done via assignments”, it makes sense just to store those assignments!

Like that, we have all information available without having to think about datamodels and other (difficult) things (like e.g. building one, normalisation or implementing changes).

Of course, to be able to store assignments, we do also need a datamodel, but that datamodel is hidden from its users (like datamanagers and developpers), as it should be.
Furthermore, our datamodel for storing assignments is a bit special in the sence that every “table” in it has exactly two “columns”: a reference and a value.

Using that kind of model, we can store the assignments in binary trees.
Do actually do this, we can build an application that does just that.



RESULTS
To ease up storing and managing data, we store assignments in binary trees.
To do this, an application ‘AssSet’ (for ‘Assignment Set’) was built.

So, we store and manage data using this application ‘AssSet’.
The term ‘Assignment’ is obvious (it’s what we store).
The term ‘Set’ means there are no doubles stored, so there is no redundant storing.

This all means: by using the application ‘AssSet’, we can store and manage data efficiently without having to construct and maintain a (constraining and difficult) datamodel!



TABLES AND GRAPHS
void



DISCUSSION
As we have seen, using the application ‘AssSet’ simplifies the storing and managing of data.
To store and manage data, we use operarotors put at our disposal by the application ‘AssSet’.
This means we can start the storing and managing data immediately after installing ‘AssSet’.
There is no need te build and maintain a so-called datamodel.
(Of course, there is a datmodel, but it is hidden!)

Now, by just storing assignments as they come, we cannot do very much maintenance.
That’s why we use the described assignment as a base to store the delivered assignment in parts (object, property and value), enabling the adding of so-called metadata.
The metadata that is standardly added is ‘creation moment’, ‘deletion moment’ (set to `later´) and ‘reference’ (providing the internal reference of the assignment).


Metadata:
If we want to know what the meaning of the data is, we should give an explanation on the level of properties of object.
We can do this trying to describe to meaning of the values informally.
We can however also give (formally) the so-called ‘value determining routine’!
Now, the problem is shifted to giving a uniform and generally useable specification of an application (an informatics-routine).
To do that, we could use the so-called `routine-proposition´ (to be helt valid by the specified informatics-routine). (see Process management, text `founding proposition´ - second occurence)

This can be the subject of some further research.



Storing and managing data is at the heart of most of human activity.
If we can simplify that, we can make quite a jump ahead.

The propositon “all data-management is done via assignments” can serve as an example of a `founding proposition´. SettingUpTheories, text `Data Management´ - first occurence (current article)


So, to conclude: In this article, we have pointed out a way to store and manage data an easy, uniform and generally useable way: store assignments in binary trees.



ACKNOWLEDGMENTS
void



REFERENCES (LITERATURE CITED)
void



About me:

My name is Stany Becker.

I have a degree in mathematics and as such, I am interested in study of space and in "theoretical studies of knowledge".

Because I am new in the subject of publishing articles however, my first article is about a methematical construction to uniformly compute sums. ("Sum via Integral")

After that, I will publish my view on the way theories could be set up. ("SettingUpTheories")

That way is then used in the next articles to point out an alternative way to contemplate reality and to manage data and processes.

Remark:
The articles having equations (like "Mathematical description of reality" and "Description of reality") will be postponed until I have found a way to properly publish equations (like in “pdf”-format).
I will test this on the articles in thread “Sum via Integral”.

If you have any ideas about this, please let me know …

 

[StanyBecker]

Be welcome on my article about "setting up theories".
Be welcome on my article about "process management".
Beneath, you will find the article itself ("Process management- article:").
After that, wou can find some informaton about myself ("About me:").
There, you will also find information about the why of these articles.
You can contact me sending a reply or via "becker-vidal@skynet.be" ...



Process management - article:

TITLE
Simplification of reusablility by a uniform and generally useable way to specify, build and store an informatics-routine (code) with a high degree of standardisation.



AUTHORS
Stany Becker



ABSTRACT
In this article, we propose a simple way to enable reusablility:
We look for a uniform and generally useable way to specify, build and store an informatics-routine:
In this article, we give a way to describe reality starting from a `founding proposition´. (see SettingUpTheories, text `Process Management´ - first occurence)
As `founding proposition´, we take “all informatics-routines can be resumed in a (logical) proposition”.
This seems to be more efficient then informally describing what an informatics-routine is supposed to do.

We can also store that routine by storing the `routine-proposition´ as its specification.
From that specification, we can then build the informatics-routine automatically (meaning without any further human interaction after start-up; a tool to do that is built).
In my opinion, this beats interpreting the informal description and then trying to built by hand an informatics-routine accomplishing that informal description.

Because the informatics-routine is built automatically, we can be sure that it does what the specification instructs.
It also creates a high degree of standardisation because every informatics-routine is built using the same techniques.



INTRODUCTION
How can we get a uniform and generally useable way to specify, build and store an informatics-routine (code) for reuse?

The branch of this article is “Computer sciences – programming language theory”.


When we have a uniform and generally useable way of specifying an informatics-routine, we can store it, share it and use it as base to build that routine from its specification.
It also could be helpfull in understanding what an informatics-routine exactly does and what it doesn’t do.

A we said before, we can build (or rebuild) the implementing informatics-routine from its specification automatically, so we can be sure that the informatics-routine follows the specification.
Because the built is automated, it follows fixed procedures.
This means the result (the informatics-routine) will also have a high degree of standardisation.

So, building just specifications is sufficient, because implementations can be generated from them automatically.

Furthermore, we can use that specification to store in a data-holding application (like for instance a database) for further sharing and reuse, since it is uniform.


So, this article is about exploring a method to get a uniform and generally useable way to specify, build and store an informatics-routine (code).
All of this to enable code sharing in order to simplify reuse.



MATERIALS AND METHODS
The question is “How can we get a uniform and generally useable way to specify, build and store an informatics-routine (code) for reuse?”.
At this time, people use some kind of informal explication of what an informatics-routine is supposed to do.

Some primarily research led to the use of a logical proposition to specify an informatics-routine, the `routine-proposition´. [see `acknowledgement´]
The meaning of that proposition is that it must always be held valid by the informatics-routine being specified.

It was also put forward that such a proposition is sufficient to fully specify the informatics-routine.
So, if the `routine-proposition´ fully specifies the informatics-routine, it must be possible to construct an implementation (in some language) from that `routine-proposition´.

As a result, our methods is: construct an implementation for a `routine-proposition´ (a logical proposition that has to be helt valid).



RESULTS
This effort to construct such an implementation finally resulted in an application called `ReqLan´ (for Requirement Language).

This application transforms a set of coupled propositions into their implementing informatics-routines (in Java for the current version).

It does so without any further human interaction or assistance after being launched.
Notice this technique can also be used to maintain these routines.

Another use of `ReqLan´ could be to produce (or reproduce) the manager-level texts, all one has to do to achieve that is to replace the ‘code producing’ parts by ‘text producing’ parts.

This implementing informatics-routine automatically has a high degree of standardisation.

We can also store the specification (a proposition) in a data-holding application (e.g. a database).



TABLES AND GRAPHS
void



DISCUSSION
So, as we have seen, it is possible to specify an informatics-routine uniformly and generally by using its `routine-proposition´!
We can store informatics-routines uniformly by storing its `routine-proposition´.

From the `routine-proposition´, one can build its own machine-specific implementation using `ReqLan´ if needed.
This implementation (a Java-routine) is built without any further human interaction or assistance after launch (as already mentioned).

It has a high degree of standardisation.

After being built, you can link this implementing routine to the `routine-proposition´ in a data-holding application.

This all can be done automatically if needed.
This means that building just specifications and elaborate them should be enough to build the application.

This is a lot simpler then developping an application and what’s more, one only needs to adjust the specifications and regenerate (only one human adjustment).


If we all use the this way of specifying, building and storing informatics-routines, the sharing and reuse of informatics-routines could be improved a lot.


So, we can say that once we have the `routine-proposition´, it is fairly easy, fast and cheap to get an implementation (using `ReqLan´).


Remains the (important) question how to get to such a`routine-proposition´!

This can be the subject of further research.

This research should include the managers, for they are the ones deciding what has to happen (they decide what should be in a `routine-proposition´).

Currently, managers are producing (or having produced) informal texts explaining what has te happen.
Imagine the progress made if the implementation could be done automatically from a formal `routine-propositions´ instead of having to follow the current difficult, slow and expensive way to implement.

We could for instance develop templates to be filled out (by managers or their respresentatives) and then have code behind it to generate the formal `routine-propositions´.


We can see this result as an example of “reverse engineering”, getting an implemenation from its specification.

The propositon “all informatics-routines can be resumed in a (logical) proposition” can serve as an example of a `founding proposition´. (see SettingUpTheories, text `Process Management´ - first occurence)


So, to conclude: In this article, we have pointed out a way to get to a uniform specification for an informatics-routine and a uniform and generally useable of building it, including a way of storing that informatics-routine for possible reuse.
All these constructions are done without any human interaction after start-up.

Notice that the manager-level texts could also be produced using another version of `ReqLan´ (with ‘texts producing’ parts instead of ‘code producing’ parts).



ACKNOWLEDGMENTS
With thanks to prof. Martin Ward (from Liverpool), who put me on the right track.



REFERENCES (LITERATURE CITED)
void


About me:

My name is Stany Becker.

I have a degree in mathematics and as such, I am interested in study of space and in "theoretical studies of knowledge".

Because I am new in the subject of publishing articles however, my first article is about a methematical construction to uniformly compute sums. ("Sum via Integral")

After that, I will publish my view on the way theories could be set up. ("SettingUpTheories")

That way is then used in the next articles to point out an alternative way to contemplate reality and to manage data and processes.

Remark:
The articles having equations (like "Mathematical description of reality" and "Description of reality") will be postponed until I have found a way to properly publish equations (like in “pdf”-format).
I will test this on the articles in thread “Sum via Integral”.

If you have any ideas about this, please let me know …

 

[origin]

The way proposed in this article is to start from a theory (the other way around)

That doesn't make any sense. How do you start with the result?

 

[StanyBecker]

That doesn't make any sense. How do you start with the result?

The whole point is to build a theory from some "founding proposition" as base!
So, I don't see a theory as a result of observations and measurements.
You could rather say that I see a theory as a result of the elaboration of some "founding proposition" (as explained in the article).

The strength of this way of working lies in the fact that we always have a consistent theory.

If some observations and measurements don'f fit the theory, we should look at the way the theory was constructed, as explained in the article.

Maybe you can read the article with the above in mind ...


As stated in the article, you can also read the examples about "Process Management" and "Data Management" as an example.
They are available now.
The example about "Mathematical description of reality" isn't yet, because I still need to find a way to publish "pdf"-equations coming from LaTeX readably.


Can you let me know what you think about this way of working? (starting from some "founding proposition" instead of starting from observations and measurements)

Is the article itself clear enough written or do I have editing-issues?


Thanks in advance for letting me know those things ...

 

[origin]

The whole point is to build a theory from some "founding proposition" as base!
So, I don't see a theory as a result of observations and measurements.

So make random guesses and then see if they match observation?

Can you let me know what you think about this way of working?

I think it is completely illogical.

 

[StanyBecker]

So make random guesses and then see if they match observation?
I think it is completely illogical.

One shouldn't make "random guesses" when establishing a "foundation proposition", but one should find a proposition that is really basic enough.
This is the hard part of the proposed way of working!


As an example, I'll explain it with the founding proposition of "Data management": "all data–management is done via assignments".
This is so because one can only initiate or change the value of a field using an assignment.
There doesn't seem to be anything more basic than that ...
{ This is the hard part I was talking about }

So, from that point, we develop a theory to find a uniform and generally usueable way to perform data–management.

You can find out the details in the article about "data management".


Another example is the founding proposition of "process management": "all informatics-routines can be resumed in a (logical) proposition".
This is a good "foundation proposition" because it can be proven that all routines can be resumed in such a (logical) proposition.
Because that (logical) proposition is the most basic you can get, it also can serve as "foundation proposition".
{ Again, this is the hard part }

The way the theory to simplify and uniformize routine-creation is built from there is explained in the article "process management".


I hope this helps you to get an insight of what I want to explain ...

Any remarks are welcome ...

 

[origin]

I hope this helps you to get an insight of what I want to explain ...

No doesn't help at all. Data management has nothing to do with scientific theories.

 

[StanyBecker]

No doesn't help at all. Data management has nothing to do with scientific theories.

Data management is mainly about storing and retrieving data to get problems solved.
For Instance, what do you all need to know to get your employees paid and how do you store thay information in an efficient, maintainable and secure way? (Database theory)

At this time, people try to understand the object involved and what information is needed from them.
This leads to a so-called Datamodel.

That Datamodel usually has redundant information (e.g. he name of an employee is registered on his function description, but also on the payroll and propably many other places).

To solve that issue, the datamodel needs to be transformed to one (ideally) without redundancy.

These are difficult, time-consuming and expensive steps and more-over, they limit severely the easy of use and flexibility.

Now, my idea is not to try model the data, but to simply store the assignments!

This works because every change involves an assignment somewhere.

It is that idea that I wanted to publish.



Also, data management is about "knowing what you know", in other words: what is the meaning of the data? (the so-called metadata)

On this point, I also would like to suggest an idea:

Up until now (for as far that I know), metadata are given by specifying by the use of the data.
This is not a unique specification (people can use the same data for different purposes).
It also makes comparing the metadata quite hard. (When can you use the same data?)

My idea to solve that problem is: specify the meaning of data by the (unique) creation-routine instead of by its use!

Now, routines can always be uniquely specified by their so-called validity-proposition, this is a (logical) proposition that must be kept valid.
So you don't need to specify a routine by its scheme, code or something else, giving the validity-propsition is sufficient.

More on that in the thread "Process Management" ...



What do you think about my ideas? ("storing assignment instead of making datamodels" and "specify data by their creation-routine instead of by their use")

Any remarks are welcome ...

 

[origin]

Ah, the problem is that I mistakenly thought this was about setting up theories! You want to discuss data management. In the future I would suggest that the title and the OP should have some relation to each other.

I will bow out now since discussing data management is not interesting to me.

Have a happy New Year.

 

[StanyBecker]

it IS about setting up theories!

I used "data management" (and "process management") merely as an example to show the use of the `founding proposition´ and what you can do with it.

I thought that the articles about "data management" and "process management" could serve well as an example to illustrate how to build a theory starting from a `founding proposition´.

Of couse, it also involves some knowledge of the subjects at hand ("data management" and "process management"), but they are not the main issue here.

Soon, I will deliver a thirth example: "description of reality", also as an example of how theories could be set up starting from a `founding proposition´.


Maybe you have some examples of a theory for which I could come up with some `founding proposition´ and (re)build that theory from there ...


I'm sorry about the misunderstanding, but you should know I have had a diagnose of autism.

As such, I may have some problems with assessing the way other people might interprete my interventions, like the fact that you should think this is about "Data Management".


Anyway, I am looking forward to your response, maybe with some theories that I could rebuild from a `founding proposition´ (with your help of course) ...