Re: The term "theory" as in "database theory"
Date: Sun, 28 Jan 2007 15:17:33 +0100
Message-ID: <45bcafad$0$320$e4fe514c_at_news.xs4all.nl>
dawn wrote:
> mAsterdam wrote:
<snip OP and cdt charter>
>> Like Bernard Peek I think building on a definition >> of "theory" in "comp.databases.theory" is working in the wrong >> direction. It is whatever the group accepts today. >> Tomorrow it may be different.
>
> That's fine, but I had some specific reasons...
Which?
<snip riddle>
>> Having said that, some comments: >> >> >From dictionary.com >> >>> "1. a coherent group of general propositions used as principles of >>> explanation for a class of phenomena: Einstein's theory of relativity. >> As soon as people started to write up inventories, IOU's, the >> problem of keeping the data up-to-date emerged. >> >> (data in a broad sense, not limited to computers - I say >> this because some even deny the existence of data outside >> computers by making computers a defining part of "data".) >> >> So, they started to see the data itself as a valuable resource. >> They started organizing the data beyond just archiving all documents. >> >> How did they do that? >> What principles did/do they use? >> How did those change? >> What triggered these changes? >> >> These could be topics for database theory in sense number 1.
>
> So, you are thinking that database theory does look at some phenomena
> and attempt to explain it with a mathematical theory?
I am not sure what you are asking here.
> This is where I
> think those who are pushing the Occam's razor comment must come out.
Other things being equal, I prefer simple constructions to complex ones. Do you?
Aside: for nuance, think of cryptology.
<snip remark towards Keith>
> With 3, 4, and 5 below, we are talking about creating the rules, the
> theory for purposes of excellence in software development (some might
> have other goals), so Occam's razor would not fit, even if simplicity
> is still a valuable attribute for a theory (in my opinion, simplicity
> for the user/developer is one of the important features, even if
> sacrificing some simplicity in the software written for their use.).
To check whether I understand you:
You are challenging some peoples uses of the term 'Occam's razor', because it should IYO exclusively apply to explanations of phenomena, not to any other explanations. You do not challenge their quest for simplicity itself, which is what they were expressing with the term.
Is that correct?
>> In Kenneth Downs' words, /database/ as a phenomenon >> whose principles need to be discovered
>
> In that case, if we are working to uncover the secrets of "the
> database" (in the sky) and model it, then Occam's razor would be
> applicable, but one would need a way to measure how closely the
> mathematical model aligns with this database phenomenon, as one can
> investigate how closely the theory of relativity aligns with reality.
I have no idea how to make sense out of this. Slow down, please.
>> He wrote: > A database is not a phenomenon whose principles need to be >> > discovered, as for instance the phenomenon of gravity >> > is the subject of Einstein's theory of relativity. >> > A database is a man-made thing whose principles are >> > crafted by the human mind to accomplish human goals. >> >> ... and so /not/ a phenomenon? I disagree. >> >> However, most participants in cdt limit the environment >> to computers.
>
> Agreed.
With the observation or with the limit?
I don't (limit the environment of data to computers), so most of the time I use terms like 'data in a broad sense', or 'meaningful' here to prevent misunderstanding.
>> Note that it (database as a phenomenon) would go somewhat >> against the cdt charter because it is about implemented databases.
>
> Or some platonic form, perhaps?
>
>>> 2. a proposed explanation whose status is still conjectural, in >>> contrast to well-established propositions that are regarded as >>> reporting matters of actual fact.
> Nice. http://en.wikipedia.org/wiki/Imre_Lakatos
>>> 3. Mathematics. a body of principles, theorems, or the like, belonging >>> to one subject: number theory. >>> 4. the branch of a science or art that deals with its principles or >>> methods, as distinguished from its practice: music theory. >> I think this is close to what the group accepts today.
>
> Yes, this one makes sense.
>
>>> 5. a particular conception or view of something to be done or of the >>> method of doing it; a system of rules or principles.RM, a vociferous group only likes to discuss this one. >> MV, you and some others like to discuss this one.
>
> Yes, this also makes sense and aligns with the charter's interest in
> data modeling, for example.
>
>>> 6. contemplation or speculation.This is just a rude #2. >>> 7. guess or conjecture."6a.
>
> agreed.
>
>> Proposal: >> Database theory: >> The part of information theory that deals with >> dynamic collections of facts as valuable resources.
>
> I don't disagree with it, but find it less useful than combining the
> meaning of "database" with 3, 4, & 5 above.
? 3 into the deal? Math is a tool everywhere, it has no business in defining a discipline outside math.
> Perhaps the charter and
> at least your def of database theory, perhaps some of these above
> related to theory, would be helpful in the glossary? Your call.
I don't know (yet). "Theory" is a little vague here (a reason to include it in the glossary), but I haven't seen real misunderstandings (maybe I missed them). To have me include it at least a few others should voice opinions.
> Thanks and sorry that when I clicked your name, I didn't verify that
> your posting was the one that came up. Dag. --dawn
Geen probleem.
Tot later :-)
Received on Sun Jan 28 2007 - 15:17:33 CET