Hi Frank,
Principles
The principles of "table relationships" come from original work on the
"Relational Model of Data" defined by the late Dr E F Codd in 1969.
A version of his paper was published in 1970 in the Communications of the
ACM with the title "A relational model of data for large shared data banks"
You can read this paper at
http://www.seas.upenn.edu/~zives/03f/cis550/codd.pdf
Parent and Child
Table structures reflect "facts" .
For example, "each Order contains many Items"
The words that I have capitalised map to tables so this fact means that
there is an Order table and an Item table with a 1:many relationship between
Orders and Items.
Here is another example:
"Person lives at Address"
This means that there will be two tables with a 1:1 relationship between the
tables Person and Address.
Aha, I hear you say, what about rich people who have more than one address?
Good point but in my conceptual model, a person can only be "living at" one
address at a time so the "many addresses" issue is catered for by using a new
fact such as
Person owns Address
For this fact you have a 1:many relationship between Person and Address
Both of these facts can be recorded in just two tables by using appropriate
primary key/foreign key structures.
As you see, there is a lot more to table design that meets the eye. The
first step is to get your facts right. I use a technique called "object-role
modeling (ORM)".
Microsoft has an ORM tool that automatically convert a set of facts in
natural language into a 5th normal form logical table structure. This saves a
lot of time.
You can read more about ORM and Microsoft's ORM tool on my website
http://www.ormfoundation.org
Hope this helps
Ken Evans