Back in lesson one and two,
one of the tables in the zoo database we

talked about, was the diet table which
describes which foods each species eats.

We looked at three ways we could store
multiple foods, for a single species.

As multiple food columns and

one row per species, as a single
column with foods separated by commas.

And has separate rows, for
each food comma species pair and

I mentioned that this form, has some
big advantages over the other two.

It works better with drawings,
it works better in comparisons and

you don't have to know in advance how
many foods there are, and we also looked

at these two tables of donors to
the zoo and their phone numbers.

In order to avoid duplication,
we split the information that

was in one object or it's record,
into two separate tables.

In database jargon,
these are called normalized forms and

the others are denormalized forms.

Normalization is the central
idea in database design.

Normalizing involves making
the relationships among the tables in

your database, match the relationships
that are really there amongst

the various pieces of data.

There's a bunch of database
theory behind this idea.

I put a link in the instructor notes to
a paper with lots of details about it.

Up ahead, I'll just tell you
the most relevant highlights.