Re: teaching relational basics to people, questions

From: Mr. Scott <do_not_reply_at_noone.com>
Date: Fri, 11 Dec 2009 23:15:17 -0500
Message-ID: <XfSdnX6v35FLhb7WnZ2dnUVZ_rednZ2d_at_giganews.com>


"Kevin Kirkpatrick" <kvnkrkptrck_at_gmail.com> wrote in message news:4be6ab47-e654-4925-b712-8e72ccb84394_at_g12g2000yqa.googlegroups.com...
> On Dec 10, 8:12 pm, "Mr. Scott" <do_not_re..._at_noone.com> wrote:

<snip>
> Hello Mr. Scott,

> If I'm understanding your explanation properly, in a database with
> only table:

> BIG_US_CITIES {CITY_NAME, STATE_CODE}

> (that is, lacking table STATES {STATE_CODE}) one must logically infer,
> from the CWA, that states with no big cities, e.g. 'RI', do not exist?

The inference is valid, but it is not due to the CWA. If the only place in the database for assertions like 'there is a state called <STATE_CODE>,' is in BIG_US_CITIES, which houses facts like 'There is a big city named <CITY_NAME> in a state called <STATE_CODE>,' then one can logically conclude that there aren't any states with no big cities. Even under the OWA, there can't be any states with no big cities. Under the CWA, all and only states with big cities are represented in the database, and under the OWA, only but not necessarily all states with big cities are represented in the database. In both cases, only states with big cities can be represented in the database. There isn't anywhere to record the assertion that there is a state with no big cities, so there can't be any. The table STATES {STATE_CODE} provides a place to record assertions like 'there is a state called <STATE_CODE>' independent of whether there is also a big city. Received on Sat Dec 12 2009 - 05:15:17 CET

Original text of this message