new research into semantic categories

ROBIN HAMILTON robin.hamilton2 at BTINTERNET.COM
Fri Feb 12 01:24:49 UTC 2010


Hi, Joel,

Our posts seem to be overlapping (I sent my second before reading this) but to cut to the chase.

> Yes, by "Boolean" variable one can understand
> "binary".  If that's a
> sufficient clarification, go to (2); Else go to (1).

K, I'll accept that, but I'd still have been happier with "binary" initially.  The problem with "Boolean" in this general context (and I admit that you did specify <"Boolean" variable> when you wrote initially, so my response was based on s sloppy initial reading of what you said) is that "Boolean" has *the possibility* of multiple confusions of reference not quite so close to the surface when the term "binary" is used.

Actually, with that cleared up, I'm not sure how far we still disagree -- and if we do still disagree, at least I'll have a better idea of what I'm disagreeing with rather than setting up a straw horse to beat to death.

     <SNIP>

> The Wikipedia article "Boolean data type" says it is a
> "data type
> having one of two values: true or false, intended to
> represent the
> truth values of logic and Boolean algebra."  A Boolean
> data type is
> thus one kind of enumerated data type, the simplest that
> can contain
> a variable (having only two possible values; if it had only
> one value
> it would be a "constant").  No need to progress on to
> Boolean
> algebra, set theory, the integers, etc. a la Russell.

Agreed, but your shift from Boolean variables to Relational Databases confirmed me in the idea, as it turns out wrongly, that you were eliding the fundamentals of a system (thus both computers and Boolean logic depend on a two-value foundation, in the one case that statements are either true OR false, in the other that a number is either 1 OR 0)with what is produced *by the system.

> The two "dimensions" from the "new research" into how
> humans classify
> non-human objects that are called "Can I eat it?" and "Can
> it give me
> shelter?" are yes/no questions, and therefore they are
> "properties"
> that have the (enumerated) data type "Boolean".

That's a different issue. I don't go along with your description above of the research done (as reported on this list -- I haven't checked back beyond that).

Best,

Robin

------------------------------------------------------------
The American Dialect Society - http://www.americandialect.org



More information about the Ads-l mailing list