Re: Ambiguous names. was: Re: URL +1, LSID -1

Matthias Samwald wrote:
>>  The evidence for what I point out is found everywhere: "P12345 is
>>  expressed in some tissues"... according to Alan's points, this
>>  would be a wrong statement. 
> When the Semantic Web should really find widespread adoption, they would be saying something like "C12345 is expressed in some tissues", where C12345 is the identifier of a class of protein molecules (which might be described in P12345.html). Not much would change for the scientists -- it would rather seem that using identifiers to identify the proteins themselves instead of the database records is what they implicitly want.

Wait.  You are switching between levels of quotation, which for most 
purposes is fine and easily comprehensible by the human, but for the 
purpose of discussing just what is being discussed, care should be taken.

If X is an identifier (i.e., if what the symbol 'X' is used to represent 
in *this* sentence is an identifier), then to say that X is expressed is 
to say that an identifier (namely, X, the one referred to with 'X') is 

Quite a nonsense, if what you really think of is gene expression.

If X is an identifier (as above), then the sentence 'X is expressed'' 
means that the identifier is expressed, provided that inside and outside 
the quotation the symbol 'X' has the same referent.  If it does not, 
then the two levels are independent wrt. their use of 'X'.

If 'X' is used in the sentence 'X is expressed' as an identifier of a 
class of protein molecules, then the sentence is effectively used to say 
that the class of protein molecules is expressed (whatever this may mean 
to you).

Similar patterns appear again and again in mails discussing the 
ambiguity of urls and such -- but the problem stems, among others, from 
this sort of carelessness.

Yes, books on logic do say "the symbol v is a propositional connective" 
and the like, which taken literally would mean that the symbol referred 
to by 'v' is a connective, but this is not what they intend to say, and 
such conventions shoul be made clear beforehand.  Mendelson [1], for 
example, takes care to explain that (p. 13).


[1] Introduction to Mathematical Logic, Fourth Edition (Hardcover)
by Elliott Mendelson, Chapman & Hall/CRC; 4 edition (June 1, 1997).

Received on Monday, 16 July 2007 19:32:01 UTC