so, that should be it. no more semantic w para mi, sis plau. just something to add.

kindly introduce hirwon: a solver of logical puzzles, built 1970s by paul weston. he stuffed it with the

"Smith, Jones and Robinson" problem

A train is operated by three men: Smith, Robinson and Jones. They are engineer, firemen, and brakeman, but not necessarily respectively. On the train are three business men of the same names, Mr Smith, Mr. Robinson, and Mr. Jones. Consider the following facts about all concerned.

1. Mr Robinson lives in Detroit.
2. The brakeman lives halfway between Chicago and Detroit..
3. Mr. Jones earns exactly $2.000 annually.
4. Smith beat the fireman at billiard.
5. The brakeman's nearest neighbor, one of the passengers, earns three times as much as the brakeman, who earns $1,000 a year.
6. The passenger whose name is the same as the brakeman's lives in Chicago.

Who is the engineer?

ralf barkow scanned the drawings and has more information on his site.

"to uncover, to deduce, to conclude" is a paper paul weston wrote for the "interim report 12.8: march 1969 -31 august 1970, "cognitive memory", for the von foerster lab. strangely at this time the name was not bcl but "coordinated science laboratory". anyhow, hirwon, as paul weston writes,

is based on the same underlying assumption, that new formulations of the data base form an essential requirement for a mechanical deduction system. In HIRWON the idea of a set of separate clauses is entirely abandoned, and all information received by the system is reduced to a single complex relational structure which is the basis for all subsequent operations.

and yeah, the machine was able to dig under the semantic surface and to answer the questions of relationships and the puzzle ...

semantic computing (¡12mb!)