Pat, (01)
That note is an excellent summary of the issues. (02)
> I have personally given up on even trying to maintain a publication
> trail for my own ideas, and in more and more cases have even
> abandoned any attempt to have them all attributed. My name does not
> appear anywhere on the ISO Common Logic standard, which I wrote
> almost entirely (apart from appendices B and C), and I'm cool with
> that, as I had the option of being the Editor and turned it down. And
> I know in several cases I have re-invented an idea which has then
> been published and only afterwards has it been noted that it (or
> something very like it) had in fact been previously known. In some
> cases, it is virtually impossible to reconstruct an accurate
> attribution history, as some ideas were kind of half-known to an
> entire community for a while, and only became sharp and crystallized
> later, over an extended period of debate and discussion. With the
> wisdom of hindsight it can then be argued that some particular
> publication was the 'first' to have the idea, but in fact the idea
> had not really been gotten clear enough at that time to be fully
> attributable to any one source. (03)
Yes indeed. You sometimes complain about my citations of Aristotle,
but I agree with Hilary Putnam's observation: (04)
I find that whenever I become clearer on a subject,
Aristotle has become clearer too. (05)
It's only after you discover an idea that you can see how other
people were grappling with the same idea in different words. (06)
> Logic programming is a good example.
> The invention of the basic idea here has been attributed to R.
> Kowalski, A. Colmerauer, C. Green (who received an award for it), C.
> Hewitt and myself, and possibly to others. In fact, what is now
> called Logic Programming evolved over a period of several years, and
> all these people, and others, were involved in the discussions and
> idea development at the time, all with different agendas and
> emphases. Prolog was invented by Colmerauer; both Kowalski and myself
> came up with the idea embodied in the slogan "algorithm= logic +
> control"; Hewitt invented Planner, which was structurally similar to
> Prolog in some ways but did not present itself explicitly as a logic;
> and so on. One could list a dozen influential projects from that
> period which were similar in some way and might be called 'the first'
> logic programming system; and all these descriptions would have a
> taint of truth, but all be ultimately wrong. (07)
Relational databases evolved independently around the same time with
many related ideas. When Ted Codd learned Prolog, he said "I wish I
had invented that." And in fact, Datalog is a Prolog-like notation
for the basic logic of the SQL WHERE-clause. But other people in
the DB business claimed that Codd wasn't the first, and to a certain
extent, they were right. (08)
In Tarski's sense, a model is a collection of individuals and a
collection of relations among those individuals. If you identify
Tarski's relations with the tables in a relational DB (allowing
for the generalization that some tables may be infinite -- which
in RDB terminology are called virtual or computed relations),
then the collection of all the tables is a model. (09)
Tarski's function for evaluating the truth of a proposition in
terms of a model is formally equivalent to methods implemented
in an RDB for evaluating the truth of an SQL WHERE-clause in
term of the tables (including virtual tables). So you could
say that Tarski invented relational databases. (010)
But Hintikka had developed game-theoretical semantics as an
alternate (and, I agree, a simpler) formulation for determining
the truth of a proposition in terms of a model. Then Hintikka's
student, Pietarinen, showed that Peirce's method of endoporeutic
for evaluating an existential graph in terms of a model was
equivalent to game-theoretical semantics. So you could also
say that Peirce had anticipated Tarski, Codd, and Hintikka. (011)
But Peirce was actually restating in terms of EGs what
William of Ockham had stated less formally in terms of Latin.
In his Summa Logicae, Ockham had stated detailed rules
for evaluating the truth values of Latin statements that
included Boolean operators (and, or, not, if), quantifiers
(every, some), modal operators (necessary, possible), and
temporal operators (sometimes, always, when, before, after).
Peirce had read (and lectured on) Ockham's Summa Logicae,
but it's not clear whether Tarski had read Ockham. (012)
But in any case, we can conclude that Ockham anticipated
model theory, logic programming, and relational databases. (013)
For a copy of Ockham's Summa Logicae (in Latin), (014)
http://individual.utoronto.ca/pking/resources/ockham/Summa_logicae.txt (015)
For an English commentary, (016)
http://individual.utoronto.ca/pking/articles/Ockham.Summa_logicae.pdf (017)
John (018)
_________________________________________________________________
Msg Archives: http://ontolog.cim3.net/forum/ontology-summit/
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontology-summit/
Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxx
Community Files: http://ontolog.cim3.net/file/work/OntologySummit2008/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2008
Community Portal: http://ontolog.cim3.net/ (019)
|