John,
Thanks very much for the links to Fred Thompson's writings. They are, of course, very impressive and thought-provoking.
Whether Tala is a Controlled Natural Language depends on one's definition of CNL. I use the following definition:
CNL = A constrained (hence, "controlled") subset of NL syntax to support representation of a constrained semantic domain.
Under that definition, Tala is not a CNL, because its goal (and to some extent, current design) is to support unconstrained NL syntax and unconstrained semantics.
Your slides give the following definition:
CNL = A subset of a natural language that has a well-defined mapping to and from a computable form.
Under this definition, Tala could be considered a CNL, since it is a language providing "computable forms" that have a syntax corresponding to unconstrained NL syntax.
However, many have viewed the problem as providing a mapping from NL syntax into some other language, or different set of computable forms, e.g. predicate calculus, conceptual graphs, frames, etc.
In contrast, the TalaMind approach, proposed in my thesis, is to support reasoning directly with computable forms that correspond to natural language expressions; to treat natural language syntax as the best way of representing natural language semantics.
(Incidentally, this is consistent with Thompson's (1972) statement "Language is the embodiment of conceptual structure." ... though perhaps he intended something different by the statement.)
Regarding your questions:
> Why haven't these great ideas become the foundation
> for the Semantic Web? Should they be considered? If not,
> why not? If so, how?
Perhaps one answer to the first question is that Berners-Lee (1998) specifically disclaimed "A Semantic Web is not Artificial Intelligence". Perhaps since people creating Semantic Web technologies were not focused on trying to achieve human-level AI, they naturally tended not to try to fully represent natural language semantics.
Regarding how to develop these ideas going forward, my recommendation is that people work on developing systems according to the TalaMind approach.
Phil
Thesis information:
http://www.philjackson.prohosting.com/PCJacksonPhDThesisInformation.html > Date: Mon, 25 Aug 2014 12:19:58 -0400
> From: sowa@xxxxxxxxxxx
> To: ontolog-forum@xxxxxxxxxxxxxxxx
> Subject: Re: [ontolog-forum] Discussion re reasoning about Time and State with REST interfaces
>
> Phil, David, Ed, Leo, Rich, Joel, et all,
>
> PJ
> > I would propose ... an artificial language, such as Tala, based
> > on the syntax of a natural language such as English, to represent
> > ontological information in general. Use other formal languages
> > and notations to represent ontological information as appropriate
> > in various domains.
>
> In the early '70s -- around the same time that Ed Lowry introduced me
> to his version of PROSE at IBM -- Fred Thompson presented a lecture
> at IBM on REL (Rapidly Extensible English). I liked both approaches.
> I also adapted an NL parser I designed to process controlled NLs.
>
> Several people at IBM (including Ted Codd and his group) adopted my
> parser to build some interesting systems. Some of them, such as Codd's
> Rendezvous and a Japanese front-end to Pascal, stopped at the demo
> stage. Others built useful tools that were used as a front-end to
> some internal IBM tools. None of them became IBM products.
>
> DW
> > I am a strong advocate of CNLs, as most of the long-term Ontolog
> > readers know, but I also want the system to be usable to support
> > existing computer systems
>
> PJ
> > Tala is *not* a Controlled Natural Language... Tala is an artificial
> > language that supports unconstrained English syntax, to represent
> > declarative and procedural concepts in unconstrained semantic domains.
>
> The range of notations called CNLs is rather broad. Any fixed boundary
> is likely to be violated by the next innovation. My recommendation
> is to embed a formally defined CNL within a very tolerant framework
> that tries to interpret anything thrown at it -- and switch to a
> dialog with menus for confirmation and clarification.
>
> PJ
> > This supports a point made earlier in this thread by Ed and John,
> > that no single model (or language) is adequate for representing
> > all ontological information.
>
> I agree. But where do we go from here?
>
> Five years ago, I presented the following tutorial at a Semantic
> Technology Conference in San Francisco:
>
> http://www.jfsowa.com/talks/cnl4ss.pdf
> Controlled Natural Languages For Semantic Systems
> A Roadmap of Directions to Explore
>
> Unfortunately, this is still a good survey of the approaches --
> because nothing more successful has appeared.
>
> Slides 22 to 26 summarize Adrian Walker's version of Executable English.
> Adrian is another of my colleagues at IBM, who couldn't make an earlier
> version of his system into an IBM product.
>
> Slides 30 to 32 summarize the REL and ASK systems developed by
> Fred Thompson, his wife Bożena, and their students at Cal Tech.
>
> Fred earned a PhD in logic from Tarski. Bożena had done research
> on MT for the Georgetown Automatic Translator (GAT), which later
> became SYSTRAN and Babelfish (which are still available).
>
> By the way, Fred had deeper insights than Bill Gates, who missed
> the boat on the WWW and the smartphone. Following is a prediction
> he made in 1992:
>
> FT
> > The next decade will see the telephone, personal computer, work
> > station, and television set combined into a single, ubiquitous
> > instrument - the telephone-computer.
>
> I largely agree with his summary of the issues:
>
> FT
> > Current literature often refers to a person's mental awareness of
> > the world as one's "cognitive model." But it is not a single model.
> > It is a large family of interrelated, comparable models - the many
> > alternatives that we visualize and choose among. The logician would
> > refer to these as the model-theoretic counterpart of our sublanguage
> > ... The linguistic formulation, we feel, grasps much more clearly
> > the characteristics of our ongoing cognitive processes.
>
> Fred not only predicted the telephone-computer, he and his group
> implemented a prototype on a Sun workstation in 400,000 lines of C.
> The link in my slides is no longer active, so I posted a copy of
> his article at http://www.jfsowa.com/misc/thomps92.htm . A note
> at Caltech: http://www.caltech.edu/content/frederick-b-thompson-0
>
> Questions: Why haven't these great ideas become the foundation
> for the Semantic Web? Should they be considered? If not,
> why not? If so, how?
>
> John
>
> _________________________________________________________________
> Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
> Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
> Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
> Shared Files: http://ontolog.cim3.net/file/
> Community Wiki: http://ontolog.cim3.net/wiki/
> To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
>