|Philoshophy and Applicability of Formal Languages - Minutes from the
Geneva, 15 September 2001
Mr. Meisingset opened the workshop explaining the background
for the workshop and expressed the belief that the workshop will become valuable
for future work on language standardization in the new Study Group 17, even if
the workshop is not constrained to this scope. He stressed that the speeches
will not present Study Group 10 results, activities and history. Each speaker is
expected to express his individual views and this may not comply with that of
their employer or role in ITU-T.
Mr. Meisingset presented in his first speech an extreme
nominalistic view on language and theory of knowledge, where phenomena instances
and classes are data in some observer, and descriptions are stated as denotation
mappings between name data and phenomenon data. From this, he claims that most
specifications are not models or descriptions.
Mr. Braek presented his views on methodology and descriptions
and provided a methodological framework identifying functionality, deployment
and implementation as three main areas to be covered by the ITU-T languages.
Within each area he made a distinction between object and property descriptions
and showed how the ITU-T languages and UML covered the six areas thus
identified. It was argued that it is meaningful to talk about specification and
design within each of the areas, using the same languages for both aspects. In
his view the distinction between specification and design is not a language
issue, but rather a methodology issue. It was mentioned in the discussion that
the presented framework could provide a basis for integration of the ITU-T
Mr. Gerome introduced semiotics and interpretation of
messages according to semiotic theory. He argued for the use of Peirce's notion
of triad (process, context and causes) to convey meaning of messages.
Mr. Meisingset presented in his second speech three aspects
of language use: denotations, instantiation and implementation.
The first part presented a proposal for a generic
specification language as an attribute grammar, where algorithmic behavior is
attached to the data nodes. The superior nodes of the syntax tree provide the
context of the subordinate nodes. Hence, in this attachment grammar, the
superior nodes must remain in the final statements, while the leaf nodes need
not be instantiated.
In the second part he presented a compiler inspired 7 level
transformation between two media. He used this framework to identify the scope
of SG 10 languages.
Mr. Jervis provided proposals for clear distinctions between
specification, design and implementation languages. He explained how MSC did not
necessarily comply to these terms. He pointed out that even if ITU-T languages
have formal semantics, that the mappings from terms to concepts were
well-defined, different users still have different real world semantics, that
they believed the terms mapped to different things and domains in the real
Mr. Sarma presented a realistic world view of science, as
opposed to idealism, which he considered the only other consistent but useless
alternative. He also compared software engineering with science, pointing at
similarities and differences. In the following discussion, he argued that the
classes do not exist in the real world, ref. the War of Universals.
Mr. Meisingset provided a short presentation of Popper's
Theory of science by falsification as Proof theory. He presented Theory of
science as a theory on the truth of statements about classes, while theory of
knowledge is concerned with truth of statements about individuals.
Mr. Subramanian presented various strategies to software
development and identified mathematical theories which could support each
strategy. He likened mathematics with specification of artefacts, ref. H. A.
Simon (1981). He uses category theory to study combination and development of
functions, e.g. for telecommunication protocols.
Mr. Sarma presented in his last speech his views on the
customer base and how we could improve the support and extension of this base.
In the following discussions, it was noted that several of
the speeches could contribute to developing a framework for Language
co-ordination. This is asked for in the Language co-ordination project, being
initiated by SG 17.
Also, SG 17 calls for a Framework and Scope Workshop Saturday
2 March 2002, and the speakers and others are asked to provide contributions
both to the Workshop and the
SG 17 meeting 27 February - 8 March.