G.S.Tseytin. Features of Natural Languages in Programming Languages. Logic, Methodology and Philosophy of Science, IV, North Holland, 1973, 215-222
A number of features is discussed, each discovered in
syntax or semantics of natural languages and considered or
introduced (possibly independently) in programming languages,
including: phrase-structure grammars and the balance of
syntactic and anaphoric links, recursive definition of
semantics, need of defaults and preferred parsings, declarative
vs imperative, gradual elimination of jumps and assignments.
G.S.Tseytin.
From Logicism to Proceduralism (an autobiographical account).
In: Algorithms in Modern Mathematics and Computer Science,
Lecture Notes in Computer Science, 122, Springer, 1981, 390-396
An account of how experience in both design of special-purpose
computer languages and study of natural language semantics led
to abandonment of plans for mapping NL semantics into a kind of
predicate logic, while direct manipulation of names
(rather than the entities they allegedly have to represent)
turned out to be more successful.
G.S.Tseytin.
On relationship between natural language and formal model. (In
Russian.)
In: Voprosy kibernetiki, Obshchenie s EVM na estestvennom
yazyke, Moscow, 1982, 20-34
(ÒÕÓÓËÉÊ ÔÅËÓÔ ÚÄÅÓØ,
ðóññêèé òåêñò çäåñü)
It is argued that a mathematical model of a language should
not attempt to describe the language as a whole (especially when
it comes to semantics), but that there should be an open
collection of partial models each representing a small fragment
of the language (words or clusters of related words, small
problem domains, grammatical paradigms, etc.) Interfaces between
such fragments can be defined in terms of selection of an item
in an environment by name, the name being supplied by one
component, and the environment by other. Comparison with
various technological "languages", including programming
languages, is extensively used.
G.S.Tseytin.
Programming with association nets
(English translation here). In: EVM
v proektirovanii i proizvodstve, vyp.2 (G.V.Orlovsky,
ed.),"Mashinostroenie", Leningr. otdelenie, Leningrad, 1985,
16-48
A rather extensive description of a formalism
to support open-ended systems (as needed for AI and NL, with
identification by name playing the key role) and a simple
implementation on IBM-like mainframes. The data structures are
association nets (conceptual graphs), the system features
modularity (in an object-oriented style) and object "merging" (a
remote analogue of PROLOG unification). Some examples are given.
G.S.Tseytin.
On relationship between mathematical and common thinking. (In
Russian.)
Izvestiya AN SSSR, ser. Tekhnicheslaya kibernetika,
1987, no. 2, 193-196
It is argued that normal thinking cannot be regarded as a
derivative or a special case of mathemathical thinking, and
therefore its study cannot be based on mapping to predicate
logic but may require independent mathematical constructs.
M.M.Zheleznyakov, T.N.Nevleva, I.M.Novitskaya, L.N.Smirnova,
G.S.Tseytin.
An attempt of building a "text -> reality" model using
association nets (English translation
here). In: Mashinnyi fond russkogo yazyka: predproektnye
issledovaniya. Institut russkogo yazyka, Moskva, 1988, 140-167
An approach is described to meaning extraction in a restricted
problem domain (with association nets as an instrument), and
some linguistic implications are discussed.
G.S.Tseytin.
Towards unit-based programming (in Russian).
Programmirovanie, 1990, no.1, 78-92
The quest for modularity (never fully achieved) is represented
as the pricipal driving force behind the development of
programming language concepts.
G.S.Tseytin.
On some mechanisms of representation of meaning in
natural languages. The Prague Bulletin of Mathematical Linguistics,
65-66, pp.5-12 (1996). (First presented to LPAR'91, a
conference on logic programming.)
A survey of the
approach to natural language semantics via restricted problem
domains, and of the linguistic phenomena covered by this
approach.
G.S.Tseytin.
Bottom-up and top-down approaches to text
analysis. ELSNET goes East and EMACS Workshop on Integration
of Language and Speech (November 9-11, 1995, Moscow, Russia),
Proceedings. Moscow, 1996, pp.210-214.
A balance of
the two approaches is needed in processing of both speech and
written text.
G.S. Tseytin.
Association Nets: An Alternative Formalization of Common Thinking.
Lecture Notes in Computer Science 1234, pp.385-398, Springer, 1997.
Invited talk at the
Symposium on Logical Foundations of Computer Science
Yaroslavl, Russia, July 6 - 12, 1997.
Association nets described in previous papers now
presented in a form close to traditional mathematical logic.
A revised version with computer-tested examples is here.
Postscript version to download.