Progress in hardware and software performance in combination with integration of information technology and telecommunication services are leading to an information society in which language and speech technology will become increasingly important. Whereas many aspects of the information age are genuinely global in character, the development of language and speech technology has an important national component.
The creation of a solid infrastructure for language and speech technology for any language requires that language and speech resources are being developed, maintained, and distributed. This requires organizations which determine the needs of education, research, and industry, which set priorities and standards, and which make sure that resources are maintained, documented, and available to interested parties. Coordination and funding by national governments is of key importance, especially for smaller languages: commercial applications will only be developed if the level of basic, publicly available, resources has sufficient quality to make the development of these applications economically feasible.
With approximately 20 million speakers, Dutch is a relatively small language (at least from the perspective of companies active in the field of language and speech technology) which is spoken in Belgium (Flanders) and the Netherlands. There are many striking differences between the language as it is used in Flanders and in the Netherlands. Nevertheless, both governments recognize the importance of a joint language policy for Dutch. For this reason, the Nederlandse Taalunie (NTU) (Dutch Language Union) was created, an intergovernmental institution responsible for language policy in the Netherlands and Flanders. The NTU is responsible for maintaining the position of the Dutch language, and as such it stimulates educational programme's (both in the Netherlands and Flanders and abroad), it supports projects leading to dictionaries, grammars, and other language resources, it promotes Dutch literature, and it advises the Dutch and Belgian government on language policy issues. One of its recent activities was advising the two governments on a spelling reform, which was implemented in 1995. The NTU also enables the Netherlands and Flanders to speak with in single voice in the international arena, for example in order to ensure the position of Dutch as one of the working languages of the European Union.
Language and speech technology is becoming increasingly important for
the NTU. For a long time, language resources consisted almost
exclusively of paper dictionaries and grammars. More and more,
however, the emphasis is shifting from paper resources to electronic
resources. For instance, it is currently investigated how the
Algemene Nederlandse Spraakkunst [Haesereyn
1997], a descriptive grammar
of Dutch produced with support from the NTU, can be made
accessible electronically. Another example is the recent spelling
reform [Renkema1996], which not only lead to the publication of
new paper word lists and dictionaries, but also necessitated the
release of new versions of spelling correction and hyphenation
programs. With a shift in emphasis to electronic resources, the NTU is also confronted with the certain responsibilities with
respect to maintenance and distribution of such resources. For
instance, the NTU will be responsible for the distribution of a
large (10 million words) corpus of spoken Dutch, which is currently
being developed by a consortium of research institutions in the
Netherlands and Belgium, as well as for the distribution of several
(bilingual) dictionary databases. Finally, the most important reason
for the NTU to be concerned with issues of language and speech
technology is the fact that language policy for Dutch will
increasingly be a matter of creating and stimulating an infrastructure
in which the production of tools and applications implementing state
of the art language and speech technology for Dutch is technically and
economically feasible.
As a starting point for its new task, the NTU is currently
investigating what exactly constitutes the language and speech
technology infrastructure for Dutch, how it compares to the
infrastructure for other languages, what the most important needs are
for education, research, and industry, and what coordinating actions
are required.