A device in order to improve the quality of machine translation, based on the correlational theory of thought
Giulio Benedetti · 2005
In this article the author describes a device in order to improve the quality of machine translation, conceived by Silvio Ceccato and based on his correlational theory of thought. Basing themselves on this theory Ceccato and his collaborators worked on the problem of machine translation from 1959 to 1966, carrying out almost exclusively theoretical researches. The project was interrupted when MT research in USA stopped because of the well-known ALPAC report 1 , before it was possible to evaluate the results this device could give. The author thinks that this device and the theory it is based on, although very little known for various reasons, are still extremely valid today. Therefore they could permit, in the field of MT, better results than the ones we can achieve with the programs available today, especially when the source language is a language with very little morphology and a lot of ambiguities, such as English, and the target language is a language with rich or very rich morphology. This is essentially because: a) the correlational theory of thought is (also) a linguistic theory which is completely new and very different from all others; b) the program based on the aforesaid device partially reproduces what the human being really does by using his intelligence when translating. Besides, this MT system is probably the only one based completely and exclusively on a single linguistic theory and conceived by the person who proposed the linguistic theory itself. The only problem it presents, which is the fact that the human work that has to precede the program implementation increases exponentially when the number of the dictionary entries increases, is not theoretical but only practical. Warning This article cannot be understood without having previously acquired some knowledge of Operational Methodology. sufficient knowledge can be quickly acquired reading my article A presentation of Operational Methodology . The present article has in fact been conceived as having the other article as its indispensable preamble. Reading another article of mine, Basic mental operations which make up mental categories, is not strictly necessary but it can be useful in order to have a more complete idea of the subject. This article has been conceived to be read also by persons lacking any knowledge of Machine Translation, but who are interested in knowing how Operational Methodology and the correlational theory of thought can have an application in this field.