Language and technology: translation vs. education dilemma

Dr Piotr Wozniak, SuperMemo R&D, Apr 12, 1995

This lecture has been presented at Language and Technology Conference, Infosystem 1995, Poznan, Poland

One of the recently promoted trends within the EEC (and not only) is the vision of the multilingual Europe without linguistic barriers. The vision assumes realistic targets for introducing computer technology and employing educational directives that would make it possible to reduce the mounting costs of language translation in the community within the framework of several years.

A far more attractive proposition, in the author’s view, is the vision of monolingual world (not just multilingual Europe) without any barriers (not just linguistic barriers). The proposed transition from the popular vision backed by the European administration to the more controversial long-term vision for the global community implies a visible difference in the implementation approach.

While multilingual community without linguistic barriers puts the main emphasis on perfecting translation tools based on natural language processing technology (with a degree of help from multilingual education), the monolingual world might require, at least in its initial phase, more emphasis on education of the standard communication language for speakers of other languages. The ultimate solution to the global language problem comes, as suggested by Mario Pei back in the 1950s, from establishing a standard language for international communication and introducing regular schooling in this language. Human and financial resources permitting, this language should be introduced at the earliest possible time in educational systems worldwide.

The author’s main interest is focused on optimization issues in the process of learning; including learning languages. Optimization of the four following areas can shorten the path to solving the language problem via the educational approach:

  1. the status of the receiving system (here: the human mind and body): optimization refers to health and the mental hygiene and currently is not aided much with computer technologies
  2. access to knowledge (apart from standard access tools, the knowledge access system should be aware of the learner’s progress in navigating a knowledge hyperspace)
  3. representation of knowledge (e.g. using mnemonic techniques)
  4. retention of the learned material, which can be best accomplished by the application of repetition spacing algorithms such as SuperMemo

SuperMemo Word has developed a software line under the commercial name of SuperMemo for platforms such as Windows [cf. SuperMemo 98], DOS, Macintosh and Commodore Amiga. SuperMemo software applies optimization techniques that make it possible to apply quasi-optimum repetition spacing in the process of learning. In other words, the fourth of the aforementioned optimization issues is dealt with in order to greatly accelerate the process of acquiring new knowledge by minimizing the time necessary for retention of the previously learned material.

Repetition spacing makes it possible to perpetually retain pieces of information in the learner’s memory by effecting less then ten repetitions of the material in the lifetime.

The future development of the SuperMemo software line will go towards the implementation of the so-called Knowledge Machine that incorporates two other optimization issues in learning: access to knowledge and knowledge representation [cf. SuperMemo 98].

At present, the easiest way to become familiar with the repetition spacing technology for language learning applications is to take a look at SuperMemo 6 Public Domain for DOS available from a number of on-line shareware houses