An attempt is made to provide a brief outline of the present situation with regard to the language used for computer retrieval. This language is increasingly based on descriptors and thesauri, which contain the key words for manual, mechanical or electronic data retrieval. After a review of the general problem, the document deals with techniques of preparing standardized lists of descriptors, considering them separately from thesauri which represent a higher and more complex level of elaboration which is not always required. Consideration is then given to the classification and construction of descriptors and thesauri, and it is also recommended that this vital aspect of data retrieval should be standardized. Lastly, an outline of some systems of descriptors and thesauri used in the international field is presented, ending with an explanation of the project which, as a subsystem, is being implemented by CLADES. This document should be considered not only as an information document, but also as a working document especially designed to help the participants in the Meeting on Modern Documentation Techniques to decide what standardization principles should be applicable to descriptors and thesauri in the Spanish-speaking region. (Author/NH)