Evidence of semantic association in technology-related vocabulary : a corpus study of collocations.

Abstract

This paper presents a semantic study of corpora-based words which meanings are related to technology, exploring how such words have been semantically coined over time regarding their most frequent collocations. In this study, we counted on three corpora (DAVIES, 2010): The Corpus of Contemporary American English (COCA) and the Corpus of Historical American English (COHA) were used as source of data about the English language. The Portuguese language data was assessed through the corpus do português NOW (News on the Web). The screening procedures were respectively: a) Selection (words that probably have changed semantically over the years); b) Comparison (check the collocates and their frequency); collection: a file was created with all the comparisons regarding frequency and strength of attachment). In order to verify the existence of statistically different groups, we applied T-test and considered the percentage of appearance for the target-collocations as well as the MI score. Results indicated meaningful changes to words related to internet milieu when time is considered and regarding collocation. Similar results have been found for both Portuguese and English. Moreover, data revealed that the three most frequent collocates for the target-words when present time was considered not only demonstrated a clear change towards internet-related meaning, but also showed a higher strength of attachment.