The bottom line : does text reuse translate into gains in productivity?

For the first time, language technologies have made possible the systematic (and semi-automated) re-use of text. They have brought text composition and text translation out of the artisan model and well into the industrial paradigm, and this has lead to gains in productivity. In the nineties, it was translation memory, with its promise of never having to translate the same sentence twice. This decade a similar pattern is emerging with the introduction of single sourcing for technical writing: write once, use many times. Along the way, content management and global content management systems are being developed to facilitate the tasks of publishing and archiving. Analysing these developments from a linguistic standpoint, it is argued that single sourcing and translation memory are driving the text towards 'chunkiness' and a-discursive language, while effective communication demands 'linearity' and the use of full natural language features. The impact of this less effective resultant language may offset those gains in produtivity. While we can't turn back the clock to the artisan age, it is important to direct the attention of localisation managers, language software developers and language professionals in a way that avoids the traps of the technologies and enhances the machine assisted production of full discursive text.