Machines, translations and memories : language transfer in the web browser

Ignacio Garcia

    Research output: Contribution to journalArticlepeer-review

    12 Citations (Scopus)

    Abstract

    Classic translation memory (TM) revolutionised technical translation in the 1990s. ‘Enhanced’ TM will revolutionise most types of translation this decade. The massive memory databases collected over 20 years can now be accessed from TM tools for exact and fuzzy matches. Sub-segmental information (terminology and phraseology) can also be gained from them via manual or automated concordancing. Stripped of metadata and formatting information, they have made possible the swift development of statistical machine translation (MT), which can be integrated with TM to provide matches where the memories themselves cannot. At the same time, cloud computing facilitates the adoption of web-based MT (and of web-based TM and translation management systems). Postediting, previously limited to a narrow section of the industry, becomes mainstream. The Read/Write Web demands (and allows for) a new model of translation, with dynamic, user-generated content not well suited to the conventional translation-editing-proofreading model. The web browser seems to drive those who want to make technical translation a profession into either semiskilled workers enduring quasi sweatshop conditions, or else into the highly skilled language engineer who helps make the decisions on which tool or process those translators will use for which task.
    Original languageEnglish
    Pages (from-to)451-461
    Number of pages11
    JournalPerspectives : Studies in Translatology
    Volume20
    Issue number4
    DOIs
    Publication statusPublished - 2012

    Keywords

    • machine translating
    • translating and interpreting
    • translation memory

    Fingerprint

    Dive into the research topics of 'Machines, translations and memories : language transfer in the web browser'. Together they form a unique fingerprint.

    Cite this