Translation as positive 'cultural' action has illustrious forebears. The Russian poet Pushkin likened translation to "a change of horses at the post houses of civilization". In a pre-technological world massively fragmented by topography, acute linguistic diversity and distance, language translation must have been one of the oldest professions.

Projects that made history

From the perspective of our burgeoning industry today, however, what stand out from the past are less the billion and one micro-acts of translation than the large scale projects that have transformed minds and behaviors. Think of the content of Roman culture, largely a huge act of translation of the Greek legacy. And there must surely have been some interesting workflow experiments during the fabled translation of the Septuagint, when 72 translators went to Alexandria from Jerusalem in the 3rd century to translate the Pentateuch (not the whole Bible) from Hebrew into Greek in just 2.4 months.

Nearer to us, a vast effort must have been made by any number of south-east Asian translators to localize the huge corpus of Mahayana Buddhism into such languages as Burmese, Vietnamese, Thai, Chinese and Japanese from the 6th to the 8th centuries. And about the same time, in one of the more curious translation cascades in history, Syriac versions of Greek philosophy and science were translated en masse into Arabic by a team of scholars in Baghdad, some of whose work was then to be translated back into Latin a few centuries later in Spain, France and Italy to form the sole source of Greek learning in Europe, before the rediscovery of Byzantine manuscripts in the East helped launch the Renaissance of inquiry in western Europe.

All of which pale beside the global resolve of the Bible translation project from multiple source languages (Hebrew, Aramaic and Greek), which over the centuries has localized the Book, often by individuals, into over 550 languages and rising. Interestingly, the source code of Islam's key texts (the Koran and the Hadith) was declared to be untranslatable, and converts had to learn the original Arabic. European scholars only began to translate the Koran in the 17th century for purely informational purposes. Perhaps the only project today comparable to these concerted transfers of social knowledge is the gradual translation of European Union law—80,000 pages of it—into the languages of EU member states by virtue of the Treaty of Rome: as I write, this process is ongoing in such next-wave national languages as Hungarian, Czech and Polish.

Jerome the Productive

We know next to nothing about the daily practices and processes that developed during these foundational endeavors, nor how much they cost in economic and human terms. Presumably translation teams pioneered such tools as bilingual glossaries and alphabetic look-up, and developed mark-up procedures for quality-controlling output and comparing different versions from different translators. One of the earliest multilingual corpora is Origen's famous Hexapla versions of the Bible from the 3rd century, which comprises six columns on large pages, one giving the Hebrew consonantal text, the second giving a spelling of the Hebrew using the Greek alphabet, a third offering a literal translation into Greek, the fourth another Greek rendering of the Old Testament, the fifth giving the Septuagint version plus critical symbols to show where modifications were necessary, and the sixth a further revision of the same Septuagint. This huge unwieldy document base was kept as a resource in Caesera until 600 CE, and was no doubt consulted by Jerome, the patron saint of translators, when he made his famous translation from Hebrew and Greek into Latin.

Jerome himself seems on occasions to be something of a speed freak. In 398, legend has it that he translated the three 'Solomon' books in 8 days, which judging by the English version of 56,000 words makes for a daily translation load of some 7,000 words a day! He almost certainly dictated to stenographers working in relays, but one wonders why his client was in such a rush for these three files. After all, the whole Bible took him 22 years to complete! A single translator working at 2000 words a day would, other things being equal, take about 2 years to translate the whole file set. A suitably resourced machine translation system today, working at 15,000 words a hour, could gist the lot in 3 days.

Globalization and its discontents

Why this historical digression? We are hearing much loose talk these days about civilizations clashing while history comes to an end. Such concepts offer us quick-think models for organizing our ideas in a world where globalization and its discontents are prompting reactions of retrenchment and fear. Our own industry is only a couple of decades old at the most, and is currently confronted by the threat of an economic recession that is likely to reduce certain categories of localization spending and re-channel investments into streamlining processes rather than exploring new markets. Without rehearsing the old saw about those who ignore history being condemned to repeat it, we can identify three findings from localization's role over the centuries that enable us all to navigate through the turbulence from the vantage point of a somewhat philosophical crow's nest.

• First, the localization industry, and the myriad daily acts that it undertakes all over the world, is itself yet another of history's great projects. This is not just self-serving megalomania. The rise of a post-modern networked society—most vividly embodied in the technology of the world wide web—represents an extraordinary transformation of ways of doing business, learning, governing and living. Localization's ultimate role is to domesticate this brave new world, to bring it all back home, to prevent it from being just a wild frontier where anything goes. By putting in place the practices and the technologies that in the end can help bridge that digital—and cultural—divide, we are, albeit in another register, the heirs of those early pioneers who localized the religious teachings that formed the foundations of culture as we know it. Concretely, this means extending proven practices and standards to translation and localization markets that we have, as an industry, perhaps been blind to, as trados' Jochen Hummel argues in his article in this issue. And as Rodrigo Vergara of Logos suggests, our own experience of fine-tuning translation processes can be exported to a variety of other professions who have to deal daily with the challenge of multilinguality hard-wired into our communication systems.
• Second, our people matter. History retains the names of a select band of famous translators—Jerome and Luther for the Bible, Hunayn ibn Ishaq for Aristotle, Budé for Greek learning into French and so on—whereas commercial localization today is performed by legions of unsung heroes who work in teams in close symbiosis with technology to complete large-scale projects—sometimes faster than Jerome's three-file blitzkrieg. Any serious comparison of the methods and economics involved is of course pointless. Yet at every stage of the localization process, however powerful our technologies, decisions are made by individuals who need all the help they can get from the community of best practice. The industry is today making a significant effort to ensure that knowledge is sharable and shared, that learning support becomes an integral part of our mutual responsibilities towards the people who drive it forward. In this issue, Yves Champollion's article shows how an freelance translator was prompted to develop a tool that would enhance individual productivity without turning everyone into a mark-up language wizard on the side.
• Third, content is process. It is tempting to think that the content we localize is independent, in some deep-structural sense, of the language versions produced. And of course in a complex way this is true. Yet scholars tell us that Luther's translation of the Bible is said to have transformed both the Book's impact and the German language, Hellenic civilization in Egypt was not Greece, Arabic translation of Galen inspired new medical investigations, post-Renaissance translations into English coined thousands of new words almost overnight. This dialectic between source content and target version continues today, as we struggle explicitly with the need to find ways of ensuring that powerful messages co-evolve with their use in new contexts. Again in this issue, Ben Martin of J.D. Edwards argues that authoring for localization requires terminological work upstream of the translation, so that source and version co-evolve over time.

So why my title—communication breakdown? Because a world without localization projects—unimaginable as history tells us—is one where ideas and knowledge, best practices and meanings will be lost. Where everyone has to re-invent processes on a near daily basis. Hence the importance of a strong industry capable of growing a market and enhancing its workforce.

But wait! The phrase has a second meaning—'breakdown' in the sense of an inventory of parts. Provided we can maintain awareness of communication's components—that communication is a global project, that it is performed by individuals working in teams, and that whatever the source, every act of communication is in some real sense an act of localizing meaning to others—we shall hopefully maintain our course in the stormy weather ahead.

Andrew Joscelyne (ajoscelyne@bootstrap.fr) is the LISA Newsletter's new editor. He has been a localization and language technology watcher for over a decade. He has worked as a journalist for most of the language industry publications, and has acted as a consultant on multilingual technology programs for the European Commission. Based in Paris, he is writing a book on the history of language technology.