The Information Revolution

At the dawn of the information age, the vital role which information plays in business success is still surprisingly badly understood. Even "traditional" product- and technology-oriented businesses are now information- and IT-driven, with the quality and quantity of information available to corporate decision makers increasingly determining the quality and quantity of what is actually being sold. "Inward-bound" flows of information, e.g. up-to-the-minute market data and competitor intelligence, are vital to corporate positioning and financial health, while "outward-bound" communications such as marketing collateral and product documentation influence both pre- and post-sales acceptance.

Equally, the decision on where products are physically manufactured and distributed is now based on standard economic criteria rather than geographical considerations. Concepts such as outsourcing, componentization, the extended workbench and the virtual factory are one relatively abstract symptom of this trend; the transfer of production facilities to low-wage economies is a more concrete one. In practice, though to differing extents, the phenomenon applies both to manufacturing industry and to what has been called the "pre-programmed" services sector. What is more, it is also increasingly independent of company size, as IT lowers the entry threshold and levels the playing field for international SMEs.

In addition, new "knowledge organizations", which are designed to solve individual customers' individual problems , and which therefore depend even more acutely on accurate and meaningful information, are gaining ground. Such organizations - consultants are a good example - need to leverage existing customer (and market) knowledge to generate ongoing revenue, and are increasingly using sophisticated computer-based information systems to do so. Other large (and equally rich) service organizations with massive databases are also moving up the customization scale, as the interest on the part of banks and insurance companies in decision support systems and data warehouses shows.

Corporate Chaos

Added to this focus on information as a means of improving customer value is a growing recognition of the need to protect it as a corporate asset. As Jim McMullan at the Graduate School of Business University of Texas at Austin says, "any company that can figure out how to give its people the organizational knowledge they need - at the point and time needed - can position itself to compete more effectively and succeed much faster. According to him, "specific knowledge management activities help focus the organization on acquiring, storing and utilizing knowledge for such things as problem-solving, dynamic learning, strategic planning and decision making. It also protects intellectual assets from decay, adds to firm intelligence and provides increased flexibility."

While this may be true, however, the picture in practice is extremely bleak. Although many organizations are beginning to realize they have a problem, "corporate knowledge and corporate knowledge management systems are terms that are being widely used by organizations but with apparently little understanding of their meaning and the extent to which the technologies are available today" .

There are several reasons for this. One main one is the popularity of the last few years of radical - and often repeated - organizational changes. If badly managed (as is all too often the case), attempts to increase the speed and quality of reactions by downsizing and decentralizing operations and by flattening corporate hierarchies actually result in the loss of key information, and hence in uncertainty, delay and error.

In many organizations, middle managers and other experienced staff - who may well be fired or at least demotivated during such changes - are the main repositories of formal and informal corporate information. In addition, insufficient thought is generally given to the fragmentation of corporate information flows and procedures caused by decentralization. This is compounded by the tendency to turn operating units into profit centers, and hence to remove any incentive to contribute to longer-term, more broadly focused initiatives such as knowledge acquisition and dissemination (or, indeed corporate language facilities).

Added to this fragmentation of corporate information and workflows are the more prosaic - because more familiar - but no less tricky problems of coping with a heterogeneous installed base, legacy data and existing investments. This extends not only to the underlying hardware and software technologies used (e.g. conventional and free text databases), but also to the ways in which knowledge is ordered and represented within individual systems.

All This and Multilinguality too?

In such a situation, and given the ever-increasing economic constraints, one could perhaps be forgiven for deciding that the only way to manage complexity is to have everyone speak only one language, preferably machine code. In fact, however, multilinguality is unanimously being seen by both customers and language service providers as not only being here to stay, but growing in importance.

This is only understandable given the increased globalization of business, in which international workforces sell internationally sourced, internationally tailored products to international audiences. The Web is helping immeasurably here - as IBM's promotional material states, it allows companies to "open a store in every city in the world".

One side-effect of this development has been the rise or final establishment of the English language as the lingua franca for the digerati , and for the international political, economic, scientific and cultural communities. The (at least passive) language of this "superset" is the English of the Web and of CNN, Hollywood, Nature and the Wall Street Journal.

Equally, corporate language policies - to the limited extent that these have been formalized - often involve using one language as a "backbone" (in most cases within Europe - and in a surprising number of cases even within Germany - this would seem to be English). For the companies involved, this is often a pragmatic rather than an ideological issue - the total number of employees who speak English as a first, second, or third language will almost always be greater than those speaking Dutch or Portuguese - or even French or German.

However, a distinction is often made between run-of-the-mill in-house communications, and official documents and external communications. While the latter two categories are more likely to be treated professionally, and to be channeled through language service departments where such exist, the criterion for the former is often "being good enough". In addition, there is often a difference between oral and written communications, with a lower standard of linguistic competence being acceptable in face-to-face communication: as has been said, "the most common language in business is broken English" .

This does not mean, however, that everyone is, or should be, speaking only English. Initial replies to the MIM Workshop questionnaire show a clear tendency towards multilinguality in multinational companies, be they service providers or their customers. This may either take the form of "regional" or "local" languages supplementing the corporate language, or it may be an explicit or implicit (because not explicitly regulated) policy of multilinguality.

In addition, as any international executive knows, linguistic issues can intrude into company politics both officially and unofficially. In extreme cases, a - conscious or unconscious - desire to restrict the transmission of information may lead to the adoption of a "minority" language. Thus a British national, Cliff Wakemann, was recently awarded compensation for unfair dismissal by his Japanese employers, who according to him regularly switched to Japanese during meetings . Conversely, the ability to speak you boss's language may bring you kudos, a reputation as a head office spy, or even both. While such problems are more serious in respect to oral information, they are not solely confined to it, as access (or the lack of it) to written memos and other documents is also an aspect of corporate power struggles.

Another, far more fundamental, point is that globalization and market saturation in the developed world - coupled with political pressure in many cases - are leading to a relative reduction in the importance of established (and monolingual) markets such as the US in favor of developing ones. Since, to quote Willy Brandt, "If I am selling to you, I will speak English, if you are selling to me, dann müssen Sie deutsch reden" , success in sales increasingly depends on multilinguality. Schemes such as the Department of Trade and Industry's "Language for Business" in the UK have had considerable success in getting this message across, even if, again according to DTI estimates, some 75% of all European businessmen do not speak English in addition to their own language.

Equally, marketing collateral increasingly has to be provided in target customer languages, as a result of either legal regulations or market pressure. Here, too, the Web offers at least a potential opportunity of reaching an audience of millions for the relatively minor price of localized site content. A further advantage is that this applies even to geographically highly fragmented and/or relatively small language communities who could not be effectively targeted by conventional means.

Last but not least, the legally reinforced need within Europe for user interfaces and product documentation to be in the language(s) of their target communities also contributes to the steady to increasing attention to multilingualism on the part of many companies.

Servicing Multilingual Needs

There are several points which need to be made here. One is that there is a wide continuum of language needs and uses within the corporate environment, and that all are large and growing. 80% of respondents to the MIM Workshop survey thought that the business problems posed by multilinguality and attempts to manage multilingual information would increase; the other 20% (who were directly concerned with localization) thought that it would remain constant. On the one hand, there is the growing need for publication-standard materials such as marketing collateral and documentation . On the other, oral foreign-language skills and "for your information" translations are more important than ever before.

My own past experience as head of a language services department and conversations with others in similar positions suggest that in many cases facilities for these are located outside translation departments - in some cases without the latters' input or even knowledge. Such fragmented language workflows can also be seen in initial responses to our survey: of the customers (as opposed to the service providers) responding, only one (a major international IT company) claimed actually to have integrated language processes into corporate workflows, although another, again a major international IT company, was planning it.

Secondly, enterprises see language facilities and translation as a major cost factor (with an urgency not shared to the same extent by service providers). In other words, language services are the means to a (business) end, not the end itself. Management's perception of the role of language and the feasibility of the solutions proposed is thus crucial to success.

In this context, it should be emphasized that conventional language services are high-cost, non-core activities which use large amounts of relatively expensive labor, increase lead times and still cannot keep up with demand. Unsurprisingly, interest in delegating the problem is great. The survey revealed three main solutions - the increased use of external services, the increased deployment of language technology, and increased integration (these are not, of course, necessarily mutually exclusive). While the importance attached to the different obstacles to effective multilingual information management varied, the lack of tools and service providers was cited by many, as were more classic information management problems such as heterogeneous media and update frequencies.

Great Expectations

These expectations are luckily mirrored by recent developments in language engineering. For example, MT has seen a move away from sterile attempts at theoretical perfection towards a "good enough" approach which allows it to provide value for money in specific situations, such as FYI translation and high-volume repetitive texts such as manuals. Industrial strength term bases and translation memories are now well established, while term extraction seems ready for widespread integration in the relatively short term. While much remains to be done before one can talk of open standards or plug and play functionality here, the first steps have been taken, in the form of major integration projects in enterprise-wide environments. These projects, the benefits they can bring and the technology they are based on form the subject of the MIM Workshop.