This post was originally published on September 20, 2010.
A Very Short Recent History of Translation, Interpreting and Localization
If the introduction of whispered translation is due to Charles de Gaulle and that of simultaneous interpretation is due to Telford Taylor and Richard Sonnenfeldt for the Nuremberg trials, the introduction of localization is due to Microsoft.
In fact, in late 1980′s, Microsoft released a corporate manual for the development of international applications, The International Handbook for Software Design, in an attempt to improve internal efficiency and reduce the costs of developing the international versions of its products.
Since then, with the growth of revenues from sales in foreign markets, the internationalization group grew in importance to eventually lead Microsoft to completely revise the product development and marketing policy, and the handbook that became a kind of sacred text.
Before then, especially in large companies, translation and localization were handled internally. In the various IBM translation centers around the world, for example, ranks of translators and localizers were trained who then gave birth to other companies even if their business and communication model was biased towards their previous experience, and proved too stiff to meet the many challenges of innovation in those years.
The following years were very important for telecommunications evolution. At first there was the fax that its instant success due to the maturation of the underlying technology, although available for several years, and to ITU’s efforts that led to the definition of a major standard, and thanks to its spread in Japan for the ability to transmit ideograms without transcription.
In that same period the first data communication tools began to spread and BBS’s were born, with FidoNet and CompuServe, although they were hard to use without sound technical skills.
In Europe, then, some multinational corporations had established their research and development departments in Sophia Antipolis “technology park”. Microsoft chose Ireland: the language was the same and the country was experiencing very hard times and needed foreign investors to bring in hard currency.
The way Ireland took advantage of the opportunities of that period and how it is today, partly because of the choices made then, is all another story, and a treaty would be just enough to tell, even because the subsequent identification of India as a technology center is basically the most obvious consequence.
The technology push in those years and the consequent demand for localization produced by marketing strategies in foreign markets led the way for the first language services companies with a strong technological flavor.
1994 marks a historic transition. Trumpet Winsock allows the connection to the Internet possible even to Windows 3.11 users with dial-up access, and helped spread the Net.
The second half of the ’90s then saw the expansion of the the so-called new economy bubble that peaked in 1998. The birth of the major companies currently in the global market dates back to those years.
Lionbridge, in fact, was founded in 1996 and was listed in 1999, SDL was founded in 1992 and was also listed in 1999. TransPerfect was founded in 1992.
RWS is the oldest of large translation companies, having been founded in 1988 from a merge with pre-existing companies the oldest of which dates back to 1958. RWS obtained ISO 9000 certification in 1992. Logos was founded prior to RWS in 1978, and his story tells of an effort in trying not to shrink to language services and extend to media.
All the large companies adopted more or less the same strategy based on the acquisition of smaller companies and technologies, but it was fatal to Lernout & Hauspie, founded in 1987, listed in 1995, and went bankrupt in 2001.
All of them also have another trait in common: their founders are not translators, and if some were, as Rodrigo Vergara, they were such out of necessity. TransPerfect is indeed emblematic of a way to entering the market and acting this industry; in fact, it was founded in a dormitory at the New York University by two students of the school of economics who were fasting the whole matter.
The appearance of customers with large spending power who can grant rich contracts led to the birth of aggressive techn-savvy businesses.
Until the mid-80s, translation companies were geographically characterized. Often, freelancers, agencies, and small companies rebuked each other for having their spaces invaded, even when these spaces were not even near.
In those years, the value added was a personal computer and the ability to deliver a job on a floppy disk, to connect to a bulletin board system to download or upload a translation job. This was a huge competitive advantage.
The explosion of the new economy brought the Internet to the attention of traditional media and the element that had been a discriminating factor for success so far became mandatory.
The abundance produced the steep fall of fees that, to a lesser extent, continues today. Suddenly technical knowledge and skills ceased to be important with the possibility of recruiting anyone, anywhere.
All this happened less than fifteen years ago. Only eleven years ago, moreover, the world struggled with the millennium bug presented as the worst nightmare to live, Altavista was the favorite search engine because Google’s founders were looking for a place to move from the garage where they run their business, and Mark Zuckerberg was in high school and looked for a picture for his class’s yearbook.
Globalization, Fragmentation, and Commodification
Is globalization, then, the cause of fragmentation of the translation market and its commodification?
Actually, it is only in part. Globalization, in fact, prompted the major companies to abandon the traditional crafting approach for a semi-industrial one, but none of them proved adequate to develop a business model other than manufacturing.
Furthermore, the illusion of having access to flexible workforce in unlimited quantities thanks to the spread of the Internet has prompted the major players to adopt a pure brokerage approach and to organize accordingly by focusing on management and administrative activities with minimum productivity and low value added.
The semi-industrial approach, however, limited the options and forced competion on price only, but the business and operation models reduced profit margins and productivity from the beginning. Over the time, the original defect of this choice, the limited downward scalability, imposed the growth in size to cope with the volume of work necessary to cover costs, thus triggering a downward spiral.
On the other hand, to attract large-size customers the offer should be adequate in terms of overall capacity.
To recover from the profits loss in the discount race to grab the customers with the higher spending power, not necessarily the best, the most obvious solution was exerting a wide variety of pressures on suppliers. The most original of these pressure that contributed most to the commodification of translation has been the discount on the so-called fuzzy matches, which, however, do not provide the final customer with any direct benefit, and are used only to reduce the remuneration for translators and allow for recovery of profits on the prices to the clients who, especially if unaware, is offered only an overall discount for leverage in general.
Too often, moreover, we forget that the goal for the birth, development and use of translation memories is to reduce the cost of human translation, not to simplify the work of translators that is a widely marginal aspect used as a decoy against translators themselves.
Anyone who has worked on projects with a large number of fuzzy matches, especially with a percentage above 85%, knows, in fact, that handling these segments requires an effort largely exceeding the discount imposed on their treatment.
For the rest, globalization has had the same effect it has had on other businesses, while the inherent weakness of the model and the approach chosen has in fact penalized an entire industry that has always shown different trends in respect with other industries in facing the crises that have occurred in the last three decades.
Although unfortunate, this business model that already pervaded the industry has definitely affirmed and extended. In the absence of barriers to a traditionally free and deregulated market, the apparent success of the most aggressive companies has also led to further fragmentation. In fact, if once fragmentation was only a local phenomenon, globalization tools made it common place and fed the illusion of accessing an unlimited market and flexible workforce in unlimited amount.
Confusion in Quality Management
With the same delay with which the industry’s leading companies implemented oudated and totally inadequate models such as the fordist model, they also tried to rebuild an eroded competitive advantage through the adoption of quality assurance practices that have applied for many years in the manufacturing and services industries, and widely regarded as mature.
Those practices were actually mature. Translation entrepreneurs were not. The obvious evidence is the still unsolved confusion about QA, but the industry immaturity was also witnessed by the deep differences within the BT/TF 138 working group convened to develop the EN 15038 European quality standard, which in fact is a poor outcome. The fact that it was actually inspiring the similar Canadian standard and was taken as a reference for the American one does not make it better. Moreover, the approach for ASTM 2575 cannot and in fact does not repair the errors of all industry standards, from the first, the Italian UNI 10574 to the German DIN 2345 and the Austrian ÖN 120X, all suffering from the same original sin: the attempt to assert a specificity that is deemed unattainable by non-experts, thus betraying the true spirit of quality standards.
These are based on three founding principles, recently enriched by a fourth:
- describe your processes in writing;
- comply with them;
- allow others to verify that the processes underway match those described and to correct deviations;
- continually review your processes for continuous improvement.
Quality system certification consists in the endorsement of the positive match between the processes described and the processes in place; the path to certification has its costs, often substantial. This investment requires the full involvement of both top management and staff.
Usually, the first noticeable benefit is the detection of inefficiencies, once corrected, the improved efficiency translates into a permanent cut their costs. Similarly, the effort towards continuous improvement allows for the continuous streamlining of processes and makes a company more agile and responsive in meeting the competition with its competitors.
Translation companies, however, borrowed the worst practices from the manufacturing industry companies, from late payments to raise and support cash flow to face the delayed payments of their clients, to the continuing requests for discounts, particularly towards traditional suppliers, usually associated with the silly motivation of large volumes (to line up more work for little money), to the foolish decoy of guaranteed continuity of orders (as if denying it was not so easy) to end with an absolute masterpiece: asking vendors to pay for using a poor software platform to continue working.
This aspect is emblematic of a technological capacity that is as vaunted as fake. A company with a winning product would sell it on the market according to the market rules, possibly through a financially and operationally independent spin-off to manage its development and marketing, instead of showing cowardice and inability to support its choices.
Relying on unequal terms to exert the pressure deemed necessary to maintain a competitive advantage or an uncertain dominance also reflect the belief that the flexible workforce in unlimited quantity that can accessed easily is an illusion and it actually consists of poor resources whose main feature is despair.
It is also true that translation is a traditionally “feminine” activity typically made for moonlighters and that, therefore, it is perfectly right for the application of “unequal terms.” In fact, this is a widespread view reflecting only very little strength and negotiating power that are actually preventing any possibile change.
The implementation of a quality system and its subsequent certification in accordance with acknowledged criteria has no role in the evolution/involution of corporate conduct. Conversely, a quality system could be used to hide a substantial uncertainty that process formalization and their forced compliance eventually boosted.
The efficiency gains should affect processes, and excessive bureaucratization should indicate to inspectors that artificial processes are in place that should therefore be corrected.
One of the first corrective actions should focus on the very concept of quality that is clearly not shared. As Renato Beninatto rightly pointed during episod 6 of his meetings dedicated to the translation with Kirti Vashee, quality should not be negotiated, as it is a prerequisite, and is relative; what really matters is the quality of service, whatever it is, that can certainly be considered a process function that leads to service delivery.
For these reasons, a major customer who commissioned a complex translation project of dozens if not hundreds thousands of words in a dozen, if not more, languages, will judge quality on metrics that are very little to do with language. The major players in this industry have built much of their fortune on these assumptions, giving different and irreconcilable visions to customers and suppliers for the sole purpose of profiting from both.
Universal Standards and Industry Standards
Therefore, it is not necessary to change processes to comply, for example, to ISO 9000:2005 and ISO 9001:2008, and there is no reason to claim that these and other standards may have in any way destroyed creativity (!) of translators and mechanize their precious cognitive processes, even when they produced laughable results that are very far from what their customers and the public expected. Regretting the good old days when no one would apply quality standards to the translation industry is an exercise of nostalgia perhaps comforting, but sterile and empty.
On the other hand, poor models can only match with inefficient or only partially effective processes and poor enforcement of standards, that are, anyway, written by those to whom they are destined.
Even the so much invoked disintermediation may not be useful for the restoration of typical market rules. Disintermediation, in fact, may serve to streamline processes and reduce costs, not to upset long-established bad practice. To streamline processes, however, good initial conditions are necessary, while the list of companies with which translators suggest to work seems saying that these conditions are missing.
In fact, disintermediation provides for the elimination of all intermediaries in the supply chain, resulting in drastic reduction of operating costs. The Net helped disintermediation by allowing for the direct connection of customers and suppliers, and e-commerce solutions are a classic example. For years the flag of disintermediation were Amazon and Dell selling directly to consumers, even though in both cases disintermediation was mitigated by the need to use vendors for the delivery of goods and/or warehouse management.
ATMs and online banking, are a perfect example of pure brokerage; online banking allowed for the creation of virtual banks offering better terms than those offered by traditional banks because of the reduction of operating costs, typically infrastructures and workforce.
Also bank customers, then, benefited from bank disintermediation.
However, banking disintermediation was possible and, above all, effective thanks to process maturity and the ample room for efficiency gains that were still possible through the use of similarly mature technologies.
Disintermediation may have room for even greater success in the production of intangible assets that the end user can directly order to the producer in real time and pay directly, thus assessing costs and benefits in their actual value.
Translation, however, is not a good that can be spared in stocks, despite memories, and without translation factories is in fact impossible to fulfill orders with the immediacy and urgency expected under the technology used.
In the translation industry, brokers has always played a purely organizational role, and organization and rationalization of production processes are the key to disintermediation. For companies that have built their business models on brokerage, the reduction of operating costs should be a primary goal, but disintermediation also requires a drastic reduction of profit margins and the launch of a cycle of continuous refinement of operating processes.
Therefore, disintermediation could be effective only if applied to basic services, and this requires a further contraction in compensation to suppliers to support profits and a fast return on investment. In addition, disintermediation alone cannot solve any cash flow shortage, which may indeed be even more serious given the investments, the abuse of delayed payments will certain not diminish.
Some predict that disintermediation in the translation industry will be made through collaborative platforms for hire in SaaS mode, because of the lower costs that professionals will face, the certainty of safe and continuous updates and the ability to continue to monitor assets.
To succeed, however, these platforms must be authentically and fully independent. They should not be owned or operated by any LSP, or worse, by any MLV. On the other hand, customers may access them to follow the progress of projects.
So far, crowdsourcing has been the bugbear of professionals who have not seen the most interesting opportunity. Companies that have used crowdsourcing have not skimped on investments in technology and have not waived professionals. They rather set priorities and areas of action according to principles and business models that sound strange translation industry, even before the professionals, simply because it is not possible to translate everything but would translate everything.
In the short to medium term, the game will be payed around the user-generated content that are, by definition, volatile and changeable, and their translation is not worth the money requested by more and more greedy LSPs and MLVs more and more crippled by organizational and development problems.
Moreover, a globalized economy requires productivity increases that affluent countries are unable to support, at least not at the same rates of emerging countries. The demand for translation is not diminished thanks precisely to the emerging countries and it will likely grow further and at faster pace, precisely because of their growth.
The trend towards delocalization is now at its peak. The search for cheap labor has also led a historical mark of excellence in commodities (tea) to move production to cheaper countries.
Current working conditions are also the result of the demands from the 60s and 70s when “how” came before “how much”. In the last few years the deterioration of payments for work has been fed, driven and encouraged to make the reshaping of working conditions acceptable, but this is the typical trap since worse working conditions cannot lead to better compensations.
The abundance of cheap labor and low dignity following the new geography of work that occurred with globalization helps new exploiters to pretend to spread an overall responsibility and blame the new slaves for irresponsibility and selfishness; they count on the ambition of those living and aspire to live from “privileged” jobs as even translation is considered, although incredibly.
It is just a trivial form of vassalage and should not be difficult to identify the correspondences between the schema that is given as established and indisputable and the various layers of feudal society.
The solution that the manufacturing industry supposedly identified consists in concentrating production in limited places and businesses and in reducing the overall supply. Only in the coming months we will know if this solution will lead to productivity and, above all, competitiveness increases.
Relocation, however, also involves a transfer of wealth. When this is complete, migration from poor to affluent areas will increase and there will be a reverse relocation.
So fare we could give for ascertained that the solution that manufacturing industry adopted may not work well in the translation industry.
Indeed, if manufacturing industry workers, thanks to strong automation and maturity of processes and methods may see themselves forced to accept an even greater schedule flexibility, restrictions on the right to strike, wage contractions, and radical reform of contracts because they are largely replaceable, a translation industry Gresham’s Law variant could be postulated stating that the continuous flow of poor resources leads to the permanent expulsion of precious resources that cannot be replaced, due to the progressive and relentless decline of the industry.
What is worse is that this is a trend that cannot be reversed, since there will be no more qualified resources to educate their sobstitutes or willing to re-enter the market.