We, the authors of this letter to the editor of MultiLingual, have thought about Jaap van der Meer’s article in the July-August issue of MultiLingual magazine and intend to present an alternative view of the translation industry and professional translators.

We, of course, agree with Jaap that data is extremely important, but find his analogy with oil to be misleading. We also question the appropriateness of his labels for types of data (“language” and “translation”).

We agree that “neural” machine translation is simply the combination of algorithms and data, but are convinced that this observation by Jaap actually works against his claim that the need for human translators will soon go away. Much more is needed besides empty algorithms applied to data.

We agree that the translation industry consists of a “mixed model” (zero-cost raw MT and paid translation from LSPs) but strongly disagree with Jaap’s contention that this mixed model is unsustainable.

We believe that we have uncovered a fundamental flaw in his argumentation, namely, an unstated assumption that is false. Revealing his assumption and its consequences might show that the current mixed model is not only sustainable but beneficial to society.

We will present an alternative characterization of the translation industry data, a discussion of differences between artificial and real intelligence, and the reveal of Jaap’s unstated assumption. All this in an article we are preparing for a future issue of Multilingual!

Drs. Alan Melby and Christopher Kurz

—

MultiLingual welcomes letters to the editor and opinion pieces such as this one from our readers. To share your ideas, please contact editor@multilingual.com.