What Price Patent Abstract?

by

Kriemhild Zerling

Once again small businesses are calling translators and editors about the translation of patent abstracts from any foreign language into standard American English. Last year the contract originated from the World International Patent Organization (WIPO); this year, from the United States Patent and Trademark Office (USPTO), which is awarding the contract not just to the lowest bidder, but the bidder who also demonstrates the highest quality. To this end, the USPTO has included sample translations as a means for assessing both the quality and the cost of the translation. Of course, the cost set by the vendors providing the translation services will not be the fee, which is offered to the actual translation providers/editors and which will vary from language to language. Emphasizing the high volume and the small size of the abstracts, many vendors are providing the USPTO a discount and passing the decrease on to the translators. If, for example, the translator's language pair is German to English, the fee may range, as in the case of the WIPO contract awarded last year, from an initial 4 cents per word to a final maximum rate of 7 cents per word. Assuming an average length of 50 words per abstract, the total remuneration may range from $2.00 to $3.50 per abstract. At the latter rate the translator would have to translate 10 abstracts to earn $35 per hour. Not surprisingly, 10 abstracts is also the rate of disposal that is frequently set by agencies to establish the editor's base fee of $35.00 per hour. Instead of measuring the size of the abstracts to be translated, I propose that the entire patent serve as the basis for establishing the rate of remuneration, since only with the entire patent can the translator begin to enfold his function of control and avoidance.

In the context of patents, control is defined here as the translator's choice of terminology that is most appropriate and that forms the knowledge available to the patent examiners in their examination of patent applications. This definition is not meant to limit the translator's impact on the efficiency and effectiveness of the USPTO and, therefore, does not preclude the broader sense, such as the control of the knowledge that serves as the documentation in the results of the patent examination and/or the control of the information used by the Board of Patent Appeals and Interferences in its adjudication of patent appeal. Whenever the USPTO has to refer to the English translation, the translator's terminology will play a vital role in the decision-making.

In this respect the translator may be regarded as the USPTO's partner in the process of examining patent applications and, hence, exhibits many similarities with the examiner. For example, when a translation comes to the vendor it is assigned a number, categorized according to language and assigned to a translator, who will be responsible for the translation. Similarly when an application for a patent arrives at the USPTO, it is assigned a serial number, classified and then assigned to an examiner, who will be responsible for examining the application. Like the translator, examiners work at a piece rate, receiving credits only for the number of applications they have examined. These credits in turn are derived from the number of expected hours it would take the examiner to complete the examination of a patent application.

Yet these superficial features fade into the background as soon as one realizes the enormous magnitude of the task that the translator and the USPTO examiner face when confronted with over 470 broad technology categories or classes and over 150,000 specific subcategories or subclasses, to which a patent may be assigned, and the subsequent rewards they receive for their performance. The procedure at the USPTO stipulates that the examiner will be selected from a very large pool of potential examiners exhibiting a wide range of expertise. At the vendor, however, the selection is made from a very small group of translators according to the language pair; and no distinction is made with respect to the class of technology. Therefore, a German to English translator will be responsible for any German language abstract in any field of technology. The translator's handicaps become even more apparent when a comparison is made between the examiner's education and experience and that of the translator. Whereas the examiner must show a minimum educational requirement of a B.A. in computer science, physical science or engineering and some engineering experience or graduate study, this technical background is optional for the translator. On the other hand, translators for this kind of work are highly educated linguists with more often than not an M.A. and 5 to 7 years of experience in technical translations; the examiner needs only 2 weeks of training to start examining patents. Nevertheless, this junior examiner receives a guaranteed entry level salary ranging from $35,000 to $76,000 and after 4 or 5 years of experience a salary ranging from $110,000 to $143,000, not including bonuses. In other words, the USPTO employees are rewarded in proportion to their signatory power. Yet, his resource person - the freelance translator -, who bears the sole responsibility for the quality of the examiner's database, always works - from the beginning of his career to the end - at the rate established by the vendor for any particular job. There is usually no negotiation: the only choice is to accept and, thus, work or to reject and not to work. Moreover, even if the translator produces the highest quality translation in a proper and timely manner, it still does not guarantee when and/or whether the translator will be paid. The only certainty that translators have is that there will be no reimbursement for their expenses, nor will they be offered any incentives or bonuses.

Instead, the translators are severely penalized whenever they strive for greater accuracy and the highest level of quality demanded by the respective patent office. To illustrate this point, let us assume that a USPTO examiner is looking for prior art references concerning oil well drilling. As stated above, the examiner is a researcher and processor of knowledge; the translator is the person who controls the terminology and, hence, the knowledge available to the examiner. Therefore, the outcome of the examiner's perusal of foreign language abstracts on oil well drilling will be a function of whether the translator had the time and expertise to render a proper translation of the abstracts within this field of technology. Will, for example, the examiner recognize the significance of the DE 33832715 A1 - Vorrichtung zur Herstellung einer Ablenkbohrung - from the English title "Arrangement for Producing a Deflecting Bore", which is certainly a correct and plausible translation, given nothing more than - say - Ernst's Dictionary of Engineering and Technology. To have gone beyond the minimum expectation, to have researched the field via out-of-print dictionaries and the Internet and/or to have conferred with specialists and, hence, to have facilitated the examiner's work with a translation that clearly indicates the exact object of the invention - i.e., A Device for Drilling a Deviated Well - would have cost the translator time and effort that could have been spent more profitably churning out another abstract. Furthermore, the translator will lose the goodwill of the vendor and perhaps even the contract, because he is not producing English abstracts as fast as the other translators. And more importantly for the translator, this generalist approach has the insidious secondary effect of becoming such a habit that, when called upon to translate technical documents that are difficult in content, the translator will no longer be able to perform in terms of avoidability, i.e. avoidance of improper vocabulary, etc. Therefore, the obvious solution is not to discourage the translator from doing research, but to encourage and enable the translator to unfold his translation capability by supplying the entire patent and not just for reference, but as the basis for compensating the translator for the abstract. Only with the complete patent can translators begin to satisfy this second requirement of avoidability.

It is quite likely that the USPTO understood the translator's function of controlability and avoidability in terms of avoiding errors, inconsistencies and/or improper vocabulary when it provided the entire patent, of which the abstract was to be translated, so that the translator could meet or even exceed the translation capability requirement. Moreover, the vendor must have also understood - at least initially - that a technically accurate translation is dependent on the avoidability function of the translator when it, too, passed the entire patent along to the translators/editors for reference. In other words, both the USPTO and the vendor stipulated that the initial sample abstract was to be translated in the context of the patent to which it relates. Under these circumstances the abstract (in German Zusammenfassung) is more than a summary; it is the leitmotif or the thread that runs through the entire patent, beginning with the specification and ending with the abstract itself and must be treated in any translation as the culmination of what is claimed.

This fluid connection between the abstract and the rest of the patent is reflected in the structure of any patent document. It is essentially the object of the invention and the solution, as the following abstract of the DE 10 2004 028 017 A1 - Zugmittelanordnung - demonstrates.

Abstract:

Belt assembly, in particular for driving camshafts of an internal combustion engine, comprising a first and a second guiding device for the belt. The first guiding device is assigned a device for tensioning the belt; and the first guiding device is connected to the second guiding device so as to transmit a force and/or motion.

[original German abstract:

Zugmittelanordnung, insbesondere zum Antrieb von Nockenwellen einer Brennkraftmaschine, umfassend eine erste und eine zweite Führungseinrichtung für das Zugmittel, wobei der ersten Führungseinrichtung eine Einrichtung zum Spannen des Zugmittels zugeordnet ist und die erste mit der zweiten Führungseinrichtung kraft- und/oder bewegungsübertragend verbunden ist.]

The first sentence of the above abstract appears word for word in the first sentence of the description (paragraph 001).

The invention relates to a belt assembly, in particular for driving camshafts of an internal combustion engine, comprising a first and a second guiding device for the belt.

The second sentence of the above abstract is repeated as the solution to the problem (Lösung der Aufgabe) in paragraph 006:

... the first guiding device is assigned a device for tensioning the belt; and the first guiding device is connected to the second guiding device so as to transmit a force and/or motion.

Both the first and the second sentence of the abstract appear again in the section disclosing the patent claims, where it forms the first independent claim.

Belt assembly (100), in particular for driving camshafts (104, 106) of an internal combustion engine, comprising a first and a second guiding device (114, 116) for the belt, the first guiding device (114) being assigned a device (122) for tensioning the belt (108); and the first guiding device (116) being connected to the second guiding device (116) so as to transmit a force and/or motion.

Even the drawing, which does not show any text, may also be regarded as reflecting the abstract, since the numbers 100, 104, 106, 108, 114, 116, etc. refer to what is claimed. Like the pattern of a carpet, the specification, the claims, the drawings, and the abstract form a monolithic unit, which rapidly loses its meaning when it is unraveled or rather in the context of the translator's job dissected and discarded as non-essential ballast.

The significance of the entire patent with respect to avoidability and controlability may also be demonstrated in a practical way. That is, in the normal course of a translation, the patent translator does not start with the abstract, but with the name of the inventor and/or the applicant (here Tryphonos, Andreos and Bayerische Motoren Werke) and the title (Belt Assembly) in order to lay a foundation for the patent in question. In this case Tryphonos and BMW are also named as the inventor and applicant respectively in WO 2005/121599 A1, a patent that discloses a belt assembly with a tensioning device, and in WO 2005/121600 A1, a patent that discloses a traction mechanism. Both of these patents include an English abstract that can be used to prepare a glossary of terms that apply specifically to the patent in question. Secondly the translator looks at the prior art references (DE 411 49 48 A1 and DE 201 02 748 U1) for additional vocabulary and, more importantly, also for a better understanding of the invention under discussion. With the aid of this glossary the next logical step is to translate the description of the device with reference to the drawings for shapes and orientation as well as for a greater comprehension of the operating mode of the device. After checking the translation for accuracy of the terminology and consistency, the translator then moves to the claims, which, as stated above, are often identical to the abstract. What is left is the first half of the patent - that is, the introductory part, the background art, the object of the invention and the solution of the problem - which is either embodied in the abstract or has already been researched. Such as inversion of the translation process helps the translator to avoid errors, inconsistencies and improper vocabulary and to control the terminology, thus ensuring the highest quality of translation.

In light of the time and expense that is involved in acquiring adequate resources to satisfy the controlability and avoidability requirement and the relationship between the patent and the abstract, it would only be logical to establish the price for the abstract at the final word count of the patent itself.