Thursday, June 4, 2015

PTAB finds Applicant arguments about misleading terms in translated reference were not supported by evidence

Takeaway: An Applicant argued on appeal that a German patent reference didn't disclose the claim limitation "helically extending" despite the presence of "setscrew," "grub screw," and "screwed in" in the English translation. According to the Applicant, the terminology used in the reference could also be used to describe a twist-and-lock structure which did not include helically wound threads. The Applicant also pointed to the lack of a clear illustration of helical threads, the fact that the drawing was "poorly described" with only a paragraph of text, and the fact that the reference was translated from German to English. The Board affirmed the Examiner's rejection. The Applicant's arguments about the drawings were contradicted by the express teachings of the text. The arguments about the terms used by Schraber were presented as attorney argument and were not supported by evidence. The Board cited In re Geisler, 116 F.3d 1465, 1471 (Fed. Cir. 1997) for the proposition that "argument by counsel cannot take the place of evidence)." (Ex parte Jackson, PTAB 2012.)

Details:

Ex parte Jackson
Appeal 2010-011356; Appl. No. 10/784,066; Tech. Center 3700
Decided:  March 16, 2012

The application on appeal was directed to a bone screw used in spinal surgery. A portion of a representative claim on appeal read:
     1. A closure for setting engagement with a structural member and comprising:
     (a) a substantially cylindrical body having an outer cylindrical surface ...
     (b) a guide and advancement flange extending helically about said outer cylindrical surface ....
     (c) at least one of said leading surface or said trailing surface [of the flange] being compound in contour and including an inward anti-splay surface component facing generally toward said closure axis...
Claim 1 was rejected as obvious over Schafer (a German patent) in view of Jackson (same inventor as application on appeal). The Applicant identified Schafer in an IDS, and provided a full translation with the IDS.

The main dispute in prosecution revolved around the teachings of Schafer with respect to the claim limitation "a guide and advancement flange extending helically." The Examiner took the position that this feature was disclosed in Schafer's FIG. 2 (below).

Three lines of argument developed during prosecution. One argument focused on the meaning of "helically wound" (although this precise language wasn't in the claim). A second argument focused on what a POSITA would understand from Schafer's drawings. The third argument related to Schafer's textual description, as translated into English from the original German.

In the first response, the Applicant explained that the perspective view of Schafer's FIG. 2 made it "hard to determine if it has a pitch of any kind," but "comparing the flange shaped openings in comparative spacing from the edge of the seat, indicates that there is no conventional thread positive pitch present on the flanks of the structure" (emphasis added). The Applicant concluded that Schafer's flange was a "twist screw in and lock structure" rather than a helical structure.

Although the Applicant's initial response didn't explicitly discuss claim construction, the Examiner addressed claim construction in the next Office Action. The Examiner explained that "any angular movement in the structure satisfies 'helically wound' as the claims don't set forth a minimum amount of winding." Further, "pitch need not be 'conventional' to be present and 'conventional' is not recited in the claims."

In a Response to the second Office Action, the Applicant argued that even though pitch wasn't in the claims, the phrase "helically wound" nonetheless implied something about pitch. The Applicant used an annotated Schafer FIG. 2 (below) to explain that Schafer's flange would be helically wound only if D2 > D1 and D4 > D3. But since these conditions did not hold, Schafer's flange was not helically wound.



The Applicant also discussed the text description of FIG. 2 that appeared in the of-record translation of Schafer (supplied earlier by the Applicant with an IDS), which read:
In the exemplary embodiment shown in Figure 2, the bifurcated head 4 of the bone screw 1 likewise has a thread, which, however, has a top flank 11 and a bottom flank 10 embodied in a stepped fashion. The shoulder of the bottom flank 10 is shaped such that it forms an undercut 17. This undercut 17, particularly by means of the shoulder 18, prevents the legs 5 from being bent radially outward while the grub screw 3 is being screwed in. A positive lock is thus produced in the radial direction between the bifurcated head 4 and grub screw 3. This positive lock prevents, as previously mentioned, any slippage of the leg 5.
(Emphasis added.)
Based on the drawing and this text, the Applicant urged that FIG. 2 was a "twist and lock" structure in which "projections are screwed or rotated ninety degrees into the receiver," rather than a "structure that will helically rotate into the receiver" as claimed.

The Examiner maintained the rejection in the next Office Action. The Examiner explicitly discussed claim interpretation: "helically wound just means it progresses in a spiral or helical fashion, i.e. is threaded (as opposed to 'twist-lock' designs)." The Schafer text actually relied on by the Applicant "teaches not only that the device has threads but that the anti-splay grub screw is screwed in" (emphasis in original). Thus, "it is quite clear that Schafer discloses a 'threaded' structure (i.e., a helically wound flange structure)."

The Applicant filed one more response before appeal and addressed the Examiner's latest contentions.
According to the Applicant,
There is nothing that requires a "grub screw," as it is translated from German, to be helically wound. Such a screw can have mating portions that rotate, for example, 90°, into each other and such are well known. It is believed that this is what is shown in the German [Schafer] reference.
When the Examiner maintained the rejection, the Applicant appealed. In the Appeal Brief, the Applicant noted that FIG. 2 was "only a partial cross section of the receiver" and the corresponding text was limited to "a single paragraph." The Applicant characterized the embodiment of FIG. 2 as "poorly described" and "apparently added as a quick afterthought." According to the Applicant, a POSITA would see in FIG. 2 that all 7 tiers of the receiver were equally spaced from front to bottom on front and back and essentially horizontal, i.e., did not have a pitch. This was contrary to the meaning of "helically wound," which implied that each tier has a pitch going from one tier to the next with each 360° revolution.

The Applicant then made this point by including FIG. 2 next to a FIG. 2 modified to be helically wound). The juxtaposition allowed a POSITA to "immediately see" that Schafer's FIG. 2 was a twist-and-lock structure rather than a helically wound structure.



Moving from drawings to text, the Applicant acknowledged that the German Schafer patent, as translated, referred to a grub screw "screwed into" a thread in describing FIG. 2. The Applicant also briefly acknowledged that Schafer used the phrase "screwed in" in describing the embodiment of FIG. 1. But according to the Applicant, this terminology wasn't dispositive:
[T]he term screw can mean simple rotation into something. ... A structure that mates upon rotating 90° can be called a grub screw and rotation or twisting can be interpreted and translated from the German into English as being "screwed in" in the Schafer disclosure to distinguish from the closures that are slid in sideways without rotating.
A POSITA looking at Schafer, as a whole and in context, would understand that the FIG. 2 receiver was not helically wound. Furthermore, a POSITA would also understand that Schafer's "closure and receiver are aligned and the closure twisted or rotated 90° to secure them together." 

In the Answer, the Examiner first reiterated that the Schafer translation relied on by the Applicant used the terms "threads and "screw." The Examiner then referred to a full translation provided with the Answer, and pointed to additional mentions of "threading" and "screwing in the setscrew." (This document was identified as "Full translation of Schafer (DE 298 10 798 U1), Schreiber Translations, Inc., May 2010.")

Moving on to address the Appeal Brief arguments, the Examiner asserted that the Appellant's "opinion" that Schafer disclosed a non-threaded design was "irrelevant since it overlooks or misconstrues what the reference explicitly says, namely, that the design shown in Fig. 2 includes a threaded set screw that is screwed in." (Emphasis added.) And although the closure itself wasn't illustrated, the text of Schafer said that the "closure is 'setscrew 3' and that it forms a friction-lock with the undercut 17 and lug 18 of the thread of the head 4 (translation, page 10, lines 5-7)."

The Examiner also criticized the Applicant's "exaggerated" side by side drawings: 
[G]iven the very small portion of the entire circumference of the threads that is actually shown, the purported pitch imagined by Appellant is unnecessary for a helical thread and implausible in a single-start thread design (only one helical thread ridge). The angle that is necessary for the threads of Fig. 2 would not be extreme (as in Appellant's sketch) in the very small portion of the circumference that is shown in the basic drawing of Fig. 2.
In the appeal decision, the Board found that Schafer did disclose the limitation at issue. The Board first addressed the text of Schafer, and its use of terms such as "screw" and "screwed in." The Board acknowledged Applicant's argument that the context of Schafer would lead a POSITA to understand these terms differently from normal English usage. However, the Board agreed with the Examiner's reading of Schafer and further noted that "argument by counsel cannot take the place of evidence, citing In re Geisler, 116 F.3d 1465, 1471 (Fed. Cir. 1997).

Turning to the Applicant's arguments about Schafer's drawing, the Board first found these arguments ran counter to Schafer's express teachings in the text (i.e., "grub screw," "screwed in"). Moreover, the Board was not persuaded by the Applicant's side-by-side comparison. Instead, the Board agreed with the Examiner that "given the 'very small portion of the entire circumference of the threads' illustrated in Schäfer‟s figure 2, a conclusion cannot be drawn as to the pitch of the threads."

Finally, the Board considered an argument about the secondary reference and found it not persuasive. As a result, the Board affirmed the § 103 rejection.

My two cents: There are several lines of argument at play here.
  • Inferences made from a drawing. 
  • Overcoming literal teachings of a reference by arguing that the reference really doesn't mean what it says. 
  • Attorney argument.
  • Translation errors.
Considering the point of whether the FIG. 2 drawing alone showed a helically wound thread, I think the Board got this one wrong. According to the Board, the Applicant made unsupported inferences about thread pitch. Maybe so, and maybe the Applicant overreached with the side-by-side drawing. But the Examiner didn't convincingly show that the FIG. 2 drawing itself illustrated helical winding.  So I think the Applicant won that round.

Even so, the linchpin of the Applicant's argument was overcoming the literal teachings of a reference. Words like "screwed in" and "setscrew" and "grub screw" strongly suggested helically wound threads. The Applicant fought hard to convince the Board that, despite these terms, Schafer's closure structure actually twisted or rotated into alignment with the receiver, and this twist-lock behavior did not require helical threads.

I thought the Applicant did a good job of marshaling facts to put together an argument against ordinary meaning. But in the end, this amounted to nothing but attorney argument. The Applicant should have entered evidence in the form of definitions, technical articles, or an expert declaration. I see no bright line rule for when it's appropriate for the attorney to expound upon the teachings of the reference, and when evidence is needed. But when the core of the argument is what a term means, I think you need evidence.

Finally, the Applicant may have held back the strongest argument. To me, the Appeal Brief statement "rotation or twisting can be interpreted and translated from the German into English as being 'screwed in' " was a hint that the translation from German to English was in error. But the Applicant didn't make this point explicit. And this making this argument would certainly require evidence.

If you wanted to argue translation error, what sort of points would you make and what sort of evidence would you submit? Are machine translations more vulnerable here?

19 comments:

  1. "If you wanted to argue translation error, what sort of points would you make and what sort of evidence would you submit? Are machine translations more vulnerable here?"

    In this case I don't think there is any evidence that the attorney could put on the record that would help him (which is likely why he didn't). A madenschraube in German translates to grub screw/set screw in English and there is little room for a differing translation. I would have liked to see the examiner find an English-German dictionary and put the page with madenschraube on the record to ensure the Board reached the correct decision, but the same could be said for the attorney and how he should have avoided the appeal. This is one of those cases where you wonder why the attorney didn't save the client money and get a translation if they thought there could be a translation issue. German-English translators charge $35-$45 an hour, and when the translation comes back confirming the machine translation and examiner's argument the client would have saved the money on the appeal. The examiner had made the prima facie case with the machine translation, and the onus was on the applicant to show there was a translation error.

    As an aside, since English is a germanic language modern machine translations are very good except from idioms.

    ReplyDelete
    Replies
    1. >why attorney didn't save the client money and get translation
      >if they thought there could be a translation issue.

      To be fair, the Applicant never did explicitly argue there was a translation error. I read one of the (extensive) remarks to suggest that.

      But I did start thinking about translation error when writing this post, and wanted to hear what others had to say. Thanks for contributing.

      >by confirming the machine translation and examiner's argument
      >client would have saved the money on the appeal

      Nah, cuz the Applicant had all these other arguments unrelated to translation. One, figures alone didn't show helical winding. Two, "screw" can mean rotate or twist and doesn't necessarily imply helical winding. Three, some of Schraber's written description would be understood as twist-and-lock rather than helically wound.
      And another argument I omitted here -- using other patents by Schaber to argue that FIG. 2 of the reference used on appeal was twist-and-lock. (Thinking about covering that last onein another post.)

      >since English is a germanic language modern machine translations
      >are very good except from idioms.

      Yep, that makes sense.

      Delete
  2. MaxDrei says thanks for an interesting item. One can easily imagine motion along a helical path in a twist lock assembly that lacks any helical surface whatsoever. So no wonder Applicant fought. Problem is though, that a German Schraube dictates a helical surface.

    ReplyDelete
  3. I think the issue of which translation is correct misses the point. The issue is what would a POSITA construe a "Madenschraube" to include. So what the Applicant needs is a declaration/affidavit from a practicing German mechanical engineer, who's also fluent in English, to state on the record what a German mechanical engineer would consider a Madenschraube to include/not include. The problem with many technical terms is they often don't translate well, and that includes British English to American English.

    ReplyDelete
    Replies
    1. Thanks for your contribution. Yep, once you frame the issue as what a POSITA understands a term to mean, that is best addressed by an expert declaration.

      Delete
  4. MaxDrei here again.

    Good comment from that anon, yesterday evening. A difference between practice in chem/bio and that in mech eng. is that the all chemists use the same terminology world-wide, whereas, in mech eng, each applicant uses their own chosen terminology, peculiar to them. The terminology they deliberately and thoughtfully chose inevitably comes with its own connotations, which themselves vary from country to country, This in turn renders so difficult, or iposible, any attempt to translate that terminology 1:1 into another language.

    Which makes inter partes mech eng proceedings at the EPO so much fun, given that the EPO has three Office languages English, French and German, and that in an opposition with more than on Opponent, all three languages can be in play, in the decisive oral proceedings.

    No surprise then, that Europe has little time for "lawyers' games" with words, and will always cut to he technological heart of the matter.

    But suppose Karen's case had been instead under Article 54(3) EPC (the one that excludes a reference from obviousness attacks), The decisive and ultimate quesion would then be: does Schafer deprive the claim of novelty?

    Tricky. Excellent quesion in fact. This case might be a good one for training Eropean patent attorneys. Thanks Karen.

    ReplyDelete
    Replies
    1. >all chemists use the same terminology world-wide
      >in mech eng, each applicant uses their own chosen terminology,
      >peculiar to them.

      Just to clarify ... are you saying that mechanical terms differ a lot by country or language (where chemical terms do not)?

      Or are you saying that mechanical terms vary a lot by Applicant even within a particular country or language?

      I read as you saying the latter. If so, I won't disagree -- I simply don't practice enough in this area to make this call, although I do read plenty of file histories for mechanical apps.

      I will say that I see *far* more commonality in mechanical terms than I do for my field -- electronics/software. In my world of CS/EE cases, there is some commonality in basic building blocks -- terms like "encoder" and "rectifier" come to mine -- but outside of that .... virtually none.

      Delete
    2. >[hypothetical] question under Article 54(3) EPC
      >(the one that excludes a reference from obviousness attacks),
      >would then be: does Schafer deprive the claim of novelty?

      I think you're trying to simplify the issue by making it novelty, instead of obviousness. Right? Or maybe I'm missing something, I don't know Article 54(3) EPC very well :-)

      I think I essentially made such a simplification in my post. As did the Examiner, Applicant, and Board. That is, the narrower sub-issue was "did Schraeber disclose helical winding". Sometimes the Board hedges by framing a sub-issue as "the reference SUGGESTS," which I read as meaning not-close-enough-for-102-but-enough-for-103.

      But not here ... I read the Board in Ex parte Jackson as saying Schraber did not anticipate the helical limitation.

      Delete
    3. Thanks Karen. Yes, the reference to 54(3) is my way to confine the issue to novelty alone, and exclude any wandering off into the swamp of obviousness.

      With my "engineering" I intended everything except chem/bio. Different engineering Applicants use different terminology, and the terminology varies with language and country. You can see this in, for example, metallurgical cases originating in Japan and translated there into English for PCT filing.

      The reference has an umlaut over the "a". So it is "Schaefer". Translated into English, that's "Shepherd". A "German Shepherd" (or in England an "Alsatian") is a "Shaeferhund". So no problem with translation there then, even though an "Alsatian" out of context would be a person from Alsace. But wait a minute. Alsace is in France, isn't it?

      Delete
    4. Alsace is in France, isn't it?
      It is now. However, it wasn't always.

      Regardless, to answer Karen's question:
      Just to clarify ... are you saying that mechanical terms differ a lot by country or language (where chemical terms do not)?
      I would agree with the original statement -- for obvious reasons.

      Delete
  5. I see from your report that, immediately after spouting some tosh about what the reference, Applicant uses the obiquitous text "It is believed that this is what is shown in " Schaefer.

    Karen, I practise in Europe. Do me a favour, will you. Tell me, is there any downside (apart from a severe loss of credibility) arising from using on the USPTO written record the passive voice and the notion of "belief". I mean, when PTO Examiners see the words "it is believed that" are they thereby sensitised, and looking out for, the recitation of nonsense. You know, as in "It is believed that, here, "black" means "white" and so on.

    ReplyDelete
    Replies
    1. >any downside (apart from severe loss of credibility) arising from using
      >on written record the passive voice and the notion of "belief".

      Nothing that comes to mind. I mean, there *is* a downside -- but you identified it: loss of credibility.

      I think the root for this passive phrasing is an attempt to avoid "mischaracterizing the reference." And they're afraid of mischaracterizing because they're worried it will be viewed as Inequitable Conduct.

      >they are thereby sensitised, and looking out for, the recitation of nonsense.

      I don't see it that way. I don't think that "believed that" signals nonsense any more than any other introductory phrase, or even lack of an introductory phrase. I'd say "believed that" signals hedging, but not nonsense.

      BTW, I think many Examiners absolutely ignore *ALL* the scaffolding we attorneys use -- references to Applicant or attorney or undersigned, positioning such as "submit" and "maintain", hedging such as "appears to" or "believed" . I think Examiners learn to quickly filter out all of the extra verbiage and zero in on the heart of the argument.

      Prosecution is challenging because we write for more than one audience. The Appeal Brief goes to the Examiner and the Board, which are two very different audiences. We also think about two other *potential* audiences: judge and opposing counsel.

      Delete
  6. That's a very thoughtful, herlpful, cheering and illuminating answer. Thanks Karen

    ReplyDelete
  7. German to English patent translator responding here.

    "In this case I don't think there is any evidence that the attorney could put on the record that would help him (which is likely why he didn't). A madenschraube in German translates to grub screw/set screw in English and there is little room for a differing translation."

    This pretty much hits the nail on the head. There's no credible expert opinion or dictionary in the world that would state that a Madenschraube is anything other than a grub screw or a set screw. Which is, of course, why the Applicant didn't/couldn't present one.

    "As an aside, since English is a germanic language modern machine translations are very good except from idioms."
    I may be a bit biased here, but this is anything but true. Machine translation has come a long, long way, but germanic or not, the highly complex structure of German makes machine translations unusable for anything other than basic gisting. Anyway, it's irrelevant here - the English rendering of the patent wasn't a machine translation. I'm personally familiar with the firm that did the translation in this specific case, and I can tell you that they do excellent work.

    "Problem is though, that a German Schraube dictates a helical surface."

    Yes. "Schrauben" virtually MUST refer to helical winding.

    "Just to clarify ... are you saying that mechanical terms differ a lot by country or language (where chemical terms do not)? Or are you saying that mechanical terms vary a lot by Applicant even within a particular country or language?"
    From my perspective both are true. I work with both chemical and mechanical patents and I can tell you that chemical terms are almost entirely consistent, both internally within the language and in translation. Makes my job a lot easier. In the case of mechanical terminology, things are a great deal more flexible, both by Applicant, within a language, a region, sometimes even a company (a certain auto manufacturer I do a good bit of work for is notorious in this regard). As a translator, there are many, many more judgement calls when it comes to mechanical patents as opposed to chemical patents. It's just that in this particular case, sometimes a grub screw really IS just a grub screw. :)

    ReplyDelete
  8. Wow, it's great to get so many comments from folks that are really in the know.

    While I was initially drawn to this case because of the foreign language reference, sounds like translation error was not an issue in this case.

    I still find this case interesting because the Applicant tried so hard to overcome the plain teachings of the reference. I wonder if the Applicant would have gotten further with declaration evidence making some of the same points ... that the reference, in a sense, didn't mean what it said on its face.

    And surely we've all been there ... the reference says something that's misleading, or only part of the whole story, or is in some cases, just plain wrong.

    ReplyDelete
  9. A Declaration, to get past the PTO objections? You set me (MaxDrei) thinking Karen, on one difference in practice between the USPTO and the EPO.

    At the EPO, the Examiners are expected to use their own specialist and high level technical credentials to pronounce on what teaching the notional "person skilled in the art" would "derive" from any given document. So filing a Declaration, to instruct the Examiner what the skilled person would have derived, as a way to advance prosecution, is worse than useless, counter-productive. In a civil law jurisdiction, where there is no possibility to cross-examine the Declarant, the Office (or Court) will simply suppose that the Declarant is a good servant, mindlessly parroting what he is being paid by his master to aver. So Declarations carry very little, or zero, weight.

    Attorney argument alone though. Now you're talking. That can be very persuasive indeed.

    Funny, isn't it?

    .

    ReplyDelete
    Replies
    1. Thanks, Max, for your explanation of EPO procedure.

      >EPO Examiners ... use their own ... credentials to say what the notional
      >"person skilled in the art" would "derive" from any given document

      See, that's exactly what's unfair about US practice. The *Examiner* is allowed to opine about what the references mean to a POSITA, about what terms mean to a POSITA, and about how/why a POSITA would combine.

      Yet when making what are essentially technical arguments, the Applicant's attorney must stick close to the literal teachings of the documents already in the record. Straying too far risks the argument being ignored as "mere attorney argument".

      OK, the Examiner's leeway is cabined somewhat, because his statements must be based on / tied to evidence (references). But still, the system gives the Examiner a lot of leeway to interpret those docs.

      Double standard.

      >mindlessly parroting what [Declarant] is being paid by his master to aver.
      >So Declarations carry very little, or zero, weight.

      They carry little weight with some Examiners, presumably for precisely this reason.

      >Attorney argument alone though. That can be very persuasive

      Perhaps the policy behind discarding "mere attorney argument" is that we can't be sure the practitioner does understand the technology?

      US patent practitioners are required have a technical background to sit for the patent bar. But once a registration number is obtained, the practitioner isn't limited to working on applications in particular technologies.

      Is this true in Europe too?

      Examiners, OTOH, are assigned to work in a particular art unit. Though from what I understand, they aren't *required* to have a technical background in that unit. (Though in practice the Examiner's background often *does* match the unit he works in.)

      Delete
  10. OK Karen, MaxDrei here. Off we go.

    The EPO is getting more like the USPTO. In the past, Examiners were put to examine only in technologies in which they are expert. That's a thing of the past. Quite why, I'm not sure. EPO management "efficiency" and need to hit "targets" I believe.

    In Europe, there are two distinct professions, attorney at law and patent attorney. In the courts in Germany, they wear different black robes, to distinguish the two professions. The patent attorney speaks on content, the attorney at law on procedural matters. This might be one reason why patent attorneys at the EPO are expected not to stray into technical fields where their knowledge and experience is inadequate, and why clients expect their patent attorney to be an expert in the science and technology of their invention. If the patent attorney cannot offer science and technology expertise, how else to justify the separate patent attorney profession?

    Your answer shows how things are much the same, the world over, despite protestations to the contrary. What I was going after though was the difference between common law "preponderance of evidence" and civil law "balance of probability". in the USA, you need "evidence" Civil law judges over here can manage quite well without.

    ReplyDelete
    Replies
    1. >I was going after difference between common law
      >"preponderance of evidence" and civil law "balance of
      >probability". in the USA, you need "evidence" Civil law judges
      >over here can manage quite well without.

      Ah, gotcha. Thanks for contributing to the comment thread.

      Delete