Priority Claims in US Patents

By Dennis Crouch

Patentees are continuing to move away from the single-patent model toward a model that involves multiple patents in multiple jurisdictions. Today, over 75% of patent applications rely upon at least one prior patent filing as a priority document. I expect that figure to grow based upon incentives coming from the recently implemented first-to-file regime. Over the past 10 years, most of the change has been in increased PCT filing and increased use of US provisional patent applications. In the US, the largest users of the provisional patent applications are some of the largest strategic users of the patent system, including Qualcomm, LG, Broadcom, Marvell, Google, Apple, Samsung, Univ of California, and Texas Instruments. More than 40% of patents with a US assignee (or US inventor owned) claim priority to a provisional application. Expect that figure to rise as well.

39 thoughts on “Priority Claims in US Patents

  1. 3

    I see too many deficient provisional applications, either prepared by the inventor or by patent practitioners on the cheap.

    The filing date of the provisional may be needed (maybe to avoid the inventor’s own activities) and I anticipate that this issue will become common in the not too distant future.

    1. 3.1

      The securing of an “effective filing date” requires that the disclosure meet 112, a description in such detail that one of ordinary skill in the art can make the invention. Practitioners who prepare provisionals “on the cheap” should be sanctioned.

      This behavior is not unlike the colleague who, many years ago, told me that when a client cannot afford to pay entire fee for the preparation of a utility patent, the colleague would charge only the amount the client said he could afford.

      I asked does that mean you provide the same service but agree to take less. Oh, no, he responded. If, for example, the client could only pay 1/2 half the fee, that’s all the client got in the patent application!

      1. 3.1.1

        Philip I get what you are saying, and any filing by a licensed practitioner that includes a claim to a prior filing that cannot meet the law’s requirement is something that practitioners routinely ‘get away with’ because there is no sanction for such a filing. Due diligence is not measured, and you do tend to get what you measure.

        It is rather similar to the lack of Rule 11 strictness, which would abate some of the “Troll” mania (a far lesser intrusion on property rights that is currently available).

      2. 3.1.2

        And I would further add that the proper avenue for going after the “bottom-feeding unethical lawyers” (or any other choice descriptions routinely widely broadcast by the RQ/HD), is correctly done under a different avenue than attacking the patent right itself.


          Let us know when you see an example of an attorney/troll who should be sanctioned under Rule 11. Preferably, let us know before the sanctions are actually handed down.


            Another reply (while correct) was too ‘brash’ and was, um, moderated…

            …so let’s see if this reply will meet this unknown ‘objective’ standard:

            Dear Malcolm, please note that I actually wrote “It is rather similar to the lack of Rule 11 strictness

            Your reply of “let us know” seems to misconstrue what I wrote. Believe me, if I ever find an example of anyone (no need to use the denigration) who should be sanctioned and such has not already been revealed for ‘our’ edification, I will let ‘us’ know just as soon as possible. I am not sure what you mean by the preference before sanctions are actually handed down, but it seems completely disassociated from the discussion and seems safe to simply ignore.


              Believe me, if I ever find an example of anyone (no need to use the denigration) who should be sanctioned and such has not already been revealed for ‘our’ edification, I will let ‘us’ know just as soon as possible.

              I look forward to that. You indicated that Rule 11 is applied more laxly than it should be, so there should be plenty of opportunities for you to let us know what the proper application should look like in the year ahead.


              Malcolm still trolls with “there should be plenty of opportunities for you to let us know what the proper application should look like in the year ahead.

              No need for such particularity Malcolm. My position on this point is crystal clear. You seem confused and want some sort of additional clarity.

              Or maybe you are just trying to say something else….? Something that has no support…? Something buried in the innuendo that you say you do not use…?

              (sigh) You remain a transparent t001.

  2. 2

    Mostly off-topic: This case deals with a priority claim, but I thought the more interesting aspect was the discussion of the claim term “receiver” – the challenger argued that it was “purely functional,” but the court disagreed. I think Prost’s opinion on this topic is highly relevant to some of the controversy on this blog.

    link to

    Prost pretty clearly (and persuasively, in my view) elaborates on the point that NWPA has made several times:

    We have stated previously that just because “the disputed term is not limited to a single structure does not disqualify it as a corresponding structure, as long as the class of structures is identifiable by a person of ordinary skill in the art. Linear Tech. Corp. v. Impala Linear Corp., 379 F.3d 1311, 1322 (Fed. Cir. 2004); see also Power Integrations, 711 F.3d at 1365 (“[W]e require only that the claim term be used in common parlance or by ordinarily skilled artisans to designate sufficiently definite structure, even if the term covers a broad class of structures.”) (internal citation omitted). Given the demonstrated familiarity that one of skill in the art would have with a “receiver” and the Board’s own factual finding that a skilled worker would know what a “receiver” is, we hold that in this case the term is not the “black box that performs a recited function” that Face would have us believe it is.

    1. 2.1

      I don’t think the point is what you make it to mean, but let’s clarify before I jump in. What point is that? What’s the legal theory being advanced?

      1. 2.1.1

        Sorry, RG, you’re right that my point was not at all clear. I was thinking of the debate about claims that include phrases like “processor configured to X…”

      2. 2.1.2

        To be even more clear, several people have offered opinions to the effect that a “processor configured to …” is at best an m-p-f limitation. NWPA, in his inimitable style, has argued that this type of limitation should be interpreted to encompass a broad range of structures, “configured” according to techniques that are well-known to the skilled person. I think Judge Prost’s opinion here supports his position, although perhaps not to the extreme that he necessarily wants to take it.


          To be even more clear, nobody here (that I’m aware of) really cares much about the use of such terms as “receiver” when they are not at the point of novelty or otherwise critical to the patentability of the claim. In many (most? all?) cases the “invention” could have been described without reference to “the receiver” and the issue would never even come up.

          Was there a particular function ascribed to that receiver that was relied upon to distinguish the invention from the prior art? That’s a different story.

          Additionally (as you seem to be aware, LB), the difference between a skilled artisans understanding of the structure of a receiver in this context is rather different than the skilled artisans understanding of the structure of “determining whether the appropriate ad is targeted to the user.”


            MM, you get it. It is not apparent that LB does.

            The key aspect of the case was that receiver was a known structure. One did not have to described in the specification any details about that known structure. The court even cited a 1881 Supreme Court case to support this point.

            The point of novelty and nothing to do with the functional aspects of receiver. If that were the case, the result should have been different.


              It is not apparent that LB does.,

              LB’s a smart guy. I think he’s just throwin NWPA a bone to chew on.


          Okay, I’ll jump in, since few people on this board seem to understand the concept:

          Morse, Gen Elec v. Wabash, Halliburton – they all say the same thing, which is that a naked claim to a function as the novel means of distinguishing your invention is disallowed as a matter of patent law. It’s unclear, frankly, whether this rested upon a 101, WD or indefiniteness basis, but today it would definitely be at least WD if not indefiniteness (regardless of if the current indefinite test today is the right one).

          In response to Halliburton, 112, 3rd was created to change the specific situation of Halliburton – where the specification had support for a particular means but the particular means wasn’t claimed as a limitation.

          Since then, there have been particular statements made about situations that are not the claiming of a novel function that a lot of very silly people on this board seem to think also apply to a novel function. The only exclusion was in Wabash, where the court said that a functional description to distinguish inherent properties of already-known things is acceptable.

          So, for example, if one argues that combining chemical mixture a and mixture b and (functional description of a class of chemicals) to create a new mixture, and that new mixture is non-obvious because of the combination, that is okay. The applicant isn’t claiming he invented A or B or the other chemicals, the invention lies in the non-obvious combination.

          Contrast that with a processor running code to do function F, where F is argued to be novel. If function F isn’t claimed in means-plus, it’s invalid because the scope exceeds the possession of the applicant. The applicant here is not claiming a non-obvious combination of well-known code (indeed it’s pretty clear the combination would be obvious), he’s saying his function is new.

          Applying this rule, you can see how phrases like “functional language is not inherently invalid” “single means claims” “exact point of novelty” became phrases that people on this board have wildly misinterpreted.

          Functional language is not inherently invalid. However the way it’s being used in computer claiming today is invalid.

          Single means claims are inherently invalid, this is because as a single limitation there’s no obviousness analysis, and it’s forced to be the novel aspect of the invention. It does not follow that a combination of means claims are inherently valid.

          Exact point of novelty means just that. It’s completely fine to claim functionally for well-known things. The attempt to take a novel thing and claim it in two broad functional non-means-plus steps doesn’t make it valid either though.

          In computer claiming, the differentiating structure is the code. The structure need not be described in minute detail though, which is why the requirement is an algorithm, not the code itself.

          Seen in that light, it doesn’t matter if NWPA is correct. If you construe “processor configured to” as functional outside of means-plus, it cannot be valid for any novel function and is likely valid for any well known function. If you construe it as means-plus, you can only broaden the structure beyond your own by admitting that it existed in the prior art. Either way, you can’t achieve what I believe NWPA is hoping to achieve, which is seek the safety of 112, 6th but get the broadness non-6th-functional with respect to a novel limitation.

          There are a handful on this board who so misunderstand what the Supreme Court and the CAFC say that they think that Halliburton’s line of reasoning was overruled, when in fact it’s been upheld as recently as Lizardtech (and probably more recently). 112, 6th is a very narrow response to a very narrow situation to overcome a particular issue which confirms the Supreme Court was upheld on the Morse-Wabash-Halliburton line of thinking. That’s really all that matters for computer software claims.


            Trying to remove the gag…

            The biggest problem that you have Random, aside from the fact that you are wrong (Halliburton was indeed abrogated and the 101/103 and the removal of invention setting by common law is completely missed in your retelling of history), is in the “use” / “configure” aspect as has been well established.

            You cannot use an ‘oldbox’ to do a new function. You bursh aside the “inherency” aspect way too quickly. See Nazomi. See also Alappat.


            Also, Random, your ‘single means’ analysis has a fatal flaw in that you attempt to use PON incorrectly and subdivide the claimed invention improperly to arrive at the ‘single means,’ thereby eviscerating the claim as a whole requirement.

            Yes, the law of 101 is a mess. But you need to look at the correct (addicted) party to assess why.


              anon: “The long-recognized problem with a single means claim is that it covers every conceivable means for achieving the stated result, while the specification discloses at most only those means known to the inventor. See O’Reilly v. Morse, 56 U.S. (15 How.) 62, 112, 14 L.Ed. 601 (1853)” – In re Hyatt

              It’s not a question of “subdividing the claimed invention”. There’s a limitation and either it is described or it isn’t. The reason single means are per se invalid is because the only thing that could save the claim (that the limitation was only describing something old and non-novel and thus should be limited to known ways of achieving something) is unavailable when there is no other means of novelty.

              Halliburton was only abrogated to the extent it different from Wabash, which is Halliburton had a finding that the spec supported a single means. Wabash is good law. Morse is good law. Halliburton is good law if the claim is not placed in 112, 6th mode. This is why the court correctly said that 112, 3rd confirmed their reasoning (as it only defined a rule for scope interpretation, not a question of whether that scope was substantively valid). Your failure to understand this is why you are so messed up on the law.

              Since 112, 3rd, Fuetterer has confirmed this reasoning, several cases have cited to Morse and Wabash (which would be improper if Halliburton had been completely overruled as you seem to think) and Lizardtech is basically the same situation as Halliburton without availing them of 112, 6th.

              Nokia is a claim construction case and has nothing to do with whether the claim is valid, there is no WD discussion at all. Alappat does not concern a claim to a novel function not claimed in means-plus, which is the dominant means of expressing claims in computer software and the issue I argued about.


              anon: Just so I am very clear here.

              “I claim an apparatus comprising:

              a processor configured to”

              Situation 1 – “Run code C” where C is novel and described in the spec, the claim is valid. The scope is limited to C.

              Situation 2 – “provide means for executing function F” where F is achieved by code C that is novel and is described in the spec, the claim is valid under 112, 6th. The scope is limited to C.

              Situation 3 – “execute function F”. If F is novel, the claim is invalid for 112, 1st. If F is not novel, the claim will fail under 102 unless another limitation renders the claim non-obvious (here there is none).

              Situation 4 – “execute function F, and also non-obvious limitation X”. If F is novel, the claim is invalid for 112, 1st. If F is not novel, the claim is valid because X renders it novel or non-obvious. F is limited to known things that cause F.

              Notice there is no situation whereby 1) the claim is valid and 2) F is not limited to C.

              Notice also that regardless of whether you’ve claimed novel or non-obvious X, F’s description will invalidate the claim under 112, 1st if it is 1) novel and 2) not in means-plus.


              Wrong Random: Halliburton is considered tossed so there is no such thing as “Halliburton is good law if

              Abrogated means abrogated. You want to fall back to a different case, then you might the “to the extent that” thinking – but in reverse – to the extent that the prior case is like Halliburton, then that too is tossed.


              It should have been “Notice there is no situation whereby 1) the claim is valid and 2) F is NOVEL AND not limited to C.”


              You do understand the difference between a substantive change and a procedural option right?

              1) Halliburton decided a case using means-plus language before there was a means plus statute. So the scope that was given to that claim was “all means of achieving something”.

              2) The claim was considered under that scope and found invalid for lack of description.

              After that, a statute was written that told the court “when you see means-plus language, you will give it the scope of the particular means”.

              That abrogates the part of the ruling that defines the scope (1). It does not abrogate (in fact it doesn’t even pretend to mention in anyone’s mind other than yours) the reasoning with respect to the validity of claim that has the same scope as they considered.

              I can’t believe you’re so bad at this, let me put it in little kids terms:
              The court says 1 is true
              The court says unrelated thing 2 is true
              Congress acts to tell the court 1 is false
              Congress doesn’t say 2 is false
              2 is not only true according to the Court, now it’s confirmed by Congress. It’s actually more true than it was before.

              So the end result is that when the Court again comes up against something that has the broad scope, i.e. novel computer software claims not in means-plus, it’s just going to reassert 2 as it has before and has Congress has specifically considered and decided not to change.

              Tell you what, show me where in 112, 3rd the statute discusses the validity of claims that are to be construed as applying to every means of achieving something.


              Random, your statement of “I can’t believe you’re so bad at this” is matched by my feelings towards you.

              What part of abrogation do you not understand?


              anon: “What part of abrogation do you not understand?”

              Oh come now, anon, for someone who is so particular with words as you, surely you are capable of word parsing. Congress passed a statute that involved one particular issue within the case. Now, does that statute remove any other holdings of the case? By definition it does not (in fact, by definition it strengthens them).

              The court construed the claims to mean one thing, Congress acted to tell them the claims should be construed differently. That is what was abrogated. What part of abrogation do you not understand?

              The court, after completing claim construction, held that claims so broad are invalid. Congress left this holding in place by not mentioning it. (The court made a ton of other holdings, like a claim has be described and enabled, I hope you’re not dumb enough to think that was abrogated too)

              You still haven’t responded to my citations of Fuetterer or Lizardtech. You still haven’t explained why Hyatt’s reasoning with single means claims holding them per se invalid is correct if Halliburton/Wabash/Morse are incorrect in their reasoning.

              I’ll give you another out: Produce even a single case in which a) a novel function to all means of doing something is claimed and b) the novel function is considered with respect to Written Description and is upheld. I mean there have been hundreds of thousands of software patents where “configured to” has been used, so you should have no problem of finding just one case that upholds a novel function, right?


              Oh come now, anon, for someone who is so particular with words as you, surely you are capable of word parsing

              Oh come now Random, surely you know that you are not allowed to parse abrogated cases and take away only that which you want to take away.


            Random, totally agree with your post and a good summary.

            Check Faulkner v. Gibbs where the Supreme Court emphasized that its problem with functional claims is where they are functional at the precise point of novelty. Old elements, combinations do not invoke the rule.

            Faulkner v. Gibbs, 338 U.S. 267, 70 S. Ct. 25, 94 L. Ed. 62 (1949). link to



              As commented on previously, you seek to augment a pre-1952 case with another pre-1952 case.

              That is quite meaningless.


            Well said, RG. I don’t believe I’ve seen anyone here pushing the line that “functional language is inherently invalid.” I know I’ve personally said the opposite on many occasions.

            Those of use who decry the excessive reliance on functional claiming in the computer-implemented arts have been making the same points you’re making for a long time. But by all means keep making them because they really can’t be made often enough! It’s common sense and eventually the courts and the PTO are going to fully “get it”. They’re starting to get it already, which is why we’re seeing a pronounced uptick in complaints by those invested in the status quo.


              That a claim to any means of achieving the function you achieved is invalid.

              Some people view this under 101 because even though there is token language about a machine in there, really the claim is to a function rather than a machine. Others would view it as a 112, 1st failure since the scope of enablement and description does not match with the claim scope.

              The 101 argument may survive today, and frankly, the enablement one should survive (but probably doesn’t in terms of computer software), but the description one definitely survives.


              The difficult part under 101 is that listed there along with machine (and manufacture) is process, and the distinction between what is a process and what is a function is often not made correctly. See Deener 94 US 780.

              Something to keep in mind is the somewhat errant jurisprudence from an otherwise great Justice Story regarding process claims is the historical note that at Story’s time his borrowing and dependence on the then English patent law was flawed as England did not recognize process patents as the US did – the flaw being that ‘process’ was second class to the hard good categories. It is a subtle, but pernicious flaw.

    2. 2.2

      LB, good case and poster child for the argument I have been making that claims to old well known structure must be given a broad construction. That was done here. Both the “receiver for” and “means for receiving” were given the same broad construction to “receiver” because receiver was a fairly well-known structure in the context of the specification.

      As well, the reason that the “receiver for” clause was given non-MPF status was because “receiver” was well-known structure.

    3. 2.3

      “Prost pretty clearly (and persuasively, in my view) elaborates on the point that NWPA has made several times:”

      You can elaborate on it for all time and it won’t make it so, and it won’t make it any less likely to be immediately overturned by the supremes when the government finally gets around to it.

      After having reviewed the case though it appears to be just another case where a word is defined by its function irl and yet is considered “structural” just because it is a proper name for devices, regardless of whether the definition of that name is nothing but a reference to the function.

      It’s claim construction from a kindergarten perspective is what it is.

      For instance, no doubt the term does in fact convey definite structure(s) to ones of skill in the art. “We also come to this conclusion, in part, because
      EnOcean has provided extensive evidence demonstrating that the term “receiver” conveys known structure to the skilled person.” <that is what the court notes and nobody questions that. What is at issue is whether the category of structures in the claim is defined, for purposes of knowing where the bounds of the claim are, ultimately in nothing more than a functional fashion. The definition of the word is apparently functional in nature, so you can't just fall back on it being a term that denotes various structures when that class of structures is just being defined functionally anyway by use of the term at issue.

      Again, if the court would get its head on straight all this would settle down right fast and we wouldn't have so many cases going to them over this sort of issue. I wonder, if in 20 years time, the court would finally acquiesce on this issue sans supreme court bashing them over the head, if there are still people arguing the same. I doubt it, they seem to be a hard headed bunch that can't get their act together to save their judicial seats. Intent on imposing their judicial will on the world to skew the patent system pro-patentee. The tom-gamery that goes on simply does not appear to be on their radar's whatsoever.

    4. 2.4

      Though I should add, it appears that in the context of a mpf limitation as here it is fine to not consider the term mpf under current practice. Lemley’s suggestion that the counter-position be willingly adopted by the court not withstanding.

      1. 2.4.1

        Lemley’s suggestion that the counter-position be willingly adopted by the court

        And thus the problem with such advocacy: it is not the Court’s role to rewrite patent law.

        Our Constitution was very clear which branch of the government has that authority (all addictions to wax nose bending aside).

  3. 1

    Dennis, one of the reasons for filing a provisional application in big corporations is that they may have patent committees that must pass on whether to spend the money to actually file and prosecute the patent application. Routinely filing provisional applications allows one time to make a decision.

Comments are closed.