In a patent application, the claims, the written description, and the drawings are supposed to fit together in a consistent way, according to a few basic rules. Disregard those rules, and — best case scenario — you’ll get to waste a lot of time and money filing corrected drawings and amendments to the specification. Worst case, you’ll get claims rejections that you can’t overcome without filing a continuation-in-part (expensive), and maybe not even then if some prior art has popped up in the meantime.

First, a bit of terminology and background. The claims are what determine the extent of your patent rights. You can think of claims as being made up of some elements, and some relationships between the elements. Referring to the canonical example invention — a chair — the claim might be something like:

A chair, comprising a seat, at least one leg extending downward from the seat, and a back extending upward from the rear portion of the seat.

In this example, the “elements” would be the seat, the leg, the back, and possibly the rear portion of the seat. Claims must also indicate how the elements relate to each other. Here, for example, “extending downward from the seat” indicates the relationship of the leg to the seat.

Now for those rules:

  1. Every element of every claim must be supported by the specification.In practice, that means that if one of the elements of a claim is “a seat”, the written description had better spell out what the inventor means by “a seat”. (Doing that optimally is more complicated than you might think, but we’ll save that topic for another day.) The PTO enforces this rule fairly vigorously, and you can pretty much count on a rejection of any claim that contains elements that aren’t specifically mentioned and described in the original specification.
  2. In theory, every element of every claim should also appear somewhere in a drawing figure. This one is somewhat less likely to be nit-picked. The rules require “a drawing where necessary for the understanding of the subject matter to be patented,” MPEP 608.02, so as long as the drawing figures reasonably cover the subject matter, examiners usually don’t go out of their way to be obsessive. However, the rules also say that “the drawing in a nonprovisional application must show every feature of the invention specified in the claims,” MPEP 608.02(d), so it’s wise to err on the side of inclusion.
  3. Elements in patent drawings are labelled with reference numbers, also sometimes called “callouts”. The rules require that the features of drawings be labelled with numbers, not descriptive text. The descriptive text belongs in the written description. Thus, in the “chair” example, a drawing figure might look something like this:
  4. The written description refers to drawing features by reference number. For example, “In an embodiment as shown in Fig. 1, a chair includes a seat 101, one or more legs 103 extending downward from the seat, and a back 105 . . .” (and then go on to describe each element in more detail, list possible variations, etc.).
  5. Reference numbers must be used consistently. If you refer to “a leg 103” in one place, don’t use reference number 103 for a “hinged fastener” somewhere else. Conversely, don’t use a different reference number for a “leg” in another figure (unless it’s different in some way that you need to distinguish, in which case you would call it something else).

These rules seem simple enough, and they are. But in the context of a 50 page written description, 20 pages of drawing figures, and dozens of callouts, the bookkeeping can be a bit challenging.

There are three main issues here from a mechanical, nuts and bolts standpoint:

  1. Orphan callouts in the drawing figures: elements that are shown with reference numbers in a drawing figure, but are never referred to in the written description.
  2. Orphan callouts in the written description: elements that are referred to by a reference number in the written description, when there is no corresponding reference number in any of the drawing figures.
  3. Incorrect naming of callouts in the written description: elements that are referred to in the written description with the same reference number but with different names or referring to different elements.

All of these things are bad.

You can minimize the errors by approaching the drafting process systematically — keep a list of elements and callout numbers as you write, for example — but it isn’t unusual for inconsistencies to slip in as you pass drafts back and forth, make corrections and changes, etc., so it’s still a wise practice to give the final draft a good thorough consistency check.

You can, of course, go through the whole specification line by line, element by element, and look for mistakes of this kind by hand, but that’s one of those tedious jobs that a computer can do better than the average human. So, I finally overcame inertia and wrote a program that automates most of the task. You can access it here, together with an explanation of what it does and how it works.

(A word about security: With patent applications, confidentiality is obviously a concern. This program is written in javascript and runs in a web browser (a reasonably up to date one), so you can run it without needing to download anything or install anything. The entire program is in the web page; it is entirely “client-side” and standalone, and does not communicate any data to the server or anywhere else. If you’re a javascript person you can verify that by reading the source code — just press cntl-U in your browser — it’s all there and it isn’t minified or obfuscated. Or, if desired, you can run the program offline, just load the page and disconnect from the internet before you run it.)

Leave a Reply

Your email address will not be published. Required fields are marked *