The Patent Office can materially improve the problem of uncertainty arising from examiner over-use of Alice-type abstract-idea rejections with improved guidance offering better definition of what might fall – and not fall – within previous guidance for abstract subject matter. Improved contrast would avoid a great deal of uncertainty and high rejection statistics, especially in the economically critical fields of e-commerce, software, and finance, which has led to a negative public response, as I have observed previously. Although I have recommended a more structured approach (with examiner training on such), but that is not enough to cure the problem if the guidance on which the examiners are trained lacks significant constraints or restrictions.
For the above reasons, the Office of Patent Legal Affairs (OPLA) – as the PTO’s policy branch, which drafted the Guidance and July 2015 Update on Subject Matter Eligibility (Section 101) – should take the lead in “reining in” the overbroad application of Alice observed in the examining corps. In general, the Office needs to take seriously the following prongs:
- Examiners need to be told what NOT TO DO – i.e., what are some of the improper forms of analysis – as well as what TO DO to make the decision.
- Instructions to examiners need to avoid open-ended language (e.g., phrases beginning with “for example …” or “such as …”) or categories that examiners can interpret more broadly than intended.
The particular areas of the Guidance and Update requiring improvement are (1) the “judicial descriptors” of categories of abstract subject matter and (2) the lists of ‘significantly more’ considerations. The Patent Office could help matters immensely by clarifying what the entries on those lists mean to include and contrasting with what they do NOT mean to include.
Step 2A: Identifying abstract ideas
The July 2015 Update says that examiners must compare ideas to concepts already found abstract, suggesting a case-by-case analysis. But the PTO also sets forth categories, which is inconsistent with the concept of merely analogizing to the closest court case. The PTO calls the categories “judicial descriptors.”
Although categorizing abstract ideas could be helpful, the use of categories expands the risk of overbreadth, especially when the categories have little definition, include sub-categories, and lack negative examples. The PTO should refine the categories of “judicial descriptors,” and do so both negatively and positively, to avoid overbroad application of Alice by examiners.
The use of “judicial descriptors” not supported adequately by court decisions has the potential to do great mischief in the area in which I practice frequently, i.e., software and Internet-related patent applications. I often seek review of claims for clients who made significant advances in such art, including through continuations of previous applications that have been granted as patents. The uncertain and changing landscape under Section 101 makes prediction difficult and outcomes seem somewhat arbitrary.
If the PTO believes it would help to retain “judicial descriptors,” it should provide more clear explanations, including what they are NOT intended to cover as well as what they ARE intended to cover, and the PTO should provide certain examples of ideas that fall within and don’t fall within each category. These need not be complete “examples” of the sorts released so far (with a specific claim, a discussion, and analysis of each step).
Criticism, including specific suggestions for improving wording, follow as to three of the “judicial descriptors” discussed in the July 2015 Update:
- Reword the definition of “economic practices.”
The Update defines this category as “concepts relating to the economy and commerce.” By use of non-limiting language, the PTO tells examiners they can place any idea in this category if it involves anything “relating to” business, of any kind. This is overbroad: It would swallow the rule that business methods are not ineligible as a class. After all, isn’t every business-method claim “relating to the economy and commerce”? The overbroad “relating to” phrase also contradicts the December 2014 Guidance, which says that a claim is potentially ineligible only when it is “directed to” the abstract idea, not merely when it “involves” an abstract idea.
The Update improves the definition by including the following clause: “such as agreements between people in the form of contracts, legal obligations and business relations.” But this clause does not cure the problem entirely because of the non-limiting words “such as.”
The PTO could improve this language by saying simply that “economic practices” refers to agreements between people in the form of contracts, legal obligations, and business relations or economic practices performed by humans.
Also, the PTO should tell the examining corps what this category is NOT intended to cover. Specifically, the PTO needs to explain that, merely because an idea includes a financial aspect of some sort or involves money or other compensation, such does not automatically mean that the claim is directed to an “economic practice” and thus satisfies Step 2A. Example cases establishing this point include U.S. Bancorp v. Solutran, Inc., CBM2014-00076, Paper 16, pp. 12-14 (PTAB Aug. 7, 2014) (method of processing paper checks held not a “fundamental economic practice” despite claim limitations involving “crediting an account” and despite economic content), and Trading Technologies Intl., Inc. v. CQG, Inc., No. 05-cv-4811, slip op. at 6-9 (N.D. Ill. Feb. 24, 2015) (improved GUI for ordering commodities on an electronic exchange does not recite a fundamental economic practice).
- Restore the Supreme Court’s “long prevalent” limiter to the definition of “fundamental economic practices.”
The “descriptor” is “fundamental economic practices,” and the Update says that the term “fundamental” refers to “foundational or basic” rather than “old or well-known.” Although this is a reasonable definition of the word “fundamental,” the Office has not taken into account the Supreme Court cases themselves, which indeed refer to “long prevalent” practices.
In Alice itself, the Court held that the concept of intermediated settlement was “a fundamental economic practice long prevalent in our system of commerce.” Alice Corp. Pty. Ltd. v. CLS Bank Int’l, 134 S. Ct. 2347, 2356 (2014) (quoting Bilski). The PTO’s Update chops off the last part of the quote from the “judicial descriptor,” and worse, tells the examining corps that it can ignore questions of “long prevalent,” because the word “fundamental” does not include “well known.” If the PTO adopts “judicial descriptors” like this, it should adopt what the Court said, nothing more; hence the limitation “long prevalent in our system of commerce” should be restored.
To support the statement that an abstract idea “need not be old or long-prevalent,” the Update relies on bio cases addressing the judicial exception for “basic tools of scientific and technological work” and the Ultramercial case holding a claim ineligible despite its novelty. These authorities are not on-point and in any event cannot override the direct Supreme Court characterizations in Alice and Bilski.
- Clarify or delete “certain methods of organizing human activity.”
Again, better definition, as well as some statements about what it is NOT included, would help avoid inconsistent application of “certain methods of organizing human activity.”
The Update says that this “judicial descriptor” is intended to “describe concepts relating to interpersonal and intrapersonal activities, such as managing relationships or transactions between people, social activities, and human behavior; satisfying or avoiding a legal obligation; advertising, marketing, and sales activities or behaviors; and managing human mental activity.”
Again, we see the phrase “such as” following a broad general category that is ill-defined. As to the category, what does “relating to interpersonal and intrapersonal activities” mean? If I brush my teeth, is that an “intrapersonal” activity, so that a novel toothbrush might be “relating to … intrapersonal activities”? I assume not. If an invention relates to communication between two people with a cell phone, is that “relating to interpersonal … activities”? Again, I assume not. The point is that the PTO should say what it means.
Turning to the words after the “such as” part of this “descriptor,” some of the examples help but some do not. For example, are all social-media inventions in the subcategory of “social activities”? I assume not, but then what does that subcategory mean to denote?
Are all inventions that involve “advertising, marketing, [or] sales” categorically patent-ineligible? The Update says, “Several cases have found concepts relating to advertising, marketing and sales activities or behaviors abstract, such as using advertising as an exchange or currency ….” Does that mean that examiners should understand that all claims having anything to do with advertising, for example, are abstract automatically? Some examiners seem to think that there is such a broad exception, but I doubt that the Update meant to signal such a thing.
Further, the inclusion of the word “certain” is singularly unhelpful, and the purported explanation in the Update does not assist: That word is allegedly used “to remind examiners that not all methods of organizing human activity are abstract ideas.” Left unexplained is the standard: How is an examiner to decide which ones are the “certain” ones that are “not all”?
Reading the language of the Update makes the answers to questions like those posed here less than clear. Lack of clarity results in excessive discretion on the part of examiners to apply Section 101 rejections inconsistently or even arbitrarily.
If the Office cannot define this “judicial descriptor” adequately, it should abandon it. The law does not require its inclusion, as others have observed. See Microsoft comment, at pp. 2-3 (no “indication that the [Alice] Court intended to identify it as a distinct category [of] abstract ideas” and it is not “announced as a category of abstract ideas elsewhere in the case law”).
- Clarify or delete “an idea of itself.”
The Update defines this “judicial descriptor” as “an idea standing alone such as an uninstantiated concept, plan or scheme, as well as a mental process (thinking) that can be performed in the human mind, or by a human using a pen and paper.” Again, this is unclear.
As to the “idea standing alone” part, although presented as though it were a definition, this phrase does not define “an idea of itself” any further. And what does “standing alone” mean, exactly? All claims are directed to ideas, presumably, so when is a claim directed to an idea “standing alone”? Put another way, what is the alternative – an idea coupled with something else significant – and if so, doesn’t that fold the analysis of part 2B into part 2A?
As for the “such as” part, does the phrase “as well as” mean to signal that both parts of the sentence are required, or should that phrase be interpreted as simply the word “or”? What is “an uninstantiated concept”? Can the Office suggest an example? One dictionary defines “instantiated” as “to represent an abstraction by a concrete instance.” As noted in point 2 above, the term “concrete” is an antonym of “abstract.” Is this just a way of saying something is “not abstract,” in which case we have a circular definition, so why is it included as an example of an abstract idea? Is “uninstantiated” just a way of saying “abstract” with a bigger word?
Another, more technically oriented, dictionary says, “Instantiation is the creation of an instance, which is a particular realization of an abstraction or template such as a class of objects or a computer process.” Must a claim include an “example” or a “particular realization” to avoid being “uninstantiated”? If so, an example of what?
As for the last phrase (“mental process (thinking) that can be performed … by a human using pen and paper”), this can cause much confusion too, and the Update should clarify what is meant – and not meant – further. It is possible to argue that ANY computer-implemented invention could be performed by a human using pen and paper. After all, a computer simply compares bits, so in theory, a human could do the same with a pen and paper. Again, if examiners use an overbroad interpretation of this clause, it will “swallow the rule” that not all computer software is ineligible. Clarification is again required.
The PTO’s latest (February 2016) training slides mention the “pen/paper test” (at slide 12), relying on the Update, and (at slide 13) gives an illustrative example (by highlighting claim language) of selected “claim steps” or “claim limitations” that allegedly “set forth the abstract idea” and make “the claim directed to an abstract idea.” The highlighted steps relate to receiving stock quotes (by a computer), filtering them using a comparison, generating an alert with a URL (link), formatting the data, and transmitting the formatted quote.
This example instructing examiners how to apply the “idea as a whole” category in practice well illustrates the manifold troubles with this “category.”
First of all, so much for considering “a claim as a whole” – apparently the PTO thinks the claim can be considered as “directed to” an “abstract idea,” namely “mental process” that could be performed using “pen and paper,” merely by examining SOME of the limitations.
Second, the PTO example fails to specify any “uninstantiated concept, plan or scheme” that the claim allegedly performs.
Third, the PTO example gives as a “reason” for this conclusion that “These steps can be performed mentally, and are similar to the concepts identified as abstract by the courts.” If the test is similarity to cases, then why have “categories” at all?
Fourth, the PTO example highlights the “similarity” to certain cases allegedly on the ground that the evaluated claim uses “categories,” does “comparing,” and is “using rules,” just like certain prior cases involving patents on different subjects. But couldn’t the same be said for virtually all software inventions? And we know that all software is not automatically abstract.
Fifth, the PTO example says, “Specifying the type of information being manipulated (i.e. stock quotes and user preferences) does not render the idea any less abstract.” Why not? Isn’t added particularly exactly what is needed to cause the claim to be less likely “an idea standing alone”?
Consider how “negative examples” would help clarify the situation. With judicious use of illustrations, the PTO could contrast the claim with (1) a claim that could be done with pen/paper but nevertheless is not abstract, (2) an example of a claim that has some computer steps that might not be considered “directed to an abstract idea,” (3) an example of a claim that does rules-based comparing that is not abstract, or (4) an example where added content that DOES “render the claim less abstract.”
To its credit, the PTO does use a contrasting example, but as to the second Mayo step, not the first. See (February 2016) training slides (compare slides 17 and 28).
Regardless of examples, the PTO ought to stop trying to defend “an idea of itself.” As others have observed, this category “is simply meaningless, and provides nothing substantive.” See Sachs, at p. 9 (commenting that the language “has been twice removed from its original context,” namely judicial cases from the 1930s that involved obviousness, not ineligibility).
In general, in attempting further clarifications, or in issuing further Updates, the Office should exercise great caution in wording instructions in vague or imprecise language. Practitioners and the examiners look to these instructions to guide the arguments they make, and use of vague or imprecise language can have severe implications.
We don’t need a flood of more words, especially vague terms or non-limiting explanations, that allow examiners still more discretion; what we need is language offering constraints and contrasts.
Step 2B: Identifying ‘significantly more’
When performing the “significantly more” analysis, examiners frequently ignore claimed computer-related limitations by saying that they are performed on known, generic computers or amount to conventional computer functions. The PTO needs to tell examiners that it is insufficient to rebut a “significantly more” conclusion just by showing by that limitations are performed on generic computers or are conventional computer functions. Many cases uphold patentability despite similar showings. The DDR Holdings case, for example, which I briefed and argued, agreed that the invention was patent-eligible despite being performed on known server equipment using conventional computer instructions. Examiners nevertheless issue gobs of rejections with this “reasoning.”
This is such a significant problem that it has caused an endemic practice of cut-and-paste form rejections under Section 101 in the computer arts. A typical “form rejection” of this sort, which can be found in literally hundreds of rejections, uses this language: “The additional element(s) or combination of elements in the claim(s) other than the abstract idea per se amount(s) to no more than recitation of generic computer structure that serves to perform generic computer functions that are well-understood, routine, and conventional activities previously known to the pertinent industry. Viewed as a whole, these additional claim element(s) do not provide meaningful limitation(s) to transform the abstract idea into a patent eligible application of the abstract idea such that the claim(s) amounts to significantly more than the abstract idea itself. Therefore, the claim(s) are rejected under 35 U.S.C. 101 as being directed to non-statutory subject matter. (See Alice v. CLS Bank, US Supreme Court, No. 13-298, 6/19/2014).”
One major fix would help immensely:
- Delete or explain one “negative consideration” that has proven troublesome in the computer-related arts.
The Guidance says that a limitation will not be considered “significantly more” by: “Simply appending well-understood, routine and conventional activities previously known to the industry, specified at a high level of generality, to the [abstract idea], e.g., a claim to an abstract idea requiring no more than a generic computer to perform generic computer functions that are well-understood, routine and conventional activities previously known to the industry.”
This part of the Guidance requires that, to count as a negative consideration for this reason, the examiner must show four different facts: that the claim element is  well-understood,  previously known,  so highly well-known as to be routine and conventional, and  specified at a high level of generality. But examiners commonly rip out of the definition the “e.g.” part referring to “a generic computer to perform generic computer functions.”
If necessary, when the patent attorney responds to a form rejection arguing that the computer functions are new, as revealed by the prior art, and thus not “well-understood, routine, and conventional activities previously known to the pertinent industry,” examiners feel free to rebut that showing by (1) pointing to the Update, which says that even novel claims might not be patent-ineligible, thereby ignoring the wording of this “consideration” in the Guidance itself, and (2) deconstructing the computerized claim element, arguing that it consists of solely of steps like comparison or data storage and retrieval, which are of course routine.
Use of language, without explanation (such as “requiring no more than a generic computer to perform generic computer functions that are well-understood, routine and conventional activities previously known to the industry”) can easily allow examiners to reject any computer-implemented invention that they choose to reject, in blatant disregard of the PTO’s assurances that there is no categorical exclusion of computer-implemented claims.
PTO assistance is required at the policy level. The PTO should tell examiners plainly and directly that, in the computer field, just because a computer invention is stated as having been programmed on a “general purpose computer,” that fact is not enough to discount a proposed limitation. Likewise, just because a computer limitation uses conventional programming, or relies on conventional computer tasks, such as comparison or storage/retrieval, such is not enough to discount it as being not “significantly more.”
Instead, an examiner should evaluate whether the computer-implemented invention is required to implement the abstract idea on the computer, adds insignificant extra-solution activity to an abstract idea, or simply applies the abstract idea in a new environment, like the Internet. In short, the above-quoted negative factor does not add to the other “negative considerations,” and the Office can delete it.
The cases at note 41 of the Guidance, cited to support this consideration, do not suggest that the Court viewed this as an independent reason for unpatentability, distinct from the other considerations; in Alice, for example, the Court simply listed particular computer functions implementing the abstract idea, after saying that mere instructions to implement the abstract idea on a computer were ineligible, and concluded, “each step does no more than require a generic computer to perform generic computer functions” to “instruct the practitioner to implement the abstract idea of intermediated settlement ….” Alice, 134 S. Ct. at 2359-60 (emphasis added). Conversely, an element, even if it is a “generic” function of a “generic computer,” that DOES do more than instruct a computer to implement an abstract idea can qualify as “significantly more” under Alice. Again, the PTO’s Guidance interprets the cases in a way that allows the exception to swallow the rule, and examiners are interpreting the Guidance beyond what the Office likely anticipated.
The PTO’s Appeal Board (PTAB) has recognized that the fact that a claim includes features that are known computer components is not disabling under the Part 2 test. E.g., Apple Inc. v. Sightsound Techns., LLC, CBM2013-00019, Paper 17, pp. 18-20 (PTAB Oct. 8, 2013) (Part 2B met even though, for “the specific computer components,” of the claim, “each was known individually in the prior art and can be found in a general purpose computer”).
If this “consideration” is retained, at a minimum the Office should add explanation sufficient to block the “form rejection” plague that the examining corps actively uses to block patents on nearly all computerized inventions. A far better solution would be to delete it entirely.
The examining corps’ over-use of Section 101 rejections, in sum, can be reined in via a more contrasting set of instructions about what is eligible and what is not. Examiners would like it; applicants would find examination outcome more predictable and know how to respond to rejections better; and reduced rejections represents good patent policy and will benefit the U.S. economy.
Join the Discussion
7 comments so far.
AnonFebruary 20, 2016 09:29 am
I find it difficult to accept at face value the assertion that the PTO management refused to allow well deserved (and presumably well articulated) 112 rejections to not be made.
Are you basing this on anything except anecdotal evidence?
PAUL MORGANFebruary 19, 2016 10:27 am
The over-emphasis on more 101 rejections comes right from the top. Examiners are actually getting criticized for not giving enough 101 rejections. so it is the Director you need to complain to about that.
The good news is that the former resistance of PTO management to examiners give well deserved 112 rejections is being overcome.
step backFebruary 18, 2016 11:43 pm
Can’t help but to agree with Me.
What better way is there for the PTO to increase revenues (think RCEs) and lighten their workload than with this Alice in the Rabbit Hole nonsense? Inventors are distracted from the substantive 102/103 arguments because they have to deal with the jiggery pokery insanity of Alice and Mayo. Win/win for the PTO and lose/lose for inventors and America.
There will not be a sir have you no shame moment because they have no shame.
NahFebruary 18, 2016 04:12 pm
“The overbroad “relating to” phrase also contradicts the December 2014 Guidance, which says that a claim is potentially ineligible only when it is “directed to” the abstract idea, not merely when it “involves” an abstract idea.”
Nah man, first you have to figure out what idea the claims are “related to” so that then you can decide whether or not they are indeed directed to, wholly preempt, or near wholly preempt the idea itself. There is no contradiction here, you just don’t understand what has to happen in order to actually review the claim.
“By use of non-limiting language, the PTO tells examiners they can place any idea in this category if it involves anything “relating to” business, of any kind. This is overbroad: It would swallow the rule that business methods are not ineligible as a class. ”
Nah it isn’t overbroad, it’s there to make sure that you’re not claiming any economic ideas, themselves. You have to claim more than the base economic idea in order to get into eligibility.
What they put down is more or less legally correct.
“Specifically, the PTO needs to explain that, merely because an idea includes a financial aspect of some sort or involves money or other compensation, such does not automatically mean that the claim is directed to an “economic practice” and thus satisfies Step 2A.”
They already have. It involves some thinking.
“These authorities are not on-point”
Yeah they are bro. Get used to it.
“How is an examiner to decide which ones are the “certain” ones that are “not all”?”
Usually this is done through this new thing called “thinking”.
“Does that mean that examiners should understand that all claims having anything to do with advertising, for example, are abstract automatically?”
It raises a red flag, yes. It’s not the end of the analysis, but yes, you’re on thin ice going in. Better have you something up in the claims to make sure you’re far from an abstraction.
“and if so, doesn’t that fold the analysis of part 2B into part 2A?”
It does in a way, and which is why the separation of the two steps isn’t really a thing in the actual legal analysis. They blend together in large part. They’re only separated to try to explain them simpletons.
“The PTO needs to tell examiners that it is insufficient to rebut a “significantly more” conclusion just by showing by that limitations are performed on generic computers or are conventional computer functions.”
That would be legal error. And perhaps your DDR Holdings case was legal error as well. Consider that. Legal error does from time to time find its way to the Fed. Circ.
“This part of the Guidance requires that, to count as a negative consideration for this reason, the examiner must show four different facts: that the claim element is  well-understood,  previously known,  so highly well-known as to be routine and conventional, and  specified at a high level of generality”
4 isn’t a “fact”, it’s a legal conclusion, part of the overall legal question of whether or not the claim is directed to an abstract idea. And 1-3 need not be established by evidence o record as the whole determination is an entirely legal one, unconnected from factual considerations. When a rejection like this is made, the office is concluding as a mater of law, that it APPEARS that this is the case, because those things APPEAR to be old/well known. If they’re not then simply argue they’re not.
“examiners feel free to rebut that showing by (1) pointing to the Update, which says that even novel claims might not be patent-ineligible, thereby ignoring the wording of this “consideration” in the Guidance itself, and (2) deconstructing the computerized claim element, arguing that it consists of solely of steps like comparison or data storage and retrieval, which are of course routine.”
A. that isn’t a rebuttal. and B. Good for them, better late than never. Should have done it on first action though.
“The PTO should tell examiners plainly and directly that, in the computer field, just because a computer invention is stated as having been programmed on a “general purpose computer,” that fact is not enough to discount a proposed limitation. Likewise, just because a computer limitation uses conventional programming, or relies on conventional computer tasks, such as comparison or storage/retrieval, such is not enough to discount it as being not “significantly more.””
Can’t help you bro, that’d be legal error :/
“The PTO’s Appeal Board (PTAB) has recognized that the fact that a claim includes features that are known computer components is not disabling under the Part 2 test. E.g., Apple Inc. v. Sightsound Techns., LLC, CBM2013-00019, Paper 17, pp. 18-20 (PTAB Oct. 8, 2013) (Part 2B met even though, for “the specific computer components,” of the claim, “each was known individually in the prior art and can be found in a general purpose computer”).”
That isn’t what the case held bro. They held that the petitioner did not adequately establish that the claims are more likely than not to be ineligible under 101. See page 16. There’s a difference. And also, it appears that some legal error crept into that board decision. Regrettable, but it happens. The petitioner stated that it was the abstract idea of “selling downloadable digital music”. But that’s not necessarily true. The claimed invention there was “selling AND DELIVERY of downloadable digital music ON THE INTERBUTS WITH COMPOOTERS” or video in other claims. Had petitioner adequately explained this, then the legal error by the board may have been avoided.
MeFebruary 18, 2016 12:12 pm
While this article is one of the better discussions of the problem of Alice, it is based on a false premise: that the Office doesn’t want the current results.
Half a year, and tens of thousands of 101 rejections, after the issuance of the Interim Guidance, the office issued an Updated Guidance which did nothing to rein in 101 rejections. Why then would anyone believe that the Office wants to get the situation under control? When the PID controller output is zero, more likely than not the controller thinks the target is being hit.
Appearance of …February 18, 2016 11:45 am
Although I agree with EG that Alice is both undefined and ultra vires, the fact remains that the process of correcting SCOTUS errors can be both slow and sometimes painful (think Slaughterhouse cases, and Dred Scott). So we need to be doing damage control. In particular, the USPTO should stop making things worse (i.e. extending the magnitude of the Alice decision by further reading other Alice cases very broadly) and start making things better (i.e. at least try to contain the damage by reading the Alice cases accurately). In this respect, Louis is making some excellent and long overdue points.
Here, however, I keep going back to the same dark thoughts that the USPTO “Alice” interpretations are no mistake, but rather reflect an unstated but high level policy that is hostile to software patents as a whole. I wish that someone high in the USPTO (*cough* Michele Lee *cough*) would address this basic issue: is it the USPTO intent to broaden “Alice” still further, or not? If not, then show us some movement in this direction, such as implementing at least some of Louis’ concepts above.
EGFebruary 18, 2016 08:20 am
I appreciate your effort to try to render “sanity” from what is currently “insanity” in terms of the how to apply the Alice test. But the fundamental problem is that the Alice test is broken at the outset. When SCOTUS won’t even bother to define what they mean by “abstract idea” and “something more,” you open Pandora’s box in the USPTO. 35 USC 101 is clear as to what is patent-eligible: a composition, apparatus, article (of manufacture), or process (method). Any other requirement which is engrafted onto the statute, as SCOTUS has done in Alice, is ultra vires.