Yesterday, hundreds of members of New Zealand’s technology industry expressed their support for a clear exclusion of software patents in New Zealand.
On the other side of this argument are primarily lawyers and patent attorneys, including AJ Park lawyer Matt Adams. I refer to Matt (in a friendly way, I hope!) because he is an expert in the field and has recently written a blog post entitled “The embedded software conundrum”, in which he claims that introducing the term “embedded software” into the Patents Bill will create “uncertainty”.
Matt’s post was timely, because it was at about the same time that myself and others were considering this very issue for the industry petition at SoftwarePatents.org.nz. In the end, we came to the opposite conclusion to Matt: embedded software should be referenced in the Patents Bill (see the requested amendment to clause 10A(2) here), and this approach has been resoundingly supported by the local industry (note the industry-supported amendment uses the interchangable term “embedded computer program”).
But what about the “uncertainty” that Matt refers to? We can’t possibly have an “uncertain” term in the Patents Bill, can we? In this post, I examine the issue of “uncertainty” that Matt raises.
Is it uncertain because there is no definition of “embedded computer program”?
There is no definition of “embedded computer program” in the Bill, nor does the industry-supported amendment include one. So because it’s undefined in the Bill, that creates “uncertainty”, right?
Wrong. Guess what else is not defined in the Bill, nor by any other law in New Zealand, nor by any Court in New Zealand: “computer program”. So how can there be objection to the lack of a definition for “embedded computer program”, when “computer program” itself is not defined? There can’t – and the same applies to many other terms in the Patents Bill.
The fact is, it is perfectly normal for undefined or hard-to-define terms to be used in laws. A major part of what Courts do is interpret laws and ascertain the meaning of words and phrases, including new or previously unconsidered terms. The famous judge Lord Denning said in Seaford Court Estates Ltd v Asher  2 KB 481:
Whenever a statute comes up for consideration it must be remembered that it is not within human powers to foresee the manifold sets of facts which may arise, and, even if it were, it is not possible to provide for them in terms free from all ambiguity. The English language is not an instrument of mathematical precision.
So far from presenting a “conundrum”, the lack of a definition capable of handling all sets of facts, and free from ambiguity, is not only perfectly normal, but such a definition is impossible. There are numerous undefined, or incompletely defined, terms in the Patents Bill, the Act it replaces, and every other law. For example, at the moment there is a legal question over whether a Segway is “motor vehicle”. Should “motor vehicle” be removed from our laws because it is currently unclear whether Segways or other devices are included? Of course not, but that is the logical conclusion of those who say that “embedded computer program” should not be used in the Bill unless a comprehensive definition is agreed upon.
But how uncertain is embedded software anyway?
Proponents of software patents in New Zealand (i.e. primarily lawyers) have not pointed to any evidence that the term “embedded software” would be uncertain or cause problems. The term “computer program” (or “software”) is not even in the current Act, let alone defined, yet that has not prevented the granting of software patents. So the ball is in their court, so to speak, to justify why “embedded software” would be problematic, when other undefined or even unstated terms (in the current Act and the Bill) are apparently fine.
But it would be the wrong starting point to ask whether people in one industry understand a term from another industry, as an indicator of that term’s “uncertainty”. If a non-meteorologist struggles with a meteorology term, does that make the term uncertain? No, just like a patent attorney unsure about the term “embedded software” does not make that term less certain.
Therefore it is important to look to the industry, in this case the New Zealand IT industry, for acceptance and regular usage of a term. The industry seems to have little difficulty with the term “embedded software”. For example, Fisher & Paykel Appliances, a leading patenter of inventions (and a client of AJ Park, I note), is currently advertising a job for its “Electronics Embedded Software Team“:
We have a range of opportunities to suit either tertiary qualified graduate level and individuals who have experience in working with embedded software development
The term can’t be too uncertain if Fisher & Paykel Appliances has an entire team named after it, and is hiring staff to work on it!
We also know that Fisher & Paykel Appliances’ inventions contain embedded software, apparently without any definitional difficulties. In fact, we have this on the excellent authority of Matt’s AJ Park colleague, Ken Moon, who said recently that:
The exclusion of embedded software inventions of the type devised by high-tech kiwi manufacturers such as Fisher & Paykel Appliances and Navman for their computer controlled appliances GPS systems is somewhat irrational.
(Good news: today’s petition shows that the NZ industry agrees.)
But it’s not just Fisher & Paykel Appliances who have no difficulty with the term embedded software. A quick search on Seek.co.nz shows there are currently 48 jobs using the phrase “embedded software”, including Embedded Software Engineers, Embedded Software Developers, and so on. So while lawyers, for example, may claim the term is uncertain, the evidence is that the software industry itself has no difficulties with it.
There are many other resources confirming the commonplace use of the term in the IT industry. For example, there are over 42,000 articles on Google Scholar discussing “embedded software”, including from authoritative bodies such as the IEEE and Association for Computing Machinery. And of course there have been numerous patents involving “embedded software” already granted. There are many people and resources that can provide a sufficient description of embedded software, with no less certainty than any other words used in either the Patents Bill or elsewhere.
So while some people in New Zealand might be confused or unclear about embedded software, that does not mean that the technology industry, or a Court, would have the same difficulty. In fact the evidence is that they would not. Therefore the onus is on the proponents of software patents in New Zealand to substantiate why a plain-English, industry-accepted term – already used in numerous patents – should not be used in the Patents Bill.
What about “as such”?
Any possible questions about the industry-accepted term “embedded software” pale into insignificance compared to the European-style “as such” exclusion recently introduced into the Bill (i.e. the clause 10A(2) that the petition seeks to replace). If you ever wanted an example of notoriously uncertain, problematic language, it is the European “as such” terminology.
But don’t take my word for it – here’s what the Institute of Patent & Trade Mark Attorneys of Australia said in their 2010 submission to the NZ Minister (and which was expressly endorsed by the Australian Federation of Intellectual Property Attorneys):
[The EPC approach] will still introduce uncertainty and confusion as to the patentability of computer related inventions… The approach in Europe has also led to considerable uncertainty and criticism amongst European industry, the European Patent Office and the Courts… the approach has been often criticised as being inconsistent and contradictory to the EPC itself… In our view, adopting the EPC approach would create a significant degree of uncertainty as to how the section is to apply and which, in any event, is inherent in the EPO approach.
Adopting an approach comparable to the EPC and UK Patents Act is also likely to lead to unpredictability, as an interpretation of the “as such” clause by the European Patents Boards of Appeal has that computer programs are only patentable if they produce a “further technical effect” so as to render the claim allowable under article 52 of the EPC… The EPO approach has often been criticised in Europe as giving rise to unpredictability in that jurisdiction. Considerable costs have been incurred in dealing with and attempting to avoid the exclusion that was imposed in 1977…
So we are considering a choice between the possibly uncertain yet industry-accepted term “embedded computer program”, versus an approach proven to be uncertain, confusing, inconsistent, contradictory, unpredictable, and costly – as warned by the Australian patent attorneys.
Once again, the onus is on the New Zealand software patent proponents to explain why a discredited European-style exclusion, heavily criticised by our patent expert friends across the ditch, is in any way superior to the industry-supported exclusion that avoids all these problems.
What about European case law?
Another claimed source of “uncertainty” of adopting a “new” term (and thus not adopting the European approach) is that New Zealand Courts would be unable to call upon the body of European case law.
Well, the expert advice from our Aussie friends strongly suggests that is a good thing! Again I quote from the Institute of Patent & Trade Mark Attorneys of Australia:
For example, the UK Court of Appeal has said “The decisions of the EPO Boards of Appeal are mutually contradictory” and its approach in two decisions “is simply not intellectually honest”
So access to contradictory, intellectually dishonest, and maligned case law would be an advantage how, exactly?
What about UK case law? The leading UK case has been criticised by the Board of Appeal of the European Patent Office (EPO) as being “irreconcilable with the European Patent Convention”. Which is another reason for not going down the European path.
OK. But didn’t the Select Committee say “embedded software” was uncertain?
No. What they said was:
After careful consideration we concluded that developing a clear and definitive distinction between embedded and other types of software is not a simple matter; and that, for the sake of clarity, a simple approach would be best.
And that is quite right. As noted above, and as Lord Denning expressed rather more eloquently, definitive definitions are generally impossible. Therefore, the Select Committee sensibly recommended a simple approach.
And that is what the industry-suported amendment preserves. The exclusion of patents for computer programs remains, but with an express carve-out to ensure patentabilty of inventions containing embedded software – not the embedded software itself. Consequently there is no need to determine whether software is embedded or not to determine whether it is patentable. No “uncertainty” arises: under the industry-supported amendment, no software would be patentable. It’s pretty simple. And it’s entirely consistent with the Select Committee’s unanimous recommendation.
On the other hand, the European-style exclusion (which as the Australian patent attorneys say is uncertain, confusing, inconsistent, contradictory, unpredictable, and costly) is entirely contrary to the Select Committee’s unanimous recommendation.
So where does this leave us?
Every conundrum has a solution, and in this case it’s the industry-supported amendment. This amendment:
- Preserves the Select Committee’s unanimous recommendation that software be excluded from patentability;
- Ensures that patents are still available for inventions containing embedded software, such as Fisher & Paykel washing machines; and
- Avoids the controversial, counter-productive and unwanted European “as such” proviso in the current clause 10A(2).
It is also consistent with the Government’s statements, consistent with what Fisher & Paykel proposed as recently as April 2012, consistent with industry-accepted terminology, and perhaps most importantly, it is what the New Zealand industry has loudly endorsed.
What’s uncertain about that?