07 October 2012

Reading Prior Art When the Times, They Are A-Changin’

Bradken Resources Pty Ltd v Lynx Engineering Consultants Pty Ltd [2012] FCA 944 (31 August 2012)

Construction – consideration of how a person skilled in the art would have interpreted and understood the prior art – time at which prior art is to be construed

Monk Reading - RembrandtThere are many issues of patent law touched upon in the 369 paragraph decision of Justice McKerracher in Bradken Resources v Lynx Engineering, however in this article we will look at just one – the continuingly vexed question of the date at which a prior art document is to be interpreted (the more common terminology in patent law being ‘construed’).

As the court noted in this recent case (at [211]) there appears to be a schism in the authorities on this question.  Two different dates – possibly very different in some cases – are generally proposed as being the relevant time to assess the meaning and content of a prior art document:
  1. the date of publication of the prior art document; or
  2. the priority date of the claim which is being assessed (e.g. as to novelty or inventive step) in relation to the prior art document.
As we shall explain, there are sound reasons why either one of these dates might be chosen, although generally for different purposes.  There are (at least) three different reasons for which one might want to understand exactly how a document should be understood:
  1. to determine what is meant by various terms used in the document;
  2. to determine whether the document shows that something was fully known and/or used at some time in the past (e.g. to assess novelty of a claim); and
  3. to determine whether a skilled person would be able, after reading the document, to arrive at something which, while not fully described, could nonetheless then be derived by routine steps (e.g. to assess obviousness of a claim).
We would argue that only the third of these reasons should involve any consideration of the state of the art, or the knowledge and understanding of a relevantly skilled person, at the priority date of the claim(s) under consideration, and even then the meaning of the document at its date of publication remains a relevant consideration.

In this case, the court construed a prior art document at its date of publication (in 1936), for the second purpose above, i.e. to assess the novelty of Lynx’s patent claims, which we consider was the correct approach.


For the prior art publication date…
The earliest authority commonly cited in relation to the date of construction of a prior art document is a decision of the Privy Council (on appeal from the Supreme Court of New South Wales) in Ore Concentration Co Ltd v Sulphide Corporation Ltd (1914) 31 RPC 206, where it was stated, at 223-224, that:

It is a general canon of construction, applicable to all documents, that the document should be construed as if the Court had to construe it at the date of publication, to the exclusion of information subsequently discovered. In Patent cases the observance of this canon of construction has great importance. It is common, in such cases, to have a number of documents placed in evidence extending over a considerable period of time, each of which is relied on as disclosed relevant information prior to the date of the Patent. If these documents require the assistance of experts to aid the Court in construction, the Court is deprived the benefit of such assistance if the witness is asked to read the document, not in reference to what was known at the date of publication, but to knowledge only acquired at some subsequent date. (Emphasis added)

Of course, this ‘general principle’ makes perfect sense when, for example, the document to be construed by the court is a contract.  It is completely irrelevant what changes might have taken place – in technology, in business, in the relationship between the parties, in the economy – since the date of execution of the contract.  The task for the court is to determine the intentions of the parties at the time.

The patent law throws up different questions for a court, which may need to be answered with reference to evidence of the relevant circumstances at a number of different dates.  As was stated in General Tire & Rubber Co v Firestone Tyre & Rubber Co Ltd [1972] 89 RPC 457 at 485:

The earlier publication and the patentee's claim must each be construed as they would be at the respective relevant dates by a reader skilled in the art to which they relate having regard to the state of knowledge in such art at the relevant date. The construction of these documents is a function of the court, being a matter of law, but, since documents of this nature are almost certain to contain technical material, the court must, by evidence, be put in the position of a person of the kind to whom the document is addressed, that is to say, a person skilled in the relevant art at the relevant date. (Emphasis added)

The first passage above, from Ore Concentration, was cited with approval by the Federal Court of Australia in E I Dupont de Nemours & Company v Imperial Chemical Industries PLC [2002] FCA 230 at [57], stating (at [62]) that ‘the Privy Council expressed the firm view that prior art is to be construed as at the date of its publication’ but that ‘this issue awaits authoritative resolution in Australia.’

At first instance in ICI Chemicals & Polymers Ltd v Lubrizol Corporation Inc [1999] FCA 345, at [70]-[78], Justice Emmett proceeded on the basis that a prior art document was to be construed as at its date of publication.  On appeal from this decision, in ICI Chemicals & Polymers Ltd v The Lubrizol Corporation Inc [2000] FCA 1349 at [40]-[46], the Full Court of the Federal Court of Australia discussed the issue, but found it unnecessary to decide at which date the prior art should be read.

In JMVB Enterprises Pty Ltd v Camoflag Pty Ltd [2005] FCA 1474, Justice Crennan stated (at [55]) that ‘when determining the degree of disclosure required [to show lack of novelty] one must construe the [prior art] document as at the date of its publication’ citing the above passage from General Tire.

And now, in Bradken v Lynx Engineering, Justice McKerracher has also concluded that ‘the prior art should be construed at the date of its publication, not at the priority date’ (at [214]).
For the priority date of the claim at issue…
The origin, in the Australian authorities, of the notion that a prior art publication should be construed at the priority date of the claims at issue appears to be the decision of the Full Federal Court on appeal in Nicaro Holdings Pty Ltd v Martin Engineering Co [1990] FCA 40, in which Gummow J cited (at [31]) a passage from Lord Reid’s judgment in C. Van der Lely NV v Bamfords Ltd. (1963) RPC 61, noting that this same passage was cited with apparent approval by the High Court in Olin Corporation v Super Cartridge Co. Pty. Ltd. [1977] HCA 23.

The references by Gummow J in Nicaro have been widely interpreted as supporting the proposition that a prior art document should be construed at the priority date of the claims at issue (e.g. E I Dupont at [61], ICI in the Full Court at [45], Bradken v Lynx at [211]).

The clearest statement of the proposition is to be found in Ramset Fasteners (Aust) Pty Ltd v Advanced Building Systems Ltd [1999] FCA 898, at [18]:

In relation to the doctrine of anticipation in patent law, the question is what would have been revealed to the skilled reader at the priority date.

However, the court provides no specific authority for this proposition.  The closest pertinent citation is to a statement of the High Court in Hart Pty Ltd v S W Hart & Co Pty Ltd [1978] HCA 61 at [17], which in fact relates to obviousness, rather than anticipation.


Justice McKerracher, in the present case at [214], observes that:

It is conceivable that the division in the authorities is driven by the fact that there should not be a hard and fast rule and that it will depend on the circumstances of each case.

In our view, it is not a matter of the circumstances in each case per se, but rather of the question actually under consideration.  A different rule may apply to different questions within the same case.

An example may be of assistance, and we trust that readers will bear with us using electronics to illustrate the point!

Power TransistorThe transistor (at least in a practical, working, form) was invented in 1947, providing a solid-state alternative to the vacuum tube for the first time since its invention in 1907.  These days, vacuum tubes are rarely seen (except in some audiophile amplifiers), largely because, objectively speaking, the only things they do better than transistors are: occupying large volumes of space; converting electricity into excess heat; and glowing attractively (while operating normally).

Vacuum TubeImagine, then, a document published in 1920 which explains the principles of design of a particular kind of amplifier, constructed (of course) using vacuum tubes.  Clearly, at its date of publication there is no respect in which this document could be said to anticipate a similar transistor-based design, since the transistor did not exist!

Now jump forward in time to 1948.  A viable transistor has only just been invented, very few people are yet aware of it, and fewer still have any good grasp of exactly how it works and what its properties are.  Eventually, all electrical engineers will be routinely taught how to use transistors in the design of amplifiers, and it will turn out that many of the principles used for designing tube amplifiers are also applicable to transistor designs.  But in 1948, this day is yet to come.

So, it is still the case that, in 1948, the 1920 publication would not be recognised, by an ordinary skilled person in the art of electronics, as disclosing a transistor amplifier design.

Jump forward again to the present day, and any relevantly skilled person would immediately recognise that the vacuum tubes in the 1920 design could be replaced with transistors and, with perhaps a few tweaks to other components in the amplifier, be made to work using modern technology.

But does this mean that the 1920 publication, which would not anticipate a claim to the equivalent transistor-based design in 1920 or 1948, is now a novelty-destroying reference for the same claim?  We would argue not.  The words of Emmett J in ICI (first instance) at [71] seem pertinent to the issue at hand:

The contention appears to entail the proposition that, even if the claim under attack would have been novel if first disclosed at the time of the publication of the alleged anticipating document, it was not novel at the time of its actual disclosure, simply because the hypothetical skilled addressee would know more in 1989 than in 1957. Such a proposition appears to me to blur the distinction between lack of novelty and obviousness. Common general knowledge is clearly an expanding, or at least variable, universe. Something, which might be obvious in 1989 may not have been obvious in 1957 because common general knowledge has expanded during the intervening time.

To put it plainly, whether or not a document anticipates a given claim is a ‘timeless’ question.  Either exactly the same thing was known or done at some point in the past, or it was not.  And, if not, then the claim is novel.

In the transistor example, for anticipation one should properly be looking in the period following the invention of the transistor for an example of the amplifier in question having already been put into practice.  If it has not, then it remains novel, even in 2012.

Obviousness, however, is a different matter.  Clearly, the common general knowledge in a field evolves, as does availability of additional publications which might be brought to bear in support of an allegation of lack of inventive step.  In contrast to the ‘timeless’ question of novelty, obviousness is to be evaluated through the eyes of the relevant skilled person at the priority date of the claim, after the content of any pertinent prior art references has been properly construed.

Even then, the interpretation of terms in an ‘old’ reference may require consideration of its meaning at its date of publication.  For example, our hypothetical amplifier publication may make reference to ‘condensers’, which might leave the modern electrical engineer scratching her head, until someone explains that this is what ‘capacitors’ were once commonly called.  In effect, the interpretation of the document as it would be understood in 1920 must be placed into the head of the skilled person as at the priority date – with any changes in terminology, notation and so forth that might be appropriate – before evaluating questions relevant to obviousness.


As Justice McKerracher has pointed out in Bradken v Lynx Engineering, there is some apparent divergence in the authorities as to the appropriate date for construction of prior art documents.

However, upon closer consideration it becomes apparent that the authorities strongly support the proposition that when considering anticipation (as was the case in Bradken itself), the relevant date is the date of publication of the prior art reference.

Of the remaining authorities, any discussion of the issue is ambiguous (General Tire in particular refers repeatedly to ‘relevant dates’ without ever expressly stating what they are) and/or is obiter dicta.

The one exception, i.e. the clear statement in Ramset that anticipation is determined by ‘what would have been revealed to the skilled reader at the priority date’, appears simply to be wrong.  As noted already, it appears to have been mistakenly inferred from statements made by the High Court in relation to obviousness.

Generally speaking, it will be necessary to construe a prior art document at the date of its publication.  An old reference may use obsolete terminology, which the modern skilled reader would not understand without assistance.  Nonetheless, if something falling within the scope of a claim has, in fact, been made or done before, the claim cannot be novel merely because the modern reader might have trouble interpreting an old text, any more than it would be novel if the prior art reference were written in a foreign language.

However, when it comes to assessing obviousness, then it is a different matter.  The knowledge and capabilities of the skilled reader at the priority date of the claim must necessarily be brought to bear, perhaps on the assumption that an ‘old’ reference has been ‘translated’ into a modern form.  (Of course, under the current Australian law, the requirement in Section 7 of the Patents Act 1990 that the prior art also have been ‘ascertained, understood and regarded as relevant’ must also be considered.)

In short, a prior art reference is to be construed at its date of publication but, depending upon the question at hand, it might also need to be (re)construed as at the priority date of the claims under challenge.  And by and large, although the issue seems never to have been stated in precisely these terms, the courts seem to understand this.

No comments:

Post a Comment