From October 2016 to March 2017 the team is joined by Guest Kats Rosie Burbidge and Eibhlin Vardy, and by InternKats Verónica Rodríguez Arguijo, Tian Lu and Hayleigh Bosher.

Friday, 8 November 2013

Does IP Help or Hinder Science?: The Problem of Replication

To what extent is IP a help, or a hindrance, to scientific research? One need go no further than the Copyright and Patent Clause of the U.S. Constitution (Article I, Section 8, Clause 8), which famously empowers the United States Congress
 “To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.” 
From this clause were spawned the U.S. patent and copyright laws. In other countries, even without any Constitutional underpinning, the Legislator has been enacting IP legislation for over 300 years. Clearly, therefore, these laws are meant to promote writings and discoveries. But not all writings and discoveries are protected by the relevant IP law. Perhaps even more critically, the IP laws do not always work in the interests of promoting scientific activity.

An excellent example of the potential disjunction between IP and science was described in the article, “Trouble at the Lab”, which appeared in the October 19, 2013 issue of The Economist, here. The focus of the article is the problem of scientific replication. As summarized in the article:
“The idea that the same experiments always get the same results, no matter who performs them, is one of the cornerstones of science’s claim to objective truth. If a systematic campaign of replication does not lead to the same results, then either the original research is flawed (as the replicators claim) or the replications are (as many of the original researchers …contend). Either way, something is awry.”
One of the major problems in replication is access to the methods and data of the original data. The issue is not merely that fewer and fewer researchers appear willing to share their research for the purpose of replication. Thus, as reported in the article, less than 50% of researchers stated that they would share their raw data if requested to do so, and it was only a similar proportion of researchers who complied with the data-sharing policy in place with the journal in which the article was published. But, in addition to the increasing unwillingness of researchers to share data, there appear to be IP roadblocks as well.

The first relates to the software used in the original study. The article notes that sometimes the researchers rely on computer code that they have painstakingly developed and which they view as valuable IP that gives them an advantage in future research. In such a case, as observed,
“[a]lthough most scientists agree in principle that data should be openly available, there is genuine disagreement on software. Journals which insist on data-sharing tend not to do the same for programs.” 
From the vantage of the research, taken as a whole, both the raw data and the related software should be viewed as merely two parts of a single overall project. However, legally speaking, there is a fundamental difference in IP treatment between raw data (which, disregarding sui generis regimes for protecting data qua data, such as in the EU, is protected only as to the structure of the data) and computer software, which is protected per se under copyright. The upshot is that copyright law might make it more difficult and, in some instances, virtually impossible to carry out a meaningful replication study, where customized software was crucial in carrying out the original research.

The second issue involves trade secrets and confidential information. As pointed out in the article by Professor Harry Collins, of Cardiff University in Wales, experiments involve the “tacit knowledge” of the experimenters. As described, tacit knowledge refers to “craft skills and extemporisations that their possessors take for granted but can pass on only through example.” This means that a replication might not succeed simply because the replicator “didn’t quite get these je-ne-sais-quoi bits of the protocol right.” Because of this tacit knowledge, the likelihood that the same result can be reached on replication is slight. Thus there needs to be a change in what counts as “the same procedure.” But the unsettling effect of tacit knowledge on the ability to replicate a study  in full could also be reduced if more of this tacit knowledge was passed on to the replicators. This is not to say that all such tacit knowledge could be so transferred as a practical matter. But, even if it could, the nature of legal protection for valuable confidential information provides the original experimenter with an unassailable legal right to refuse to share such tacit knowledge with the replicator. Once again, IP seems to work to the potential detriment of scientific progress.

This Kat is aware that the utilitarian nature of trade secrets and computer software means that the legal protection of such rights does not share an identical focus with the outputs of scientific research. Still, he wonders whether it might be valuable for increased dialogue on the points of intersection and divergence between the two.


Anonymous said...

I take issue with the assertion that "copyright law might make it more difficult and, in some instances, virtually impossible to carry out a meaningful replication study, where customized software was crucial in carrying out the original research".

If there were no copyright protection for software code, researchers would be less likely, not more likely, to share their source code. At least copyright would allow the possibility of drawing up an agreement to share software for the specific purpose of replication, but not for any other purpose.

Anonymous said...

The insidious effect of companies legal policies which for defensive reasons dictate that all information is confidential so as to protect the IP means that scientific knowledge is not shared.

Similarly the same lawyers exploit loopholes in IP laws to either troll for income or stultify invention by using massive legal costs threats to stop innovation at competitors.

The IP industry needs to be aware that external observers might conclude that a lot of IP is like the 'emperors new clothes' and will be constrained when subject to the cold light of day.

Anonymous said...

Yet another article that spins the wrong message.

In the guise of "hindering" progress, you left out the fact that patents are not infringed if duplicated for scientific curiosity.

Such a blatant mistake casts a pall over the entire article and makes me wonder why you wrote it in the first place.

We deserve better journalism.

ipdraughts said...

There seems to be a typo in the comment from Anonymous at 12:16. I think it should read:

"Similarly the same _businessmen employ_ lawyers _to_..."

Anonymous said...

The author's underlying assumption seems to be that advancement of science is the ultimate goal. I disagree. The ultimate goal, however, is not science worship but rather to fashion a society so that free individuals can work and create and pursue their happiness as they see fit. Private, personal property rights are the cornerstone of such a society. The advancement of science is but a peripheral benefit.

Anonymous said...

In reply to the earlier comment that patents are not infringed if replicated if duplicated for scientific curiosity ...

This is also incorrect, a huge oversimplification of a complex and much litigated question.

Few academic or simply curious replicators are ever sued for patent infringemeent, but to assume they are exempt might be foolhardy.

Anonymous said...

I don't think IP is to blame for researchers being unwilling to share their related software, any more than property law is to blame for holding back science by making scientists reluctant to share their physical resources. If I wrote to CERN and said 'I'd like to replicate the research leading to evidence of the Higgs boson; can I come and use the LHC?', the result would be either: yes, and this is what you will have to pay to use the facility; or no, go and build you own particle collider. Same with the software; either pay towards the time and effort invested in creating it (or invested in persuading a research funding organization to pay for it), or write your own software. And before anyone mentions 'what about publicly funded research' that taxpayers have paid for; I wouldn't expect to go to someone's lab and use their apparatus, so why should I be able to use their software?

Anonymous said...

Anonymous at 13:47.
I call BS on your "huge simplification."

The matter is really that simple.

Now if YOU want to complicate things and use that scientific curiosity to make a profit or otherwise work around the right of the patent, THEN you have an issue - but you are sidestepping the actual simplicity involved.

Ron said...

In the USA it used to be the case that any government-funded research published by the Government did not normally attract copyright.

For example, raher than keping things to themselves, the fruits of the joint US/UK research and development of microwave radar that had been carried out during the war at the MIT Radiation Laboratory were published after WWII as the 28 volume "Radiation Laboratory Handbook". Normally this would have been published under the auspices of the US government and therefore free of copyright. Under the prevailing post war conditions there was insufficient official capacity to do it justice, so it was farmed out to a commercial publisher, under the condition that the copyright term would be restricted to 10 years.

I don't know what the present situation is with US government-funded works. You certainly see patents filed on behalf of US government departments.

Tim said...

The disclosure in a valid patent must be repeatable by the skilled person without undue effort. So the standard of disclosure may need to be higher than in a scientific paper.

Roufousse T. Fairfly said...


What is your source on the RadLab series' 10-year limited copyright?

Artech House republished them in the 80's/90's, but at a rather exorbitant price.

I own a couple of paperback volumes reprinted in the 1960s by Dover. I don't know whether they did the whole series, or the nature of the license they might have had.

The pride of my bookshelf are original editions of the Waveguide Handbook (Vol. 10) by Nathan Marcuvitz (I salvaged the well worn lab copy with the legendary Cape Cod and Magnetron Anode logo from a previous job. I later learned that the company librarian had scrapped the rest of the collection, despite my begging on both knees), and Radar Aids to Navigations (Vol. 2) by John S. Hall (acquired from a second hand bookshop).

If these really are out of copyright, I might consider scanning vol. 10, which is still relevant today, and place the file on the web.

Anonymous said...


I would (lawyerly) add, "or not."

You forgethat the mythical PHOSITA to whom you charge that capability has full and ever-at-the-ready knowledge of everything published. This makes the notion of "undue efforts" decidedly different for legal effect than the commone parlance would lead one to understand.

Tim said...

I'm not sure that the skilled person trying to repeat an invention can necessarily be assumed to have in front of her the specific disclosure that makes it work in practice. This is different from novelty or obviousness. Do you have an authority for your proposition?

Anonymous said...


Choose any authority that you want that defines PHOSITA.

You do know that there is only one such definition, correct?

Joe Wyse said...

In response to my learned colleague Neil Wilkof's article ,

I do not think that IP laws have a negative effect on the replicability of original research for the following reasons:

1. "Good" research usually goes into "good" Journals , such as Nature, PNAS, Cell, BMJ etc. In fact, in every field some journals are recognized as being pretty definitive and prestigious and other journals are known to be less important and less weighty.
2. Science usually works like this: a group a publishes a piece of work. Other groups draw inferences from the work and push on until the original hypothesis breaks down. Often the first signs of this are the lack of replicability of the original work. There then follows a period of backtracking to find the mistake, error, falsification or misinterpretation, and the process starts again. The quest for scientific truth is unending.
3. It can be assumed that scientists in research institutions that don't share data or the means of obtaining the data are over -competitive or have something to hide. Either way, they will get their come- uppance by not benefitting from collaboration. Non cooperative groups soon gain notoriety.
4. IP that is valuable can be filed as patent applications, or is copyright protected, depending on the subject matter. It is the job of university business development organizations to identify potentially valuable IP and protect them prior to publication in a scientific journal. It is remarkably inexpensive to draft and file a provisional patent application in the US, good for a year's worth of priority more or less all over the world. During that period, the original researcher can publish, and give data or software for verification , the IP having been safeguarded.

It seems to me that the protective mechanisms inherent in IP law (patents and copyright) can actually be harnessed by journals and research institutions of repute to increase disclosure for the purposes of replicability in significant areas.

Ron said...

The 10 year term is mentioned in the preface to some, but not all, volumes of the set. I have a complete set myself, but it is presently in store, so I can't quickly state which volumes it appears in.

However, there is a specific reference to the 10 year term in the publication"MIT Libraries' News" Spring 2000 issue, pages 1 and 3, available on line here:

" .... But the stipulation was added that McGraw-Hill place the works into the public domain within a short number of years in order to assure the typical open and free access to government-funded research. Letters between MIT President Karl T.Compton and the McGraw-Hill Book Company reveal the high level of cooperation and agreement which brought about the unusual terms of publication. ...."

Other than the prefaces referred to above, this is the only mention I have found of the copyright-free status of this publication, which remained the microwave engineers' bible for more than 50 years.

I think that there may already be some copies on the web. I came across a web site some time ago that was asking for volunteers to assist in a project to scan the images that appear in the original printed volume at better quality, but can't trace the source now.

Ron said...

On re-reading the 2000 MIT News item, I find it does not actually state the term, but only refers to the "short number of years". The 10 year term is what I recall having read in the prefaces. Possibly this information was omited from the prefaces of subsequent editions?

My own set is a well-used first edition, salvaged from the skip when I was working in industrial practice, and stamped "Marconi's Wireless Telegraphy Company".

Anonymous said...

Why must it be completely protected or not at all?

Why not legislate that a company or a researcher may expect to receive up to a dollar amount but that at that dollar amount they must accept the offer or even set up a system of arbitration guaranteeing that they sell it for some price?

It is simply not acceptable to allow people to just refuse no matter the price when this may hold back the progress of research and innovation. Certainly if the government has the power to establish exclusive use rights it can also establish less-than-exclusive use rights, that is limited rather than absolute intellectual property rights. In fact the right to do so would be inherent in the government's "eminent domain" powers.

Subscribe to the IPKat's posts by email here

Just pop your email address into the box and click 'Subscribe':