by John Sheridan

There is no other form of written texts quite like legislation, nor a form so suited to the web. In retrospect, readers of legislation had been waiting a long time for a hypertext system, such as the web, to be invented. It turns out, we needed one all along. The web has been transformative, primarily in terms of its impact to enable access to the law. Ready and easy public access to legislation (ideally, up to date, revised versions of the text) is of crucial importance to the rule of law, as Carl Malamud wrote in his Birthday Message as part of this series.

Widening access

The web has widened access to the law beyond all recognition in just a couple of decades. As services like legislation.gov.uk demonstrate, non-lawyers can and do routinely consult primary sources of law. We see millions of visits every month to legislation.gov.uk and we know most users of our service are not legally trained or qualified. This is to be celebrated, not least as an important contribution to the rule of law. In publishing the first legislation on the web in the early 1990s, Tom Bruce and his team at Cornell LII saw the potential of the new technology very early on and have blazed a trail ever since. We are still working through the implications of that transformation.

Legislation as data

Part of this transformation is to think about legislation as data. Anne Washington writes eloquently in her post Documents to Data – A Legal Memex, this is in itself an enormous leap forward. As the official publisher of legislation in the UK, we designed legislation.gov.uk to be both an end user service and to provide legislation data as infrastructure through an open API. We deliver legislation as open data, which others can readily access, use and re-use in their own products and services.

Non-linear content

There’s another story here too. Despite the drafter’s best efforts to create a narrative structure that tells a story through the flow of provisions, legislation is intrinsically non-linear content. It positively lends itself to a hypertext based approach. The need for legislation to escape the confines of the printed form predates the all major innovators and innovations in hypertext, from Vannevar Bush’s vision in ” As We May Think“, to Ted Nelson’s coining of the term “hypertext”, through to and Berners-Lee’s breakthrough world wide web. I like to think that Nelson’s concept of transclusion was foreshadowed several decades earlier by the textual amendment (where one Act explicitly alters – inserts, omits or amends – the text of another Act, an approach introduced to UK legislation at the beginning of the 20th century).

Function and form

There’s long been a relationship between the framing of law and the form of law – “we shape our tools and they in turn shape us” so to speak. England has a body of statute law dating back to the 13th century. Not least through the records we hold at The National Archives, it is possible to trace the evolution of the related concepts of Parliament, Law and Government over 800 years – and through that evolution see the impact of the technologies used to encode the law. From the Statute of Marlborough (the statute being the physical roll onto which the agreements of the Marlborough Parliament were written – only later, in part thanks to printing, did we come to view these portions of texts as separate Acts), the modern form of legislation has emerged gradually over time.

The printing press had an enormous impact on the form of law, codified for the those of us with the Westminster system, in the 19th century by Lord Thring in his book Practical Legislation. More recently, modern drafting tools, that enable the easy management of cross-references within and between texts, have enabled far more deeply entwined laws. The complexities of the modern world have given us deeply intertwingled laws.

Legislation is deeply intertwingled

Today’s statutes can be conceptually dense. Yet there’s no rhetoric or repetition to aid the reader, nor, in the UK at least, are there often purpose provisions, examples or diagrams to elucidate the meaning. Legislating in a common law environment, an Act of Parliament, even a principal Act, does not so much set out the law in an area, as set out the changes to an existing body of law – changes across both statute and common law. Statutes rely on the existing law as well as change it. They will use terminology around which a body of pre-existing law exists, either through Interpretation Acts, or established case law for terms such as “reasonable”, or “having regard to”. Moreover, they will be silent on topics where the existing law does not need to change. The intertwingled nature of modern statute law provides a significant barrier to understanding and comprehension, for lawyers and non-lawyers alike. As Alison Bertlin writes in her Loophole article, What works best for the reader? A study on drafting and presenting legislation, “Readers seem to have very little grasp of how legislation is structured and organised. Their ‘mental model’ of it is simply not very good.” It is here that we publishers can best make a contribution to accessible law through how we present legislation on the web – retrofitting the mental model the reader lacks, through the user interfaces we design.

Limits on machine processable laws

As a technologist by background, I recognise the desire to think of the body of legislation as being the Operating System for a country. In many ways, I like the analogy and we are making progress towards machine processable laws. There are valuable standardisation efforts like LegalRuleML, which seeks to define a rule interchange language for the legal domain. The aim is to enable the structuring, evaluation and comparison of legal arguments represented using the rule language. In the era of digital government this idea has real allure. There are potentially enormous benefits and efficiency gains to be had from running legislation as code. However, I often wonder whether these efforts perhaps miss the nature of the law, or at least the common law. The intertwingled nature of the law – a feature that aids the common law’s resilience and ultimately underpins our notion of justice – poses significant engineering challenges. There may be better prospects for machine processable legislation where there are formal legal codes, which sets out all the law in a given domain. It will be interesting whether legislation becomes more codified, modular and self-contained in future, in a drive towards machine processability.

There is much still to do

We have turned our legislation into hypertext and massively widened access, but we have yet to escape the printed form of legislation or to realise the full benefits. This is compounded by the limitations of the web as a hypertext system and the complexity of legislation – imagine what we could do if we had something closer to Ted Nelson’s original vision! It is a revolution still underway, and for all of us interested in legal informatics, there is much still to do.

John Sheridan is the Digital Director for the National Archives of the United Kingdom.  He was previously Head of Legislation Services at The National Archives, where he led the team responsible for developing legislation.gov.uk.

by Anne Washington

When the United States first mandated the preservation of government proceedings, clerks wrote in bound blank books. In 1789, Congress passed a law providing for the “safe-keeping” of legislative and government documents (1 Stat. 68).

When the Cornell Legal Information Institute started, digital information was shared on CD-ROMs and 3-1/2 inch disks. In 1992, two people on a cold day in upstate New York turned on a server box to provide law on the Internet.

To celebrate the remarkable work of the Cornell Legal Information Institute in its 25th year and the 228th year of the first US Congress, this is a provocation about what we might celebrate in future.

Vannaver Bush in 1945 imagined a machine he called a Memex that would enable the instant retrieval of “trails” of memory. He briefly discussed how it could apply to the law: “The lawyer has at his touch the associated opinions and decisions … The patent attorney has on call the millions of issued patents, with familiar trails to every point of his client’s interest.” (Bush, 1945, p 8)

What features might be in a future legal Memex ?

Imagine that it is 2042 and the 50th anniversary of LII.

The Vice President is listening to a debate on the Senate floor about a bill related to a campaign issue. She tracks public opinion on the social media site Face-oogle, particularly noting what the Majority Leader’s constituents are saying. More importantly, she receives an automated predictive score from the Congressional Data & Budget Office (CDB) about which existing laws might be impacted by a proposed amendment.

She focuses on the predicted laws that meet her preferred portfolio factors. She realizes that this would change an important law. She requests 3-5 potential placements of the amendment in the United States Code and creates composite texts. After comparing the composite text with the official United States Code source, she is ready to spread the word.

The daughter of the first librarian in the White House notifies the Attorney General.

The Attorney General quickly turns to the Legal Information Institute which now tracks all USC citations and their use in case law. The law has been used in several appeals court cases. LII has a detailed description of a pending case before the US Supreme Court that is related to this law. The Attorney General specifically notes the cases that have appeared before the appeals courts associated with the Majority Leader’s state.

In a brief summary, the Attorney General gives legal advice about potential enforcement issues. The message contains the proposed amendment, as it would appear in the United States Code if passed.

In an unprecedented continuation of two American political dynasties, Vice President Barbara Pierce Bush, sends a confidential instant message to President Mahlia Obama. The two are known for their ability to bridge by partisan divides through quick access to open government documents.

The President wants to understand the impact of the amendment on administrative law. A quick LII search provides a list of federal regulations that may be related to changes in the United States Code. In combination with internal government sources, her assistant is able to identify pending Federal Regulations that might be impacted by the amendment. He reviews the historic point-in-time directories of the Code of Federal Regulation. He completes an analysis of how this amendment might streamline workflows or create conflicts between agencies.

The Vice-President reaches across the aisle to the Majority Leader and shares her analysis. They agree on slight changes to the language in order to meet constituent opinion, address the policy concerns, and to keep the government running effectively.

When the amendment passes, LII notifies all lawyers, who have argued or written about the law and who have opted-in to notification services. The Senate immediately publishes everything as open data and it appears on G! government entertainment live-casting.

My legal Memex builds a network of the people and laws available in the public records of politicians and organizations. The infrastructure for this vision relies on open data, free access to law, and instantaneously availability. The text analysis and machine learning assumes a neutral analysis that is both defendable and leaves room for interpretation.

How far are we from meeting this vision?

As a scholar of legislative organizations, I consider the necessary institutional foundations for a successful legal information institute (LII). The past gives us encouragement. The past also shows how some fundamental principles are necessary to create reliable, accurate, timely, and authoritative data.

A few examples in the history of the United States show the progression necessary for creating the basis for the transition from documents to data.

Two weeks before the end of the first Congress, the chambers pass a law which orders the printing of government records. By the 13th Congress, they realized that distribution is just as important as printing. The Federal Depository Library Program (FDLP), which ensures that documents and skilled librarians are available in every jurisdiction, has its origins in 3 Stat 140 (1813).

The idea that there was too much legal information for any one person to track was evident by the early 20th century. The United States Code, finished in 1926, provided subject access to general and permanent federal laws. Congress thought it was a good idea to create indexes and summaries of all pending legislation as well. The Digest of Public General Bills, which first appeared in 1936, continued to be published in annual volumes through 1990 when it was subsumed into an online database. Congress, as early as 1999, embraced structured data formats so today we have at least the United States Code, votes, legislation, administrative law (Code of Federal Regulations) published in XML.

Law must be documented, printed, distributed, indexed, and structured. While important to the free access to law movement, these developments are equally vital to the internal procedures of public sector organizations. The legislative, judicial, and executive branches could learn from each other and grow together.

The transition from documents to data is vital to modernizing the functioning of our centuries-old bureaucracies around the world, as my other 25 for 25 authors have attested. The projects initiated by Sarah Frug, Sylvia Kwakye and others working on LII in 2017 lead the way with innovative solution for tomorrow.

Law and legislation do not sit still. For instance, the promulgation of laws in the form of the US Records Act was updated in 1795, 1796, 1814, 1816, 1820 and 1842. We need patience as we move forward. Policy has always been iterative.

It is time for the next generation of legal, policy, and technology pioneers to move the free access to law movement forward.

What is your ideal legal Memex?

Anne L. Washington is a digital government scholar who specializes in informatics and technology management. Her expertise on government data currently addresses the emerging policy and governance needs of data science. The National Science Foundation has funded her research on open government data and data-intensive political science. Her work draws on both interpretive research methods and computational text analysis. She is an Assistant Professor at the Schar School of Policy and Government at George Mason University in Arlington, VA, where she teaches organizational ethnography, socio-technical analysis, and electronic government. Prof. Washington serves on the Advisory Board of the Electronic Privacy Information Center (EPIC) and the Open Government Foundation. She has also served on the United Nations World E-parliament Working Group on XML in Parliament, the Oasis LegalXML technical committee of citations, and Federal Web Content Managers Usability Task Force. She was an invited expert to the W3C E-Government Interest Group and the W3C Government Linked Data Working Group.She holds a Bachelors of Arts (BA) in computer science from Brown University, and a Masters in Library Information Science (MLIS) from Rutgers University. She earned a PhD in Information Systems and Technology Management from The George Washington University School of Business with a secondary field in Organization Behavior.  Prior to completing her doctorate, she had extensive work experience in information architecture and information technology after years with the Congressional Research Service at the Library of Congress, Barclays Global Investors, Wells Fargo Nikko Investment Advisors and Apple Computer.

 

[ Ed. note:  two of Professor Perritt’s papers have strongly influenced the LII.  We recommend them here as “extra credit” reading.  “Federal Electronic Information Policy” provides the rationale on which the LII has always built its innovations.  “Should Local Governments Sell Local Spatial Databases Through State Monopolies?” unpacks issues around the resale of public information that are still with us (and likely always will be )]

In 1990 something called the “Internet” was just becoming visible to the academic community. Word processing on small proprietary networks had gained traction in most enterprises, and PCs had been more than toys for about five years. The many personal computing magazines predicted that each year would be “the Year of the LAN”–local area network–but the year of the LAN always seemed to be next year. The first edition of my How to Practice Law With Computers, published in 1988, said that email could be useful to lawyers and had a chapter on Westlaw and Lexis. It predicted that electronic exchange of documents over wide-area networks would be useful, as would electronic filing of documents, but the word “Internet” did not appear in the index. My 1991 Electronic Contracting, Publishing and EDI Law, co-authored with Michael Baum, focused on direct mainframe connections for electronic commerce between businesses, but barely mentioned the Internet. The word Internet did appear in the index, but was mentioned in only two sentences in 871 pages.

Then, in 1990, The Kennedy school at Harvard University held a conference on the future of the Internet. Computer scientists from major universities, Department of Defense officials, and a handful of representatives of commercial entities considered how to release the Internet from its ties to defense and university labs and to embrace the growing desire to exploit it commercially. I was fortunate to be on sabbatical leave at the Kennedy School and to be one of the few lawyers participating in the conference. In a chapter of a book published by Harvard afterwards on market structures, I said, “A federally sponsored high-speed digital network with broad public, non-profit and private participation presents the possibility of a new kind of market for electronic information products, one in which the features of information products are ‘unbundled’ and assembled on a network.”

The most important insight from the 1990 conference was that the Internet would permit unbundling of value. My paper for the Harvard conference and a law review article I published shortly thereafter in the Harvard Journal of Law and Technology talked about ten separate value elements, ranging from content to payment systems, with various forms of indexing and retrieval in the middle. The Internet meant that integrated products were a thing of the past; you didn’t have to go inside separate walled gardens to shop. You didn’t have to pay for West’s key numbering system in order to get the text of a judicial opinion written by a public employee on taxpayer time. Soon, you wouldn’t have to buy the whole music album with nine songs you didn’t particularly like in order to get the one song you wanted. Eventually, you wouldn’t have to buy the whole cable bundle in order to get the History Channel, or to be a Comcast cable TV subscriber to get a popular movie or the Super Bowl streamed to your mobile device.

A handful of related but separate activities developed some of the ideas from the Harvard conference further. Ron Staudt, Peter Martin, Tom Bruce,  and I experimented with unbundling of legal information on small servers connected to the Internet to permit law students, lawyers, and members of public to obtain access to court decisions, statutes, and administrative agency decisions in new ways. Cornell’s Legal Information Institute was the result.

David Johnson, Ron Plesser, Jerry Berman, Bob Gellman, Peter Weiss, and I worked to shape the public discourse on how the law should channel new political and economic currents made possible by the Internet. Larry Lessig was a junior recruit to some of these earliest efforts, and he went on to be the best of us all in articulating a philosophy.

By 1996, I wrote a supplement to How to Practice Law With Computers, called Internet Basics for Lawyers, which encouraged lawyers to use the Internet for email and eventually to deliver legal services and to participate in litigation and agency rulemaking and adjudication. In the same year, I published a law review article explaining how credit-card dispute resolution procedures protected consumers in ecommerce.

One by one, the proprietary bastions fell—postal mail, libraries, bookstores, department stores, government agency reading rooms–as customers chose the open and ubiquitous over the closed and incompatible. Now, only a few people remember MCImail, Western Union’s EasyLink, dial up Earthlink, or CompuServe. AOL is a mere shadow of its former self, trying to grab the tail of the Internet that it too long resisted. The Internet gradually absorbed not only libraries and government book shops but also consumer markets and the legislative and adjudicative processes. Blockbuster video stores are gone. Borders Books is gone. The record labels are mostly irrelevant. Google, Amazon, and Netflix are crowding Hollywood. Millions of small merchants sell their goods every second on Amazon and eBay The shopping malls are empty. Amazon is building brick-and-mortar fulfillment warehouses all over the place. Tens of millions of artists are able to show their work on YouTube.

Now the Internet is well along in absorbing television and movies, and has begun absorbing the telephone system and two-way radio. Video images move as bit streams within IP packets. The rate at which consumers are cutting the cord and choosing to watch their favorite TV shows or the latest Hollywood blockbusters through the Internet is dramatic.

Television and other video entertainment are filling up the Internet’s pipes. New content delivery networks bypass the routers and links that serve general Internet users in the core of the Internet. But the most interesting engineering developments relate to the edge of the Internet, not its core. “Radio Access Networks,” including cellphone providers, are rushing to develop micro-, nano-, pico-, and femto-cells beyond the traditional cell towers to offload some of the traffic. Some use Wi-Fi, and some use licensed spectrum with LTE cellphone modulation. Television broadcasters meanwhile are embracing ATSC 3.0, which will allow their hundred megawatt transmitters to beam IP traffic over their footprint areas and – a first for television – to be able to charge subscribers for access.

The telephone industry and the FCC both have acknowledged that within a couple of years the public telephone system will no longer be the Public Switched Telephone System; circuit switches will be replaced completely by IP routers.

Already, the infrastructure for land mobile radio (public safety and industrial and commercial services) comprises individual handsets and other mobile transceivers communicating by VHF and UHF radio with repeater site or satellites, tied together through the Internet.

Four forces have shaped success: Conservatism, Catastrophe forecasts, Keepers of the Commons, and Capitalism. Conservatism operates by defending the status quo and casting doubt about technology’s possibilities. Opponents of technology have never been shy. A computer on every desk? “Never happen,” the big law firms said. “We don’t want our best and brightest young lawyers to be typists.”

Communicate by email?  ”It would be unethical,” dozens of CLE presenters said. “Someone might read the emails in transit while they are resting on store-and-forward servers.” (The email technology of the day did not use store-and-forward servers.).

Buy stuff online? “It’s a fantasy,” the commercial lawyers said. “No one will trust a website with her credit card number. Someone will have to invent a whole new form of cybermoney.”

Catastrophe has regularly been forecast. “Social interaction will atrophy. Evil influences will ruin our kids. Unemployment will skyrocket,” assistant professors eager for tenure and journalists jockeying to lead the evening news warned. “The Internet is insecure!” cybersecurity experts warned. “We’ve got to stick with paper and unplugged computers.” The innovators went ahead anyway and catastrophe did not happen. A certain level of hysteria about how new technologies will undermine life is normal. It is always easier to ring alarm bells than to understand the technology and think about its potential.

Keepers of the Commons—the computer scientists who invented the Internet—articulated two philosophies, which proved more important than engineering advances in enabling the Internet to replace one after another of preceding ways of organizing work, play, and commerce.  To be sure, new technologies mattered. Faster, higher quality printers were crucial in placing small computers and the Internet at the heart of new virtual libraries, first Westlaw and Lexis and then and then Google and Amazon. Higher speed modems and the advanced modulation schemes they enabled made it faster to retrieve an online resource than to walk to the library and pull the same resource off the shelf. One-click ordering made e-commerce more attractive. More than 8,000 RFCs define technology standards for the Internet.

The philosophies shaped use of the technologies.  The first was the realization that an open architecture opens up creative and economic opportunities for millions of individuals and small businesses that otherwise were barred from the marketplace by high barriers to entry. Second was the realization that being able to contribute creatively can be powerful motivators for activity, alongside expectations of profit. The engineers who invented the Internet have been steadfast in protecting the commons: articulating the Internet’s philosophy of indifference to content, leaving application development for the territory beyond its edges, and contributing untold hours to the development of open standards called “Requests for Comment” (“RFCs”). Continued work on RFCs, services like Wikipedia and LII, and posts to YouTube show that being able to contribute is a powerful motivator, regardless of whether you make any money. Many of these services prove that volunteers can add a great deal of value to the marketplace, with higher quality, often, than commercial vendors.

Capitalism has operated alongside the Commons, driving the Internet’s buildout and flourishing as a result. Enormous fortunes have been made in Seattle and Silicon Valley. Many of the largest business enterprises in the world did not exist in 1990.

Internet Capitalism was embedded in evangelism. The fortunes resulted from revolutionary ideas, not from a small-minded extractive philosophy well captured by the song “Master of the House” in the musical play, Les Miserables:

Nothing gets you nothing, everything has got a little price

Reasonable charges plus some little extras on the side

Charge ‘em for the lice, extra for the mice,

Two percent for looking in the mirror twice

Here a little slice, there a little cut,

Three percent for sleeping with the window shut.

Throughout most of the 1990s, the established, legacy firms were Masters of the House, unwilling to let the smallest sliver of an intellectual property escape their clutches without a payment of some kind. They reinforced the walls around their asset gardens and recruited more tolltakers than creative talent. The gates into the gardens work differently, but each charges a toll.

Meanwhile, the Apples, Googles, and Amazons of the world flourished because they offered something completely different–more convenient and more tailored to the way that consumers wanted to do things. Nobody ever accused Steve Jobs of giving away much for free or being shy in his pricing, but he made it clear that when you bought something from him you were buying something big and something new.

The tension between Commons and Capitalism continues. In the early days, it was contest between those who wanted to establish a monopoly over some resource – governmental information such as patent office records, Securities and Exchange Commission filings, or judicial opinions and statutes –and new entrants who fought to pry open new forms of access. Now the video entertainment industry’s  Master of the House habits are getting in the way of the necessary adaptation to cord cutting, big time. The video entertainment industry is scrambling to adapt its business models.

Intellectual property law can be an incentive to innovate, but it also represents a barrier to innovation. Throughout much of the 1980s, when the Internet was taking shape, law was uncertain whether either patent or copyright was available for computer software. Yet new businesses by the hundreds of thousands flocked to offer the fruits of their innovative labors to the marketplace. To be sure, as Internet related industries matured, their managers and the capital markets supporting seek copyright and patent protection of assets to encourage investment.

Whether you really believe in a free market depends on where you sit at a particular time. When you have just invented something, you think a free market is great as you try to build a customer base. Interoperability should be the norm. Once you have a significant market share, you think barriers to entry are great, and so do your potential investors. Switching costs should be as high as possible.

The Master of the House still operates his inns and walled gardens. Walled gardens reign supreme with respect to video entertainment. Popular social media sites like Facebook, Snapchat, Twitter, and YouTube are walled gardens. Technologically the Master of the House favors mobile apps at the expense of mobile web browsers; it’s easier to lock customers in a walled garden with an app; an app is a walled garden.

An Internet architecture designed to handle video entertainment bits in the core of the Internet will make it more difficult to achieve net neutrality. CDNs are private networks, outside the scope of the FCC’s Open Network order. They are free to perpetuate and extend the walled gardens that Balkanize the public space with finely diced and chopped intellectual property rights.

Net Neutrality is fashionable, but it also is dangerous. Almost 100 years of experience with the Interstate Commerce Commission, the FCC, and the Civil Aeronautics Board shows that regulatory schemes adopted initially to ensure appropriate conduct in the marketplace also make the coercive power of the government available to legacy defenders of the status quo who use it to stifle innovation and the competition that results from it.

It’s too easy for some heretofore un-appreciated group to claim that it is being underserved, that there is a new “digital divide,” and that resources need to be devoted primarily to making Internet use equitable. In addition, assistant professors seeking tenure and journalists seeking the front page or to lead the evening news are always eager to write stories and articles about how new Internet technologies present a danger to strongly held values. Regulating the Internet like telephone companies provide a well-established channel for these political forces to override economic and engineering decisions.

Security represents another potent threat. Terrorists, cyberstalkers, thieves, spies, and saboteurs use the Internet – like everyone else. Communication networks from the earliest days of telegraph, telephones, and radio, have made the jobs of police and counterintelligence agencies more difficult. The Internet does now. Calls for closer monitoring of Internet traffic, banning certain content, and improving security are nothing new. Each new threat, whether it be organization of terrorist cells, more creative email phishing exploits, or Russian interference in American elections intensifies calls for restrictions on the Internet. The incorporation of the telephone system and public safety two-way radio into the Internet will make it easier for exaggerated concerns about network security to make it harder to use the Internet. Security can always be improved by disconnecting. It can be improved by obscuring usefulness behind layers of guards. The Internet may be insecure, but it is easy to use.

These calls have always had a certain level of resonance with the public, but so far have mostly given way to stronger voices protecting the Internet’s philosophy of openness. Whether that will continue to happen is uncertain, given the weaker commitment to freedom of expression and entrepreneurial latitude in places like China, or even some places in Europe. Things might be different this time around because of the rise of a know-nothing populism around the world.

The law has actually has had very little to do with the Internet’s success. The Internet has been shaped almost entirely by entrepreneurs and engineers. The two most important Internet laws are shields: In 1992, my first law review article on the Internet, based on my participation in the Harvard conference said:

Any legal framework . . . should serve the following three goals: (1) There should be a diversity of information products and services in a competitive marketplace; this means that suppliers must have reasonable autonomy in designing their products; (2) users and organizers of information content should not be foreclosed from access to markets or audiences; and (3) persons suffering legal injury because of information content should be able to obtain compensation from someone if they can prove traditional levels of fault.

It recommended immunizing intermediaries from liability from harmful content as long as they acted like common carriers, not discriminating among content originators—a concept codified in the safe harbor provisions of the Digital Millennium Copyright Act and section 230 of the Communications Decency Act, both of which shield intermediaries from liability for allegedly harmful content sponsored by others. In that article and other early articles and policy papers, I urged a light touch for regulation.

David Johnson, David Post, Ron Plesser, Jack Goldsmith, and I used to argue in the late 1990s about whether the world needed some kind of new some kind of new Internet law or constitution. Goldsmith took the position that existing legal doctrines of tort, contract, and civil procedure were perfectly capable of adapting themselves to new kinds of Internet disputes. He was right.

Law is often criticized for being behind technology. That is not a weakness; it is a strength. For law to be ahead of technology stifles innovation. What is legal depends on guesses lawmakers have made about the most promising directions of technological development. Those guesses are rarely correct. Law should follow technology, because only if it does so will it be able to play its most appropriate role of filling in gaps and correcting the directions of other societal forces that shape behavior: economics, social pressure embedded in the culture, and private lawsuits.

Here is the best sequence: a new technology is developed. A few bold entrepreneurs take it up and build it into their business plans. In some cases it will be successful and spread; most cases it will not. The technologies that spread will impact other economic players. It will threaten to erode their market shares; it will confront them with choosing new technology if they wish to remain viable businesses; it will goad them into seeking innovations in their legacy technologies.

The new technology will probably cause accidents, injuring and killing some of its users and injuring the property and persons of bystanders. Widespread use of the technology also will have adverse effects on other, intangible interests, such as privacy and intellectual property. Those suffering injury will seek compensation from those using the technology and try to get them to stop using it.

Most of these disputes will be resolved privately without recourse to governmental institutions of any kind. Some of them will find their way to court. Lawyers will have little difficulty framing the disputes in terms of well-established rights, duties, privileges, powers, and liabilities. The courts will hear the cases, with lawyers on opposing sides presenting creative arguments as to how the law should be understood in light of the new technology. Judicial decisions will result, carefully explaining where the new technology fits within long-accepted legal principles.

Law professors, journalists, and interest groups will write about the judicial opinions, and, gradually, conflicting views will crystallize as to whether the judge-interpreted law is correct for channeling the technology’s benefits and costs. Eventually, if the matter has sufficient political traction, someone will propose a bill in a city council, state legislature, or the United States Congress. Alternately, an administrative agency will issue a notice of proposed rulemaking, and a debate over codification of legal principles will begin.

This is a protracted, complex, unpredictable process, and that may make it seem undesirable. But it is beneficial, because the kind of interplay that results from a process like this produces good law. It is the only way to test legal ideas thoroughly and assess their fit with the actual costs and benefits of technology as it is actually deployed in a market economy.

A look backwards warrants optimism for the future, despite new or renewed threats. The history of the Internet has always involved argument between those who said it would never take off because people would prefer established ways of doing business. It has always been subjected to various economic and legal attempts to block its use by new competitors. The Master of the House has always lurked. Shrill voices have always warned about its catastrophic social effects. Despite these enemies, it has prevailed and opened up new pathways for human fulfilment. The power of that vision and the experience of that fulfillment will continue to push aside the forces that are afraid of the future.

Henry H. Perritt, Jr. is Professor of Law and Director of the Graduate Program in Financial Services Law at the Chicago-Kent School of Law. A pioneer in Federal information policy, he served on President Clinton’s Transition Team, working on telecommunications issues, and drafted principles for electronic dissemination of public information, which formed the core of the Electronic Freedom of Information Act Amendments adopted by Congress in 1996. During the Ford administration, he served on the White House staff and as deputy under secretary of labor.

Professor Perritt served on the Computer Science and Telecommunications Policy Board of the National Research Council, and on a National Research Council committee on “Global Networks and Local Values.” He was a member of the interprofessional team that evaluated the FBI’s Carnivore system. He is a member of the bars of Virginia (inactive), Pennsylvania (inactive), the District of Columbia, Maryland, Illinois and the United States Supreme Court. He is a published novelist and playwright.

 

From an equally long time ago, and in one of those galaxies so far far away it is sometimes mistaken for the mythical Oz, we received Tom Bruce’s call for reflection on the history of free access to legal information. “Here’s what we *thought* we were doing, and here’s what it really turned into”, he suggested, so I have taken him up on that. Andrew Mowbray and I started the Australasian Legal Information Institute (AustLII) in 1995, and our second employee, Philip Chung, now AustLII’s Executive Director, joined us within a year. We are still working together 22 years later.

AustLII had a back-story, a preceding decade of collaborative research from 1985, in which Andrew and I were two players in the first wave of ‘AI and law’ (aka ‘legal expert systems’). Our ‘DataLex Project’ research was distinctive in one respect: we insisted that ‘inferencing systems’ (AI) could not be a closed box, but must be fully integrated with both hypertext and text retrieval (for reasons beyond this post). Andrew wrote our own search engine, hypertext engine, and inferencing engine; we developed applications on IP and on privacy, and had modest commercial success with them in the early 90s. Tools for relatively large-scale automation of mark-up of texts for hypertext and retrieval purposes were a necessary by-product. In that pre-Web era, when few had CD ROM drives, and free access to anything was impractical and unknown, products were distributing on bundles of disks. Our pre-Web ideology of ‘integrated legal information systems’ is encapsulated in a 1995 DataLex article. But a commercial publisher pulled the plug on our access to necessary data, and DataLex turned out to have more impact in its non-commercial after-life as AustLII.

Meanwhile, in January 1995 Andrew and I (for UTS and UNSW Law Schools) had obtained a grant of AUD $100,000 from the Australian Research Council’s research infrastructure fund, in order to explore the novel proposition that the newly-developing World-Wide-Web could be used to distribute legal information, and for free access, to assist academic legal research. A Sun SPARCstation, one ex-student employee, and a part-time consultant followed. Like Peter & Tom we sang from Paul Simon’s text, ‘let’s get together and call ourselves an Institute’, because it sounded so established.

What were we thinking? (and doing)

What were we thinking when we obtained this grant, and put it into practice in that first year? We can reconstruct this somewhat, not simply from faulty memories, but from what we actually did, and from our first article about AustLII in 1995, which contained something of a manifesto about the obligations of public bodies to facilitate free access to law. So here are things we did think we were doing in 1995 – no doubt we also had some bad ideas, now conveniently forgotten, but these ones have more or less stuck.

1. End monopolies – Australia had been plagued for a decade by private sector and public sector monopolies (backed by Crown copyright) over computerised legal data. Our core principle was (polite) insistence on the ‘right to republish’ legislation, cases, and other publicly funded legal information. We appropriated our first large database (Federal legislation), but got away with it. The High Court told the federal government to supply ‘its cases’ to AustLII, and other courts followed.
2. Rely on collaboration – Our 1995 ‘manifesto’ insisted that courts and legislative offices should provide the best quality data available to all who wished to republish it. Insistence on collaboration was a survival strategy, because we would never have enough resources to manage any other way. From the start, some courts started to email cases, and adopt protocols for consistent presentation, and eventually all did so.
3. Disrupt publishing – Much Australian commercial legal publishing in 1995 was not much more than packaging raw legal materials, with little added value, for obscene prices. We stated that we intended to force 2nd-rate publishing to lift its game (‘you can’t compete with free’). It did, and what survived, thrived.
4. Stay independent – While we had material support from our two Law Schools, and an ARC start-up grant, we tried from the start to be financially independent of any single source. Within a year we had other funds from a Foundation, and a business group (for industrial law), and were negotiating funding from government agencies. Later, as the funds needed for sustainability became larger, this was much more of a challenge. However, independence meant we could publish any types of content that we could fund, with no one else dictating what was appropriate. A 93 volume Royal Commission report on ‘Aboriginal deaths in custody’ for which the federal government had ‘lost’ the master copy was an early demonstration of this.
5. Automate, integrate, don’t edit – The DataLex experience gave us good tools for complex automated mark-up of large sets of legislation, cases etc. Collaboration in data supply from official bodies multiplied the effect of this. We edited documents only when unavoidable. Sophisticated hypertexts also distinguished the pioneering work of the LII (Cornell) and LexUM from the chaff of commercial publishers. AustLII inherited from DataLex a preoccupation with combining the virtues of hypertext and text retrieval, most apparent from day 1 in the ‘Noteup’ function.
6. Cater for all audiences – Our initial grant’s claim to serve academic research was only ever a half-truth, and our intention was to try to build a system that would cater for all audiences from practitioners to researchers to the general public. The LII (Cornell) had already demonstrated that there was a ‘latent legal market’, an enormous demand for primary legal materials from the public at large.
7. All data types welcome – We believed that legislation, cases, treaties, law reform, and some publicly-funded scholarship should all be free access, and a LII should aim to provide them, as its resources allowed. This was a corollary of aiming to ‘serve all audiences’. In AustLII’s first year we included examples of all of these (and a Royal Commission report), the final element being the Department of Foreign Affairs agreement to partner a Treaties Library. It took us much longer to develop serious ‘law for the layperson’ content.
8. ‘Born digital’ only – In 1995 there was already more digital data to gather than AustLII could handle, and scanning/OCR’ing data from paper was too expensive and low quality, so we ignored it, for more than a decade.
9. ‘Comprehensiveness’ – As Daniel Poulin says in this series, AustLII was first to aim to create a nationally comprehensive free access system, or to succeed.  But the initial aims of comprehensiveness were limited to the current legislation of all 9 Australian jurisdictions, and the decisions of the superior courts of each. That took 4 years to achieve.  Addition of decisions of all lower courts and tribunals, and historical materials, were much later ambitions, still not quite achieved.
10. ‘Australasian’ but ‘LII’ – We asked Cornell if we could borrow the ‘LII’ tag, and had vague notions that we might be part of a larger international movement, but no plans.  Our 1995 article exaggerates in saying ‘AustLII is part of the expanding international network of public legal information servers’ – we wished! However, the ‘Australasian’ aim was serious: NZLII’s superb content is a major part of AustLII, but PNG content found a better home on PacLII.
11. Neutral citations, backdated – As soon as AustLII started receiving cases, we applied our own ‘neutral citations’ (blind to medium or publisher) to them, and applied this retrospectively to back-sets, partly so that we could automate the insertion of hypertext links. As in Canada, this was a key technical enabler. A couple of years later, the High Court of Australia led the Council of Chief Justices to adopt officially a slight variation of what AustLII had done (and we amended our standard). The neutral citation standard set with ‘[1998] HCA 1’ has since been  adopted in many common law countries. AustLII has applied it retrospectively as a parallel citation, for example ‘[1220] EngR 1’ and so on. Later, the value of neutral citations as a common-law-wide interconnector enabled the LawCite citator.
12. Reject ‘value-adding’ – We saw invitations to distinguish ‘value-added’ (now ‘freemium’ or chargeable) services  from  AustLII’s ‘basic’ free content as a slippery slope, a recipe for free access always being second rate. So AustLII has stayed 100% free access content, including all technical innovations.
13. ‘Free’ includes free from surveillance – Access was and is anonymous with no logins, cookies, advertisements or other surveillance mechanisms beyond logging of IP addresses. We used the Robot Exclusion Standard to prevent spidering/searching of case law by Google etc, and most (not all) other LIIs have done likewise. This has helped establish a reasonable balance between privacy and open justice in many common law jurisdictions. It also helps prevent asset stripping – AustLII is a free access publisher, not  a repository.

This ‘bakers dozen’ aspirations comes from another century, but the issues and questions they address still need consideration by anyone aiming to provide free access to law.

Why we were lucky

In at least four respects, we did not know how fortunate we were in Australia: the Australian Research Council awarded annual competitive grant funding for development of research infrastructure, not just for research; all Australian law schools were willing to back AustLII as a joint national facility (already in 1995 ‘supported by the Australian Committee of Law Deans’); UNSW and UTS Law Faculties backed us with both material assistance and academic recognition; later, we obtained charitable status for donations; and our courts never required AustLII to redact cases (contrast Canada and New Zealand), they did it themselves where it was necessary. Our colleagues in other common law jurisdictions were often not so fortunate.

Cornell, LexUM and AustLII were all also fortunate to be better prepared than most commercial or government legal information publishers to take advantage of the explosion of  public usage of the Internet (and the then-new WWW) in 1994/5. None of us were ‘just another publisher’, but were seen as novel developments. Later LIIs did not have this ‘first mover advantage’, and often operated in far more difficult circumstances in developing countries.

Unimaginables

Given what AustLII, and free access to law globally, have developed into, what did we not imagine, back in 1995? Here are a few key unforseens.

Digitisation from paper did not became financially feasible for AustLII until about 2007. Since then, capturing historical data has become a major part of what AustLII does, with results such as the complete back-sets of over 120 non-commercial Australasian law journals,  and almost all Australasian reported cases and annual legislation 1788-1950. The aims of both ‘horizontal’ comprehensiveness of all current significant sources of law, and ‘vertical’ comprehensiveness of past sources, is new and no longer seems crazy nor unsustainable.

We did not envisage the scale of what AustLII would need to manage, whether data (currently 749 Australasian databases, and almost as much again internationally), sources (hundreds of email feeds), page accesses (about 1M per day), or collaborations (daily replication of other LII content), nor the equipment (and funding) demands this scale would pose. Independence allowed us to obtain hundreds of funding contributors for maintenance. Innovative developments are still supported by ARC and other grants. The future holds no guarantees, but as Poulin says, history has now demonstrated that sustainable large-scale LII developments are possible.

While AustLII’s initial aims were limited to Australasia, by the late 90s requests for assistance to create similar free access LIIs involved AustLII, LexUM and the LII (Cornell) in various countries. The Free Access to Law Movement (FALM) has expanded to nearly 70 members, has directly delivered considerable benefits of free access to law in many countries, and has encouraged governments almost everywhere to accept that free access to legislation and cases is now the norm, in a way that it was not in the early 90s. The delivery of free access content by independent LIIs has, for reasons Poulin outlines, turned out to sit more comfortably in common law than in civil law jurisdictions, and no global access to law via a LII framework has emerged. However, although this was not envisaged back in 1995, AustLII has been able to play a coordinating role in a network of collaborating LIIs from many common law jurisdictions, with compatible standards and software, resulting in access via CommonLII to nearly 1500 databases, and to the daily interconnection of their citations via LawCite. This extent of collaboration was not foreseeable in 1995.

Every free access to law provider has a different story to tell, with different challenges to overcome in environments typically much more difficult than Australia. Somewhere in each of our stories there is a corner reserved for the pioneering contributions of Martin, Bruce and the LII at Cornell. The LII (Cornell) continues to innovate 25 years after they set the wheels in motion.

Graham Greenleaf is Professor of Law & Information Systems at UNSW Australia. He was co-founder of AustLII, with Andrew Mowbray, and Co-Director (with Andrew and Philip Chung) until 2016, and is now Senior Researcher.