Prior Information Technologies

PRE-DIGITAL LAW :
HOW PRIOR INFORMATION TECHNOLOGIES
HAVE SHAPED ACCESS TO AND THE NATURE OF LAW

Peter W. MARTIN
Cornell Law School
USA

I- INTRODUCTION

Digital distribution of law is not a future fancy, but present fact. The rapid shift from print to computer-based systems, now well underway, represents visible change for sure, but of what deeper significance? Will the change have other than superficial implications for the set of activities or functions we call law? Is this likely to be a case of same basic activity, new platform, or more unsettling change?

For good reasons, the technology of law is normally background technology, largely transparent to those who deal with it, especially those who deal with it regularly. As a beginning law student, I learned how to do "legal research"; I was not led on an exploration of the history nor the strengths and weaknesses of the legal information system of the early sixties. Having mastered a set of information access techniques, I assumed, as my instructors seemed to, that they would become part of a professional competence that would allow me to do the real work of lawyering.

The change to digital media has forced these tools into the foreground. Today, those who work with law cannot take legal information technology for granted. And our normal inattention to these matters, puts us at a disadvantage in facing change. Without understanding the important questions that have surfaced at past points of transition or how key features of current information tools have evolved, it is difficult to comprehend the issues and opportunities posed by new ones. Furthermore, a failure to note the evolutionary process that has shaped the system we have taken for granted can easily lead to exaggerated or distorted notions of the difficulty of dealing with its imminent change.

This paper undertakes an examination of the past in order to put the issues, challenges, and possibilities of digital law in some perspective. Since objects called "law books" and collections called "law libraries" have existed for hundreds of years there is a natural tendency to project present understandings of those words on that past and believe, as a consequence, that legal information technology has been stable for quite some time. It has not.

The quickest way to break through the "book is a book", "library is a library" fallacy is by examining a few specimens. While on a hike in Yoho National Park, Canada, this past summer, I met and learned from a geologist whose research focused on lake beds. As we climbed by a number of small mountain lakes, he explained how he obtained information about the region's past by boring into the sediment beneath them, studying small samples drawn from a few locations, in detail. His technique involved inferring changes in the lake, climate, and surrounding life forms from what he found at different levels in those samples. In loosely analogous fashion, the following reflections rest on or are at least illustrated by samples drawn at a range of depths from sediment found beneath the lake called "Law".

Before examining these samples, consider briefly what we might be looking for. The area of this exploration, law, has had a long and, may I say, fluid history, interacting with its surroundings in strong but complex ways. Its boundaries have fluctuated over time. The long range dynamic appears almost certainly to have been expansion. The following samples have been drawn not only from different depths but from different locations -- including one proximate to an area we now call by the name "legislature", another next to a growing cluster of "administrative agencies", several from the ancient region known as the "judiciary" and a few from spots in between.

Since the inquiry concerns information technology, rather than particular law content, analysis of these samples will focus on three components of communication that appear fundamental to the legal system however it flows. These components are: 1) transmission (i.e., the straightforward process of sending out legal information from sources of legal authority); 2) access or retrieval (the process of acquiring relevant legal information); and 3) interpretation and comprehension (this involves both sorting out bogus representations about what the law is from the accurate and translating law into the terms in which life and commerce are understood by those affected). While this paper concentrates on these three elements, the specimens will surely reveal other striking features of the legal orders of their times. The inquiry, though, will not be what does a particular law fragment say but what does it suggest about the means of communication. The underlying aim will be to understand how these fragments fits into functioning legal systems, and, in particular, what they reveal of how law-related official information was, at some particular past time and place, being communicated.

II- EARLY SAMPLES

How deep should we dig? With more time, we might try to pull a sample from an oral society. There is ample evidence that law goes back that far. However, finding any observable trace of the communication elements of law in such a sample is near impossible because we are dealing with soft tissue organisms, without exoskeletons. Under the press of thousands of years few traces survive. These problems notwithstanding, useful speculation is possible. Modern scholarship has suggestively illuminated the characteristics and techniques of law and information transfer that likely operated pre-writing -- the importance of repetition, of poetry, and acoustic rhythm, of narrative and public occasions, notions of justice or rightness expressed in examples, instances rather than abstract statement. See E. Havelock, The Muse Learns to Write (1986) and Greek Concept of Justice (1978).

If we probed for the oldest writings on the subject of law, that would still take us down several millennia. At that level even if we struck a cuneiform tablet, say, or similar piece of ancient law writing, we would encounter insurmountable difficulty trying to determine for whom and what purposes it was set down -- essential questions for one seeking to understand what function the writing performed. Scholars debate quite vigorously, for example, whether the codes of Hammurapi and other Mesopotamian rulers counted as law with Babylonian judges. Compare J.N. Postgate, Early Mesopotamia: Society and Economy at the Dawn of History 291 (1992) with R. Westbrook, Cuneiform Law Codes and the Origins of Legislation, 79 Zeitschrift fur Assyriologie 201 (1989). In other words, reliable information about sources of authority and how law was communicated during this period are missing.

Because of these concerns, this paper's samples begin at a much more recent level. I've chosen one of great turbulence and therefore special interest, the onset of print. From the early days of printing, one can find both intact specimens and, significantly, sufficient information about the surrounding legal structure to draw conclusions about how they were used.

Littleton's Tenures (1581 Edition)

Here from a layer 400 years down is Littleton's Tenures -- judged by most to be the first law book printed in England, being first printed in 1481 or 1482. The edition shown here, published in 1581, was its 47th printing, the 14th printed by Richard Tottel. This early use of print to convey legal information contains many striking features -- most surprising, perhaps, to a contemporary eye is the language in which it is written. In the 15th century during which Littleton lived and wrote, the language of law in the King's courts in England was not the language of the people, but so-called "Law French". For over a century and a half of law printing in England, Law French was its principal language. To the extent that the effectiveness of law depends on effective communication, this fact tells us, right off, that law books of this era held a different place in the legal information flow than those printed today. It also serves as a reminder of how directly communication can relate to control. To the extent that those directly affected by official information cannot themselves understand it, they are forced into dependency upon those of the social, professional, educational class, who can access, read, interpret, understand the "language of the law". Revolutions and civil wars have been waged, in part, over law's language.

Treatment of Dower in Littleton

Language aside, this early law book specimen has several other features that illuminate the legal information process of its time. There is ample evidence this book was not intended to be an authoritative law document, a source of law. It seems to be, quite simply, an effort to describe, to interpret law in a particular field for a very distinctive audience, the next generation. Littleton's Tenures began as a teacher's book, written, in fact, for Littleton's son who was preparing for a career in law. It was not intended for wide dissemination. Weight as authority came to Littleton's writing only after his death, in consequence of its quality and uniqueness. As for the work's relationship to truly authoritative legal materials, observe this astounding fact: Littleton (the original Littleton) has neither footnotes nor citations.

This is a short book -- a book meant to be read from cover to cover -- and mastered, that is to say, remembered, not verbatim necessarily, but in its main. This is a book that by 1581, as it was being printed and reprinted, was meant to be used as a personal information resource in a second way -- one few if any contemporary law books fulfill. Most 16th century editions of Littleton were, like this one, printed with large margins (many were far larger than these) intended and used for the owner's notes. This 1581 edition of Littleton once had a cover and a table of contents showing its division into three books; it never had an index. It did, however, incorporate a information technology breakthrough that subsequent editions of Littleton have carried forward, unchanged, for over 400 years, a functional feature critical to any law book's use as authority. I am referring here to the printer's inclusion of a system of identifying smaller units of the whole. If you and I, both familiar with Littleton, had sought to bring his text to bear on a discussion of a point of the law of dower in 1560, we would have no means beyond referring to the subject and initial words of the relevant paragraph to identify the passage in our different editions. Richard Tottel, the genius who published this issue of Littleton with section numbers, also commenced consistent pagination of successive editions of law reports -- at considerable expense (and despite all sorts of deviations within the page). In addition, Tottel established the practice of numbering the judicial opinions within each term (e.g., H.42.E.3.pl8) in printing the Yearbooks.

Lacking citation and, originally, any means of being cited, written for highly personal use, Littleton's book was not designed to fit into an extensive print system. A person could and for years, it is reported, lawyers did read Littleton from cover to cover as a annual refresher.

At a layer only 50 years or so higher up our boring than Littleton we find this related specimen.

Coke on Littleton

This is the first volume of Sir Edward Coke's Institutes, published in English. Subsequent volumes of the Institutes were freestanding Coke, but volume one was "Coke on Littleton". Coke's preface explains why he translated Littleton and presented his own commentary in English -- namely, that "the nobility and gentry of the realm...may understand ... seeing that ignorance of the law is no excuse."

Treatment of Dower in Coke on Littleton

>At least as significant as the change in language is the very different approach and purpose that can be detected in this book. Like the original, Coke on Littleton began as a personal reference; it was quite literally Coke's notes on his own copy of Littleton. But Coke turned his personal notes into an interpretive work that acknowledges and points to other elements of a distributed legal information system. Coke's book is not a memory refresher but a reference work, pointing by citation to countless cases.

Following the direction of Coke's references let us move laterally in the same stratum to the judicial area and draw up a collection of judicial opinions. Being a property teacher, my eye is caught by decision from that field -- Spencer's case decided in King's Bench two years after publication of the edition of Littleton we have been inspecting, 1583.

Spencer's Case, As Reported in Volume 5 of Coke's Reports

This is a landmark case. For well over three centuries the judgment, reported here, has had attributed to it three requirements that must be met before a covenant will run with an interest in land, binding successors. Landmark or not, as one property text observes, the ironic fact about this case is that there is true doubt about what the judgment, in fact, held and even more astoundingly who won. An anonymous case reported in Moore's Reports from about the same time and with a remarkably similar set of facts allowed the plaintiff's action. The conventional view is that the case in Moore is a different and aberrant case, and that the decision did in fact, as reported by Coke, hold against the plaintiff.

How can there be such doubt? The same Sir Edward Coke who wrote Coke on Littleton, first ventured into law publishing with a set of reports on judicial decisions. The initial volume issued in 1600. In all, Coke edited eleven volumes of reports between 1600 and 1616. No printed report of Spencer's case appeared in 1583 when it was decided, nor in 1590 or 1598. It was not until 1605 that volume 5 of Coke's Reports carried news of this decision to an anxious bar and bench. That delay tells us, of course, that printed judicial opinions played a profoundly different role in law at the beginning of the seventeenth century than they do today.

In samples drawn from this region, one doesn't find regular reports brought out more or less contemporaneously with the decisions until a full two centuries later -- namely, King's Bench 1785. Coke's Reports were not that. Like his distinguished predecessor Plowden, Coke was no mere scribe; he intermixed his own commentary with opinions he chose to reported upon. Unlike Plowden, he did not use italics to separate one from the other making it difficult at times to distinguish Coke's views or notes of earlier cases from his report of the judicial opinion. But to Coke and his contemporaries, the precise text of a judicial opinion, rendered orally, seems to have been less important than the soundness of the court's doctrine. Coke seems to have viewed his reported cases as illustrations of doctrine, not sources of law. The reports were intended to be instructional material using actual cases (more casebooks than law reports). In volume 5 of Coke's Reports Spencer's case is not sandwiched between opinions on diverse topics decided on or about the same date but rather gathered with cases on related topics under an appropriate heading.

In sum, it appears even as law material began to be printed in books, courts and those arguing law before them continued to operate within an oral and handwritten tradition. Education required years spent with common lawyers whose memories held the law and cases that exemplified important points of law. The only official records of decisions were the plea rolls, but they contained only the pleadings and formal dispositions, omitting the detailed discussion of legal points. If the attorney wished to cite to such a skeletal record, he would "vouch the record."

On the other hand, one can see that print as a medium was beginning to have an effect on ideas about law, how it was made and transmitted. By the time of Spencer's case cases were being cited, for the existence of reported cases (the body of Yearbooks) had led to their being viewed increasingly as at least a useful means of communicating about law. Individual lawyers indexed points and authorities for themselves and by the mid sixteenth century there had appeared in print two major abridgments which organized reported cases by topic. While the shift did not occur abruptly, one can observe it in the contrast between Littleton's Tenures and the important sixteenth century law books, which were filled with case references.

Spencer's case was decided during this period of changeover. There were already some one hundred printed law books; all the Inns of Court had small collections they called libraries. By 1558, the Yearbooks had been printed. An information explosion that would lead by 1800 to over 1,500 legal titles led to creation of new access devices, abridgments, and new forms of text. While contemporaneous access to Spencer's case depended on oral and manuscript accounts, paradoxically, this particular decision fell into the new technology in a way that enabled it to become an accessible precedent with impact that would have startled a common lawyer in 1583. It would not have had this impact if it had been decided a century later or ten years earlier or if Coke had not picked it to illustrate the law of real covenants. Spencer's case became, as nineteenth century reported decisions routinely were upon delivery, precedent, an authoritative communication of a point of law.

These early print samples, I would suggest, show law in transition from a system in which legal information was passed from generation to generation largely by education, consultation, and professional culture. Writing was part but a relatively small part of the process. And information about law moved to members of the general public without its use altogether. Coming in was a system in which the published record of the law became a key mechanism for transmitting it, initially vertically (from one generation to the next) and subsequently laterally (printed word of a decision or statute going out at the time as rapidly as the technology allowed).

III- SAMPLES DRAWN FROM MORE RECENT PRE-DIGITAL TIMES

By 1800 or a little after numerous U.S. law samples begin to appear, with an explosion of them by the end of the nineteenth century. However one measures an enduring unit of legal information, that was a time of enormous growth in two respects. First, there was simply much more law data. Not only had reported decisions avalanched -- 18 volumes in 1810; 800 or so in 1848; nearly 4000 by 1885; double that number by 1910; but statutes had come to address wider and wider bands of legal topics. Second, those whose professions required them to work with legal information, judges, lawyers, large interests affected by law had, increasingly, to be concerned about legal information from more than one law-making jurisdiction within the U.S. federal system. More and more legal issues or problems crossed state boundaries as individuals and businesses did so.

The stratum holding greatest interest for our investigation lies even closer to the surface, roughly 15-20 years down. That puts us at a time before law via networks and on disk became significant parts of the legal information system. From that level, I shall pull up, brush off, and examine four samples -- a case report, a collection of statutes, a volume of regulations, and a treatise.

CASE REPORT

Contemporary U.S. Law Report (Published by the West Publishing Company)

This volume of New York State decisions (c. 1981) has several characteristics worthy of note. Compared to our prior case report sample, it is both comprehensive and current. (Here and in the following observations, terms like "current" or "up-to-date" refer to historic standards rather than the possibilities opened up by computer-based systems.) A penciled mark inside the cover reveals that it arrived in the Cornell Law Library less than a year after the latest decision it contains was handed down. Of course, because of the rush to print the volume holds information that is out of date. Several of the decisions it reports were within a year or two reversed on appeal. It is comprehensive only in comparative terms. This volume and its many companions ("473" and "2d Series" indicate that this is part of a large collection.) do not carry all legal opinions of New York courts. They do carry all decisions handed down by the highest court of the state with full opinion, with less and less complete coverage the further one moves down the judicial ladder.

Like the majority of U.S. case reports of this era this volume was published by the West Publishing Company. Unlike many others, its full contents can be found in another publisher's volumes, a set published under contract with the state's court system. (And selected opinions from it have also been gathered in additional specialty volumes or loose-leaf services dealing with topics like workers compensation, commercial law, and state practice.) This book provides information, case by case, sufficient to enable users of either set of reports (West's or the "official" reports) to communicate with one another about a particular case. In addition to delivering verbatim texts of the opinions of New York judges, this report includes a reasonably thick layer of summary, interpretation, and legal categorizing. Each decision has been read by the West editorial staff and fit into a legal issue matrix (the West Key Number System).

In these features the volume shows the traces of two solutions to the nineteenth century explosion of print, one more clearly than the other. The "information overload" caused by viewing all judicial opinions distributed in print as precedent led first to organizing and subsequently to filtering solutions. In 1879, the West Publishing Company in St. Paul began printing the Northwest Reporter. By the end of the century, West's full package of innovations was available -- a system of regional reporters that covered the nation, with a similar system of federal reports that for all but the Supreme Court displaced other forms of print distribution. Compared to any competing "official" reports West reports were swift to appear, but their greater advantage lay in an indexing framework, reaching from "abandonment" to "work and labor" against which all these decisions were linked by West editors. This allowed all decisions published by West to be collated in the companion American Digest system, first published in 1897. The West digest brought every reported case (West claimed there to be over 500,000) into a single data access system, printed in 50 volumes. Through West's comprehensive organizing scheme, the search for precedents directly in point was made more manageable.

However, by the latter part of the twentieth century, "every reported case" seemed too many for most courts, and courts that had previously fed their entire output into the precedent system began, with varying degrees of system or principle, to hold decisions back. By the mid-1980's over fifty percent of the decisions of U.S. Courts of Appeals were "unpublished" (the figure was 61% for 1987) meaning simply that the courts in question declined to turn them over to the West Publishing Company, at that point the only distributor in print of their opinions. They held them back on the ground that not all their opinions were valuable sources of law. Several intermediate level state courts took the same approach. As already noted, the New York case reports are not comprehensive. Those of many other states are even less so.

Case reports like this specimen are both static and lumpy. This poses some perplexing issues about how to treat obsolete law data. Print has the characteristic of being fixed once issued. Occasionally, decisions will be withdrawn by a court between the printing of advance sheets and a full volume, and when that happens they are truly withdrawn from the printed volume, with a small epitaph to mark their passing. There appears to be none of that in this sample volume. New York courts do, like others, correct decisions between decision day, temporary prints, and their final publication. Almost certainly that has happened here but the changes were not noted. However, once printed in final form the volume is fixed. It stays as printed for the rest of the volume's long shelf-life.

Printed case reports, in this model, also package law data in large mixed bundles; that is what I mean by lumpy. There is no way to dispose of the decision on an issue of eminent domain that appears at page 451 in volume 437 even though it was later reversed by the New York Court of Appeals (in an opinion reported in volume 453). Yet once reversed the earlier decision has truly been rendered obsolete. It continues to lurk in volume 437 as a trap for any who might fail to use a citator to determine the decision's continuing authority. (Indeed, it is fair to view this book as incomplete without a companion set of citators. Most U.S. law libraries place Shepard's citations directly adjacent to the reports they track.) Many more of the decisions in this volume are obsolete even though not reversed. A 1976 empirical study by Posner and Landes calculated the median age and depreciation rate for precedent cited by the Supreme Court and U.S. Courts of Appeals. They found, for example, the median age of non-Supreme Court decisions cited in Court of Appeals decisions to be 4.3 years. (Supreme Court decisions had a half-life over twice that long and the age of cited cases did, they found, vary some according to field.) W. Landes & R. Posner, Legal Precedent: A Theoretical and Empirical Analysis, 19 J. of L. & Econ. 249 (1976). The basic point of that study for the present exploration is that appellate decisions, distributed in print, hold value as precedent (on average) for a very short time.

I conducted a small empirical experiment with this volume of New York decisions (c. 1981). Using Shepard's citator, I checked to see how many of its 1050 or so opinions had any citing reference during the two years 1989-1991. Only slightly more than ten percent of them did, 109 to be exact. Four of the volume's decisions appear to having significant continuing importance being cited 15 or more times in this period. And one opinion in volume 437, People v. Farrar, seems to contain the definitive formulation of a judge's responsibility in sentencing, following a plea bargain. During the period 1991-1993, this one decision by the New York Court of Appeals was cited no less than 358 times.

Such a count of citing references draws attention to another feature of this volume (and our subsequent samples). How does one direct another to one of its decision or to a particular paragraph within it, someone who may have access to the opinion from another source? Statutes and regulations pose the same issue, even though they have, in print, commonly dealt with it differently. The issue is simply how one identifies an authoritative passage to another lawyer, judge, citizen who may want to examine it. Volume 437 shows citation conventions that are totally premised on print (namely, page numbers from particular printings even though they break decisions into fragments, mid-thought and mid-sentence) and a very limited number of potential sources.

A final observation about this book concerns its habitat. Volume 437 was not designed to roam. Its proper and useful place is close by its many kin and companion citator in a law library. The book's language is English, albeit a dialect many ordinary citizens would have trouble following. But their greatest difficulty in finding the important plea bargain case it contains has to do with tracking it down.

Statute Sample -- Fragment of a Code

The second law sample of recent vintage is a volume containing a collection of federal statutes. The states of the United States, with the exception of one, Louisiana, are still said to be common law jurisdictions. But that doesn't mean they are without codes, that is to say, organized compilations of statutes that touch on many facets of their legal order from torts and contracts to municipal law. Codes have at least a one hundred year history in the U.S. They can be viewed as another important response to the legal information explosion of the late nineteenth century. In 1865, a New Yorker, David Dudley Field, published a general Civil Code, divided into four parts. Until his death near the end of the nineteenth century, Field pressed, without success, for adoption of his code in New York and elsewhere. Field's code was designed to supplant the common law in important areas with statute -- whence so much of the opposition. His New York opponent, James Carter, believed codification wrong because it moved too much law making authority from courts to the legislature -- an untrustworthy body, passionately addicted to the short run. But the shift in gravity which Carter fought was, like gravity itself, inescapable. And codification as a means of dealing with the accumulated enactments of a legislature succeeded where Field's codes saw only fragmentary success.

The chronological gathering and printing of legislative enactments creates even greater problems for the lawyer or citizen seeking the relevant statutes on a topic than chronological printing of court decisions does for those seeking the most relevant precedent. Subsequent enactments must be collated with prior ones to determine the effect of amendments, adding, deleting, or modifying key language, indeed to be sure that a section has not been altogether repealed. Enactments dealing with the same or related topics must be brought together.

The U.S. Congress authorized the first codification of the Statutes at Large of the United States in 1866. It resulted in the Revised Statutes of 1875 -- which performed all these organizing tasks. In varying mixtures of legislative and private publishing responsibility, with or without annotations linking the codified sections to court decisions, codified statutes became a tool of ordering statutory law on which we have come to depend.

A Volume of Title 35 of the U.S. Code as Published by The Lawyers Cooperative Publishing Co. The Table of Titles (Inside the Front Cover) Indicates Title 35 is One of the Titles That Has, As a Codified Title, Been Enacted by Congress

The Following Page. Note: No Public Official Certifies the Text as Authentic. The Private Publisher Asserts a Copyright in the Editorial Material Added to the Statutes But Not the Statutes Themselves

Like the sample case report this statutory volume is put out by a commercial publishing house as one small part of a set. Some U.S. jurisdictions, including notably the federal government, still offer an official public printing of their code. However, for reasons similar to those that have led law reports from commercial publishers to displace "official" publicly printed editions of court decisions during the twentieth century, commercially published codes have for most users and purposes supplanted publicly printed ones. Although U.S. copyright law does not extend to court decisions or legislative enactments, it has been established for over a century that a law report or annotated code, combining uncopyrightable public material and editorial notes, tables, indices, and the like, can be copyrighted. See Callaghan v. Myers, 128 U.S. 617 (1888). Therefore, as with this volume virtually all commercial law publications in the U.S. wrap public domain legal information in such copyrighted material. In some jurisdictions an authorized private code publication carries a certification of accuracy from a public official; but most appear like this one with only the publisher's reputation for accuracy as the user's assurance of authenticity.

Regulation Sample -- Code of Federal Regulations

Code of Federal Regulations, Published by the Federal Government -- A Topically Organized Compilation of the Regulations Issued in the Federal Register (Lagging the Latter by As Much as a Year and a Half and Carrying Minimal Editorial Enhancement, Compared to Commercial Publications of Statutes or Specialty Collections)

This third sample shows a very recent law document type, being a compilation of agency regulations, required by statute to be published in an official Federal Register as issued, later to be compiled into the Code of Federal Regulations. Distribution of full collections of regulations at the state and federal levels remains largely in public hands, although distribution of those affecting a particular industry or important to a particular segment of law practice (such as tax) have strong and successful private publication channels offering not only selection but swifter distribution and editorial enhancements similar to those found in privately published case reports and statutes.

Treatise Sample

Modern Treatise in Its Natural Habitat

The final sample of this set seems, at first glance, a direct descendent of Littleton's Tenures. Like that early law book, this is expert writing about what authoritative law makers have done rather than the very law they have written. But in several ways this book has a far stronger connection to the other samples of its era than to its ancient predecessor. To begin, it lives with and depends on them. It is to be found, in a law library, surrounded by other books on the same or related subjects drawn together by virtue of a standard classification system.

Wolfram, Modern Legal Ethics

This book was prepared for the institutional library. While written out of its author's experience, its completion required the resources of a large law school library. Designed for institutional ownership, it has no space for personal notes. Indeed, to place notes in it would in most settings constitute an offense. This book is built to stand up -- not lie down. It carries its identity on its spine. The text's references assume ready access to a vast array of information resources; indeed, it is fair to say that the book's principal intended use is with such resources at hand. Point after point, it provides references to case reports, statutes, rules, and even agency material. Indeed, for many readers, its job is done once it has brought them to authoritative statements found in these other sources. While written in relatively jargon-free English, it is still a professional's book, largely because of its dependence on a professional's information resources and understanding of how to access them.

This last work reveals even more clearly than the prior three that during the twentieth century legal information has come to be delivered not so much through books (i.e., individual volumes or even a series of them) as through libraries. At the turn of the century, a strong lawyer's or law school library (At that time, they were essentially the same.) filled a few bookcases. Forty years ago there were still law librarians who maintained that their experience and memories made card catalogs unnecessary. The majority of U.S. law libraries had no classification scheme as late as the mid-fifties. In truth, it is only recently that large collections of law books, intentionally interconnected by their publishers and overseen by information professionals, have become the fully integrated information systems we know. The size and integration of these systems constitute at once a source of great strength and serious limitation. They provide up-to-date and comprehensive legal information of a unparalleled quality to those who have access to them and who possess the necessary skills to use their collections. The information system they embody works far less well for ordinary citizens and all who live and work any distance away. The public institutions founded in the U.S. to expand access to legal information -- notably court and county law libraries -- have been seriously underfunded. They have, as a consequence, inadequate collections, staffing, and services. Finally, in an era that sees increasing value in global access to domestic legal information, the costs of porting useful portions of any nation's law (i.e., significant library fractions) internationally loom frustratingly large.

IV- A FEW CONCLUSIONS, LOOKING FORWARD

A short twenty-four years ago, LEXIS first offered U.S. lawyers a computer-based federal tax library, including digital counterparts to all four of the preceding sample types. It was a novelty then. From our current vantage point, we know that it represented momentous and unsettling change. In the ensuing years, computer-based law systems moved from being powerful print supplements used by a few to print replacements relied on by many. These digital delivery systems (whether employing disk, online connection, or some combination of the two) are neither static nor lumpy. They are fast and flexible and reach effortlessly across great distances. They have become mature, broad, deep, and far more current than their printed counterparts.

Within reach if not present reality is the creation of computer-based legal information systems offering access to a complete set of judicial opinions, statutes, regulations, as well as useful expert opinion in many fields of law. Since many law-making bodies now produce law digitally (preparing final versions of their statutes, regulations, or judgments on computers) one can imagine widespread access to authoritative versions of their output directly from the source. Characteristics of the digital medium may make it possible for the language of the law to be at once more available and more understandable to ordinary people.

The few samples examined here do suggest that the activity we call law has changed as the technology available for communicating legal information has improved. But they neither assure that the direction of change will be more effective distribution nor that working out the proper roles for public and private sectors will be swift and easy. It is the prospect of the former that makes it so critical that the latter be addressed.

C.L. Mardi 6 juin 1995