HyperLaw Letter to Clerk, Eighth Circuit Re ABA Citation Proposal, March 28, 1997

3/21/97 HyperLaw,

Back to HyperLaw Home Page
Back to Citation Page

March 28, 1997

Michael E. Gans
Clerk of the Court
United States Court of Appeals
For the Eighth Circuit
U.S. Court & Custom House
114 Market Street
St. Louis, MO 63101

Dear Mr. Gans:

[1]     I have reviewed a copy of your letter to the Automation
Committee of March 14, 1997 concerning the ABA Citation
Resolution. The letter was written by you as chair of the
Appellate Court Clerk's Advisory Committee. Your letter
does not mention the positive experience of the Third and
the Sixth Circuits which have been sequentially numbering
opinions for the last several years.

[2]     I am troubled by your letter because is seeks to
provide the impression that the views of the Appellate
Clerks were not solicited or evaluated by the ABA Committee.
You state in your letter:

     When we forwarded our comments to the
ABA's Special Committee on Citation Issues
last April, we invited the Committee to open
a dialogue with us on issues related to
opinion issuance and citation. We regret that
the Committee did not see fit to respond to
our letter or address our concerns. As the
court officers charged with the
responsibility for issuance of federal
appellate court opinions and maintenance of
the courts' official records, the clerks of
court have a vital interest in this topic and
are in a pivotal position to help shape
opinion and practice. We hope that the
Committee on Automation and Technology will
give serious consideration to the concerns
expressed in this letter.

[3]     Unfortunately, the April 1996 appellate clerk comments
came six months after the ABA Committee first asked for
public comments. Your comments were never made public by
your committee. The ABA Committee held public hearings in
November, 1995. The Committee's appointment by the
President of the American Bar Association was well
publicized in August of 1995. There were articles in
various legal publications. The hearing was well
publicized. West sent over ten people to the meeting. Had
you chosen to participate, or file public comments prior to
or immediately after the ABA hearing, then your concerns
could have been addressed, and, indeed, would have been
responded to by others participating in the process --
indeed, I expressed some of your concerns about the docket
number. No one at the hearing was able to respond to
questions of the ABA Committee as to the internal docketing
procedures and available automation. Now, after the close
of public comments to the Automation Committee, for the
first time the arguments of the appellate clerks have been
made public, and may now be scrutinized for logic,
consistency, and practicality.

[4]     In some ways letter seems to focus on practical issues
in implementing the system. Problems which are easily
overcome are overstated and exaggerated. Moreover, your
letter fails to consider larger issues, including the needs
to address the problems of the district courts, and the
impact on the public.

[5]     Your letter the does not in any meaningful way respond
to the policy issues raised in the ABA proposal and states:

     In summary, we dispute the need for the
alternative case citation system proposed in
the ABA's resolution.

The question is need for whom? Need for the clerks? Need for
the public? Unfortunately, your letter does not in any way
attempt to refute the detailed rationale for the new
citation described in the ABA report. Your letter then
states as part of the summary:

We believe "Blue Book" citation of
unpublished opinions (by case number, court
and date) is still the most effective method
for citing electronically reported opinions.

[6]     Although you refer only to "electronically reported
opinions", the ABA citation is a method for print and
electronic publishing of opinions -- this leads me to
believe that you are still have in mind the 1992 ECS
proposal, which solved no problem because it was not a
permanent citation. The ABA proposal is for an "immediately
available" and "permanent citation," and not just intended
for electronically reported decisions. Second, your
statement appears to address only "unpublished" opinions.
Are you stating that the sequential number are indeed
appropriate for "published opinions", but not for
"unpublished." Even if you so intended to distinguish
between "published" and "unpublished" opinions in this
statement, how does one address the problem at the district
court level, where there is no official concept of
"publication." There is no official court record at the
district court level for most opinions designating whether
an opinion is published or not.

[7]     You may not be aware, but the Sixteenth Edition of the
Bluebook explicitly endorse the ABA style citation with
sequence numbers. Hopefully, the members of the Columbia,
Harvard, and Pennsylvania and Yale Law Reviews who adopted
the proposal after their own study in 1996 are this year
ensconced as clerks for federal judges and perhaps will be
able to provide a balanced view of why they changed the
rule.

[8]     The present opinion citation system, which
you seem to prefer, is based upon completely
arbitrary numbers of where an opinion appears in a
book printed by a private publisher, and is in no
way coordinated with the docket number. However,
you letter states:

We strongly oppose the creation of an opinion
citation system which identifies opinions by
anything other than the case number assigned
by the issuing court.

Are you saying that it is appropriate to have as an official
citation a privately established citation that does not
include the docket number, but that it is not appropriate to
have a public citation that similarly does not include the
docket number? I appreciate the concern of your members to
have a cross reference between the sequential citation and
the docket number. However, the appellate clerks never place
the West citation for an opinion on the docket sheet, and
thus one cannot find the West citation in the public court
records. If one only knows the West citation of an opinion,
using court records alone, one cannot find easily the
official court version of the opinion.

[9]     Your letter attempts to assume the higher ground of
authority based upon practical knowledge in disseminating
appellate opinions electronically and your work with the law
school Internet site:

     We wish to begin our discussion by
noting that we have strongly supported the
electronic dissemination of federal appellate
court opinions. The federal courts of appeals
have taken the lead in the development of
electronic bulletin board systems for the
distribution of opinions and case-related
information. So we approach the ABA
resolution with a wealth of practical
experience as to how electronic distribution
systems work and how these cases are cited.

[10]     I will now hold you to the higher ground you have
sought to assume. The opinions which have been
electronically disseminated cannot be cited in a brief to an
appellate court or used in a legal memoranda for several
reasons. Many of the opinions do not include the
pagination, or if they do, only include the manuscript and
not slip pagination. Even were a federal appellate court to
look kindly upon a brief that cited a 1993 published opinion
by docket number with regard to the West citation, one could
still not cite to a pin-point location. Moreover, many
opinions have been modified after release on the bulletin
boards, and the amending orders were never made available
electronically. For many courts BBS and Internet opinions,
a reader cannot determine if the version of the opinion in
her possession is the latest version, because many courts
adamantly insist upon modifying opinions, but keeping the
same file name. Finally, by palming off responsibility for
archives upon law schools, the courts have thrown the
concept of accountability out the window.

[11]     Those circuits such as your own that rely upon docket
numbers in the electronic environment seem to be at a loss
in dealing with multiple opinions in the same case. It is
not at all unusual to have earlier opinions overwritten by
later opinions on the bulletin boards. For example, one
needs to look no further than the Eighth Circuit and the
opinion in the case McMorrow v. Little, No. 95-3862. Two
different opinions were disseminated on your bulletin board,
and both have the same file name 953863.P8. One is dated
the January 8, 1997 and the other is dated March 17, 1997.
Although the earlier opinion is not on your BBS, both of the
versions are available on the Eighth Circuit site maintained
by Washington University School of Law. These files may be
found at the following locations.

ftp://ftp.wulaw.wustl.edu /8th.cir/970108/953862.P8
     January 8, 1997
ftp://ftp.wulaw.wustl.edu /8th.cir/970317/953862.P8
     March 17, 1997

But, oddly enough, the only reason that the earlier version
was not overwritten is that the Washington University School
of Law does not organize the opinions by docket number, but
organized the files in directories based on date. Other
school Internet sites store them in a single directory, and,
in those situations, different files with the same name are
merely overwritten. If one is seeking to download a file by
docket number on the Washington University site , one has to
search through numerous directories organized by docket
number, which makes one wonder about the sanctity of the
docket number.

[12]     One circuit has such a problem in dealing with docket
numbers that if a matter has multiple docket numbers, then
multiple opinion files are disseminated, one under each name
-- the Fifth Circuit does this.

[13]     Of course, it is true that the problem of overwriting
file names is also possible with the use of sequence
numbers. That is why it is necessary to indicated amended
files. See the following two Third Circuit files on the
Villanova WEB site. On January 10, 1997, the file <a
"ftp://ftp.vcilp.org/pub/law/Fed-Ct/Circuit/3d/96a1461a.wp5"
was uploaded which amended "ftp://ftp.vcilp.org/pub/law/Fed-
Ct/Circuit/3d/96a1461p.wp5".

[14]     The lack of Circuit accountability for the law school
web sites is a problem. Only the Federal Circuit runs a
simple straight-forward court WEB site with access to
individual opinions, but then decided to not archive more
than 90 days of opinions, casting this responsibility on to
the law schools. The other Circuits palm of their
responsibility on unaccountable law schools, some of whom
then attempt to copyright their HTML versions of the
opinions. For example, in the Sixth Circuit site maintained
by Emory Law School:

One cannot find a case searching by the docket number
The list of cases does not include the docket number.
One cannot find a case by searching for the case by it
Sixth Circuit cite, say "97a00001p.06.
One list of cases does not include the Sixth Circuit
cite.
The opinions are not paginated.
The rtf versions of the cases cannot be easily obtained
in batch form -- there is no FTP capability.
Emory attempts to copyright the HTML versions of the
opinions.

[15]     The same is all or partly true as to the other Circuits
on the Emory Law School cite. So, who has the
accountability. Your letter states " So we approach the ABA
resolution with a wealth of practical experience as to how
electronic distribution systems work and how these cases are
cited" -- so, it is my assumption that the appellate clerks
are responsible for this chaos. One reason for my
preference for sequential numbers is that it is a system
that has built in accountability: if an opinion shows up
that was published and has no sequence number or has not
been made available, then it is clear there was a screw-up,
and, two, that the office that assigns sequence numbers made
the screw up and should fix it.

[16]     As to pin-point citation, the biggest problem is that
most of the published opinions just cannot be cited in any
brief filed in a court in the real world, and cannot be pin-
point cited because the page numbers frequently just do not
make it to the electronic versions. Just as the courts do
not wish to put in the extra 5 minutes per opinion (as
estimated by the Fifth Circuit) to insert paragraph numbers,
the courts do not want to put in the extra time to insert
the pagination in hard form, and in most courts, do not put
in the page number as it appears in the printed slip
opinions (as an example: the Fifth and Seventh). Based upon
my experience, it is as hard if not harder to insert hard
page numbers than to insert paragraph numbers after a
document is finalized, because the process of inserting page
breaks and page numbers can itself change the pagination and
location of footnotes [the publisher's analogy to the
uncertainty principle]. But, if your suggestion is accepted
to use page numbers, someone is going to have to do this
work which is being ignored now. Only the Ninth Circuit
includes the page number as they appear in the printed slip
opinions [except for the courts that do no use outside
printers.] Of course, the Ninth Circuit disseminates the
electronic file from their slip printer and it is no extra
work. But, if a court decides to disseminate opinions in
electronic form and use the page number, then someone is
going to have to be willing to step up and place the hard
page breaks and hard page numbers in the final electronic
text and, one that retains formatting and fonts.

     [17]As a parenthetical, some day traditionalist judges
are going to wake up and realize that when their courts
disseminate their treasured opinions electronically in ASCII
form [which is what your Circuit does on the Internet] then
anything italicized for emphasis goes to naught and all the
tedious formatting for citations to assure compliance with
style guides goes to naught as well. I am bemused, because
I suspect the most traditionalist judges who are concerned
with the aesthetics of italicization and bolding probably
are the least likely to surf the Internet to obtain the
dumbed down electronic versions that the law schools and the
clerks offer to the masses. I suspect that if we dig deep
enough to explain why this is happening, we will find out
that font styles were thrown out the windows as an for
expediency to avoid putting page numbers in the electronic
text since it is so simple to "print to file" and preserve
the page numbers, but lose everything else. Using paragraph
numbers would avoid these problems.

[18]     I also would challenge your statement that the present
method of citation is not a problem and that no attorney has
ever presented this as an issue. Let's be clear why West
promotes this view: they wish to make it sound like it is
only their sneaky pirate competitors who are out to do this
for monetary gain, and the members of the bar and the public
do not care. Contrary to the impression you wish to convey,
this is a broad based movement both within and without the
bar. It is also a topic that has garnered great interest at
the college and law school level, as students adept in the
Internet are savvy enough to see the problem. Your letter
states:

     Our experience since we first addressed
this issue has confirmed our belief that an
alternative citation system based on
sequential opinion number is not needed in
the federal courts. Neither attorneys
practicing in our courts nor judges
researching and writing opinions have
expressed the view that the current system of
citation based on case number, date of
issuance and court hinders their use or
citation of federal court opinions. Likewise,
no judge or attorney has ever expressed to
any of us any desire to adopt a citation
system such as the one proposed in the ABA
resolution. In our opinion, the proposal to
create a citation system based on sequential
opinion numbers is a solution in search of a
problem.

[19]     The issue of a sequentially oriented citation was first
proposed in 1991 -- by lawyers and judges within the federal
court system. This was when CD-ROM case law reporters was
in its pre-infancy, and the impending ability to disseminate
opinions at nominal cost to millions of people over the
Internet was known only to a few. At that time, it was a
few forward looking judges and staffers who saw the future
and saw what was needed. But at lot has happened in the
last six years and the few have now become the many.

[20]     I believe it is important for each of the clerks to
take a long evening as I did last night to review the 1200
pages of submissions to the Automation Committee from the
lawyers and non-lawyers of every type. A number of judges
support this system and do say so in the response. And, to
say; "no" attorney has an interest in this, is misleading to
the extreme. Moreover, the proposal itself was bar
generated and has been approved not only by the American Bar
Association, but numerous other bar associations. After
all, the Litigation Section of the American Bar Association
was soundly behind this proposal -- and, I think those
attorneys, do indeed practice before the federal courts.

[21]     I thought you ought to be aware that the phrase
"solution in search of a problem" which you repeat in your
report is the exact phrase authored by the spin-masters at
West Publishing Company to undermine this proposal. It is a
demeaning comment and is doubly so knowing its original
authorship. West even published a position statement with
that slogan as a subtitle in 1992, even while the Canadian
company Thomson and its subsidiary Carswell was using
paragraph numbers in Canadian reports. Now that Thomson
owns West, I guess that a lobotomy is required to keep the
left and right hands operating efficiently.

[22]     I am most surprised at the suggestion that maintaining
cross-reference tables for sequence numbers and docket
numbers is some type of insurmountable tasks. I reject that
proposition out of hand. Certainly, it needs to be done,
but it is a simple task. Your letter states:

     Assuming for the sake of argument that a
new system of citing opinions is required
because of the widespread use of electronic
case reports, it has always been our position
that the system must be based on the case
number assigned by the issuing court. We take
this position for several reasons.

     First, the issuing court case number is
the key to the history of the case, its
current status and all of the documents in
the appeal. An opinion identifier which does
not include the actual case number of the
appeals is, by itself meaningless. A reader
who has only the sequential number will
always have to take an additional step to
determine the case number before the file or
procedural history of the case can be
accessed. In effect, reliance on a sequential
number to identify opinions requires use of a
"key" to translate the opinion number into a
case number. This is especially important now
that the courts offer on-line access to
docket information. Future plans to create
on-line access to the actual documents in the
case file will make the case number even more
critical to researchers.

     Several key questions about this
translating key are not answered by the ABA's
resolution. Who will create and maintain the
key? How will researchers access it? How long
will it have to be maintained for future
readers' use? We think the answers to these
questions are that the clerks of the circuit
courts will be required to create and
maintain these keys, that we will have to
provide the information to users through our
existing staffs and resources, and that we
will have to maintain the keys forever.

[23]     I do not believe that the average professional familiar
with the basics of spreadsheets and databases would except
these exaggerations of complexity. Maintaining a table of
cases is so easy to accomplish. The federal courts have the
computerized docketing system that could be easily modified
to do these simple, straightforward computer tasks.
Finally, there are services like Insta-Cite, Auto-Cite and
Shepards, that could handle this task. The solutions are
many, if they are sought. For example, I was able to create
a list for the 3rd Circuit sorted by docket number, and by
the Third Circuit file sequential file name. I used as a
source the index file on the Villanova FTP site. I then
created an HTML list. At least for the Third Circuit, this
is going to be very easy to do. These tables were posted
yesterday on HyperLaw's WEB site at:

<a href="http://www.hyperlaw.com/3rdseq.htm"> Sorted By
Sequence Number</a><br>
<a href="http://www.hyperlaw.com/3rddok.htm"> Sorted By
Docket Number</a><br>

[24]     Were the courts assume the minimal task of tagging the
docket number and the sequence number at the beginning of
each case, then search engines could do this task. Thus,
each opinion would have at the beginning a tag such as:

<Circuit>8th</Circuit>
<DocketNo>96-0191</DocketNo>
<CaseType>Civil</CaseType>
<Sequence>0086</Sequence>
<Year>1997</Year>
etc.

[25]     This header would be attached at the beginning of each
file, and would make it a snap to locate opinions and even
generate the tables needed.. This is what the SEC does in
the EDGAR system that makes authoritative documents
available and searchable. But, courts do not like this
because if one attaches a header at the beginning of an
opinion, then it messes up all subsequent pagination, unless
someone has taken the effort to hard paginate and insert
page numbers. [Similarly, on a separate topic, the thought
of permitting parties to file documents electronically with
the courts without requiring some type of embedded tagging
in a header as is done with EDGAR is a step in the wrong
directions.]

[26]     I do not understand the necessity of enumerating the
changes in procedure as you suggest here:

     Second, the use of a sequential number
to identify opinions complicates our internal
operations. Each clerk's office will have to
create and maintain new databases to track
this information, and we will have to train
deputies in their use. Additionally, some
courts have opinion issuance procedures and
agreements with printing vendors which would
cause confusion with respect to the date of
issuance if a sequential numbering system is
adopted. For example the Fifth Circuit
designates certain opinions for issuance
through their printer only, while others
issue in manuscript form. The opinions issued
through the printer are actually released
some time after they are received in the
clerk's office due to their printing schedule
and, as a result, their numbering would fall
outside the numbering sequence anticipated by
the ABA resolution.

Are you saying that new procedures would need to be
implemented, or that it is extremely difficult to resolve
these problems.

     Third, we believe the sequential
identifier is no simpler to use than a case
number identifier. For example, nearly all of
the circuits issue more than a thousand
opinions a year. Is 1997 8th Cir. 1234U any
easier to use or remember than the current
system based on case number, court and date
of issuance? The difference between the two
forms of citation is simply not significant
enough to justify the loss of the important
information provided by a citation which
includes the actual case number and the date
of issuance.

What if there is more than one opinion per case (more common
at the trial court level), what are you now suggesting as to
simplicity? You then state:

     Finally, the actual date of issuance of
information contained in the current form of
citation is important to many users, as it
provides direct access to information on our
opinion bulletin board and Internet sites,
nearly all of which organize opinion
directories by date of issuance.

[27]     Now I am really confused -- if the docket number is so
great, why organize the opinions by date -- and surely, the
Internet sites could be reorganized. Actually, the opinions
are organized on the system directories on most of the
Internet sites by docket number. However, externally
generated tables organize the opinions by date, frequently
leaving out the docket number [note, that the date and
sequence number organization would be substantially the
same]. Your have really argued the point against yourself.
This proves that people do not want opinions organized by
docket number because, with the years that it may take an
opinion to travel through the system [especially at the
district court level], then opinions issued years apart are
located together. But, what the law schools do is prove
that the externally generated tables can provide order to
the chaos that results from naming the files with the docket
number. I do not oppose using the complete date, and there
may be some details to work out. But, the point of this
discussion is that docket numbers are not as essential as
suggested as long as the simple cross-reference information
is available in the form of printed table, WEB lists,
embedded tagged information, or databases.

[28]     Finally, using a docket number makes any possibility of
print publication extremely difficult: one would have to
wait years to print permanent bound volumes with the cases
in docket number order. And, I shudder at the thought of
shoveling tens of thousands of opinions on a CD-ROM with
only a docket number identifier. That is why HyperLaw
assigns a unique number to each opinion on its CD-ROM.

[29]     As to the enormous task of paragraph numbering, you
state:

     First we object to the provision that
the courts should add paragraph numbers to
the opinions. We believe this would entail
substantial additional work for chambers and
for clerks' offices. In our opinion the court
should be responsible for the text of the
opinion and initial page numbering, and all
other services, including paragraph
numbering, should be left to the marketplace.

[30]     The Fifth Circuit (which appears to be the leader in
opposing any citation change at all) states that this will
take 5 minutes per opinion. Based upon your statistics of
1000 opinions per court per year, that is under 100 hours a
year per court. You do not state the number of law clerks
and secretaries available to the 200 or so judges who write
these opinions, but I believe it is three each, so we have
600 people to spend 1300 hours, or a couple of hours per
person per year. If this is all it takes, I say it is worth
it.

     Second, the resolution does away with
the current practice of providing a parallel
pinpoint cite to the Federal Reporter
citation. For the foreseeable future, the
Federal Reporter will continue to be a
significant research tool for judges,
lawyers, prisoners and other pro se
litigants. Eliminating the parallel pinpoint
cite leaves many readers either greatly
inconvenienced or at a significant
disadvantage, while including it requires
little additional effort on the part of the
writer. Although the ABA may well be correct
in its prediction that primary reliance for
case citation will eventually shift to
electronic case reports, the clerks believe
this day is far enough off that parallel
pinpoint cites must still be required.

[31]     Actually, a better suggestion for prisoners and pro se
litigants would be to make available to them searchable
authoritative case law. Should not federal prisoners and
moderate income criminal defendants have the same access to
Westlaw that is provided to federal prosecutors and judges?

[32]     Finally, I note that I had previously made comments
urging that the concept of using a docket number be
considered by both the AALL and ABA Committees, partly
because I wished to have those committees address the
concerns that had been raised. Although I do believe that
docket numbers convey useful information, and assume that
they will be used as a cite base for opinions that are not
assigned sequential numbers, I am confident that the ABA
Committee has in fact addressed your concerns in its report
and in its submission to the Automation Committee.

[33]     I believe that sequential numbers are important because
they importantly permit one to know that a source of
opinions indeed has collected all of the opinions. This
point was made by John B. West in 1909, and his point as is
valid today as it was then. And, it will make print
publication more feasible. I also have since become more
familiar with the federal court docketing computer systems,
and, am convinced that sequence numbering is something that
could be programmed into the system in a straightforward
manner. Using docket numbers is a lazy way out, and permits
the type of confusion and file overwriting detailed above.
Finally, at the district court level, forcing the District
Court judges to assign sequence numbers to the very same
opinions that they have historically sent off to West is a
good thing and makes this process more public.

[34]     As noted above, the Judicial Conference did permit the
Circuits to experiment, and the Third and the Sixth Circuits
did assign sequence numbers. And, guess what -- it works,
and it has helped those circuits and users of opinions from
those circuits to have some comfort that all the opinions
were being placed on the bulletin board and the internet and
made public. As a result, only in those circuits can a user
of federal court opinions determine that the user has indeed
collected for her own use all of the opinions. It is also
possible for another entity operating a web site, such as a
law school, to ascertain that all of the opinions are
available.

[35]     Had the 1992 proposal been adopted, the public would
now have available to it over 100,000 federal appellate and
district court opinions with a permanent public domain
citation -- that was a missed opportunity. But, with more
mature computer systems today, there is no reason to delay
this any more.

[36]     As the federal courts move forward with automation, I
hope that that there will be more opportunities to exchange
ideas. In that vein, if you are in Washington on April 3,
1997 for the Automation Committee hearing, I invite you to
attend the Press Conference being held at the National Press
Club after the hearing and to further exchange ideas on this
important topic.

Sincerely,

Alan D. Sugarman
March 28, 1997
Page 14 of 11