    Survey on the Software Copyright "Look and Feel" Lawsuits
                               by
                        Pamela Samuelson
                               and
                        Robert J. Glushko


     The software copyright look and feel lawsuits have created a
climate of uncertainty in the user interface design field.[1, 2]
Although individuals may have an opinion about how these lawsuits
ought to be decided, it is difficult for them to know how
representative their views are.  This column will report on a
survey on the user interface field's perspective on these
lawsuits which was conducted at the sixth ACM Conference on
Computer-Human Interaction (CHI '89) on May 2, 1989.  This forum
was chosen because the annual CHI meeting is the largest and most
important gathering of user interface researchers, designers, and
developers, the people who have the most to gain or lose by the
outcome of the look and feel controversy.  (A fuller report on
the survey findings, which includes detailed statistical
analysis, is available elsewhere.[3])  The column will go on to
compare the results of this survey with a report jointly written
by ten intellectual property scholars[4] concerning copyright
protection for "look and feel" and other aspects of user
interfaces.  The CHI survey results are, in general, consistent
with the scholars' conclusions based on copyright principles.[3]
The legal experts see a basis in copyright law for denying
copyright protection to "look and feel," which is what the user
interface field thinks would be in the field's best interest.

     The results of the CHI survey can be briefly summarized:
More than eighty per cent of the 667 respondents to the CHI
survey opposed copyright protection for the look and feel of user
interfaces, although they strongly supported copyright protection
for source and object code.  They regard the kind of strong
copyright protection being sought in the current look and feel
lawsuits as likely to have a negative effect on their own work
and on the user interface design community and industry.  They
oppose strong copyright protection for user interfaces because
they think such protection would adversely affect the climate of
open exchange and discussion of research and design innovations
that has brought progress to the field.

Background on the CHI Survey

     To assist the user interface design community in becoming
more aware of the legal issues involved in the current round of
look and feel lawsuits, the first author (Samuelson) organized
and moderated a 90 minute legal debate on copyright protection
for user interfaces as a plenary session at CHI '89.[2]  The
debate featured Jack Brown, the chief lawyer for Apple Computer
in the Apple-Microsoft litigation, and Thomas Hemnes, a former
defense lawyer in the Lotus case, debating the legal perspectives
on the pros and cons of protecting the look and feel of software
user interfaces through copyright law.  Dr. Michael Lesk of Bell
Communications Research was an industry discussant.

     After a brief introduction to copyright principles by
Samuelson, Mr. Brown made his argument, emphasizing the
significant amount of creative work that went into the design of
a user interface and the role of copyright in protecting those
whose intellectual labor had produced a valuable product from
those who found it easier to imitate a creative work than to do
something creative themselves.  Mr. Hemnes argued that not every
valuable intellectual product was protectable by copyright law,
and that where nonprotection of certain aspects of intellectual
works would further progress in a field, copyright law would and
should respect that. An article on the debate published in the
New York Times reported that Brown and Hemnes made "equally
persuasive" arguments.[5]

     The authors realized that the CHI legal debate would provide
a unique and efficient opportunity to survey a large sample of
the user interface field about the legal issues. Such a survey
should not be distributed until after the lawyers had had their
say, so that the audience would have been educated about the
terms of the legal controversy and each side had presented its
view.  As the debate drew to a close, Samuelson informed the
audience about the survey that was being distributed to them, and
said that though it was the judges who would make the final
rulings on the look and feel lawsuits, this was a chance for
them, as representatives of the user interface design community,
to "vote" on the legal issues, and urged them to do so.  Six
hundred sixty-seven members of the audience filled out the survey
before they left the auditorium.

     Before delving into what the survey respondents had to say
about the legal issues and our interpretation of the survey
results, it may be useful to have a profile of the survey
respondents.

A Profile of the Survey Respondents

     The survey asked a number of questions about the respondents
and their firms so that it would be possible to analyze whether
characteristics of the respondents or their firms might predict
their views on the look and feel lawsuits and related issues.
For example, respondents were asked to select from a list of job
functions the one or two descriptions that best fit (1.68 was the
average rate of response by job function).  Table 1 shows the
respondents' profile by job function for the most frequently
indicated categories.

          Table 1
          Respondent Population by Job Function
          User interface designer       44%
          Researcher                    32%
          Software engineer             29%
          Human factors engineer        15%
          Manager                       15%
          Faculty                        8%
          Consultants                    6%
          Students                       6%

     Respondents were also asked to identify the one best
description of the organization or company for which they work.
Table 2 reflects the results of this question.

          Table 2
          Respondent Population by Employer
          Computer manufacturer         26%
          R and D organization          23%
          University                    20%
          Software vendor               14%
          Other                         17%

The survey did not ask respondents to identify the organization
or firm for which they worked, but since the respondents made up
42% of the total conference registration, the registration
information provides a reasonable substitute without compromising
the identity of particular respondents.  The ten organizations
with the highest number of attendees at CHI'89 were, in
decreasing rank order: IBM, Hewlett-Packard, Apple, MCC (the host
organization in Austin), AT&T, Texas Instruments, Xerox, Bell
Communications Research, the University of Michigan, and
Carnegie-Mellon University.  In addition to "look and feel"
litigants Hewlett-Packard, Apple, and Xerox in positions two,
three, and five, there were several representatives each from
Ashton-Tate, Lotus, and Microsoft in attendance at the
conference.

     None of the factors characterizing the survey respondents
were found to predict statistically significant differences in
their answers to the questions concerning the role of copyright
and patent in the protection of various aspects of software or
concerning predicted effects of strong copyright protection.  In
view of the support the survey gives to the "minimalist"
interpretation of the appropriate reach of copyright law as
applied to software,[2] it is worth pointing out that the
respondents were among the leading designers and researchers in
their field, responsible for creating many of the most
commercially valuable user interfaces in the software industry.
They typically worked for commercial firms that rely on copyright
law to protect their software products.[3]

Survey Findings on Protection of Look and Feel

     One of the principal findings of the survey was that the
user interface field thinks that the look and feel of user
interfaces should not be given protection by copyright or patent
law.  Seventy-seven percent of the respondents with an opinion
felt that look and feel should not be given protection by either
copyright or patent law.  Eighty two per cent opposed copyright
protection for look and feel.

     Quite a few of the survey respondents explained their
reasons for opposing legal protection for look and feel of user
interfaces.  Some said they weren't sure what look and feel
meant.  Some were unclear how similar interfaces could be in look
and feel before infringement might be found.  Others thought that
look and feel related largely to functionalities of the interface
which copyright should not protect.  Still others expressed
concern for the effect on the users, as well as the industry, if
the pending look and feel lawsuits established strong copyright
protection for user interfaces.[3]

     In addition to asking about look and feel, the survey asked
for views about legal protection of five other aspects of user
interfaces, all of which (either explicitly or implicitly) are at
issue in the current round of look and feel cases.  The survey
revealed even stronger opposition to copyright protection for
commands, user interface functionalities, and screen sequence
than to look and feel protection, as Table 3 shows, as well as
strong (but somewhat less opposition) to such protection for
screen layouts.  Icons were the aspect of user interfaces for
which there was strongest sympathy for protection, but not even
this feature enjoyed majority support among respondents.
Strongest opposition was registered as to protection of commands
and user interface functionalities, with more than nine of every
ten respondents objecting to their protection by copyright.

                             Table 3
     Support for Copyright &/or Patent for Software Aspects
aspect         cop.      pat.      both      neither
source code    71%       10%       12%          7%
object code    65%       10%       11%         15%
pseudocode     39%        7%        6%         48%
module design  18%       16%        6%         60%
algorithms      8%       32%        7%         53%
commands        6%        4%        2%         88%
icons          37%        3%        3%         57%
scr. layout    25%        4%        2%         69%
scr. sequence  13%        6%        2%         79%
look & feel    15%        5%        3%         77%
UI funct'ity    4%       12%        2%         83%

     Because many of these features overlap significantly with
the kind of look and feel being sought to be protected in the
lawsuits, it is not surprising that the respondents would view
protection of these aspects of interfaces in much the same way as
they viewed protection of look and feel.

     Opposition to protection of look and feel was, however, not
part of wholesale rejection of intellectual property protection
for software, as Table 3 also shows.  The respondents
overwhelmingly supported intellectual property (and mainly for
copyright) protection for source and object code.  Fully 93% of
those with an opinion supported intellectual property protection,
either through copyright or patent, for source code.  The 85%
support for object code protection was nearly as strong.

     Although, as a group, the respondents strongly supported
copyright protection for source and object code, they did not
support copyright protection for pseudocode or modular design.
Least of all did they support copyright protection for
algorithms, although nearly one-third of the respondents
supported patent protection for algorithms.

Predicted Negative Effect From Copyright Protection for Look and
Feel

     Another major finding of the CHI survey was that the
respondents regarded the kind of strong copyright protection
being sought in the look and feel lawsuits as likely to have a
clear negative effect both on the industry/community and on their
own work.

     In response to a survey question about the effect such
protection would have "on your own work", the average rating (on
a five-point scale ranging from "1" for "significant negative
effect" to "5" for "significant positive effect") was 2.049, a
clear overall expectation of a negative effect.  But it was not
just a minority of respondents with "significant negative" votes
who swayed the average.  Seventy-two per cent of the respondents
expected a negative impact on their own work (ratings of "1" or
"2") if the current lawsuits established strong copyright
protection, and only nine percent expected the effect on their
work to be positive (ratings of "4" or "5").

     The predicted effect "on the user interface design
industry/community" was even more strongly negative, with an
average rating of 1.646 on the same five-point scale.  Indeed,
eighty-six percent of the respondents expected the kind of strong
copyright protection for user interfaces being sought in the
"look and feel" lawsuits to have a negative impact on the
industry, while only ten percent viewed the prospect as positive.
Table 4 gives the results of the respondents' predictions about
the likely effect of strong protection on their own work and on
the industry.

                             Table 4
     Predicted Effect Of Strong Copyright for Interfaces
                    -                   +
     effect         1    2    3    4    5
     own work      35%  36%  19%   7%   2%
     ind/comm      57%  29%   4%   7%   3%

     Lest the reader interpret these results as only reflecting
the opinion of "worker bees" or "ivory-tower" types, we hasten to
point out that 15% of the respondent population identified
themselves as managers, and their responses to the prediction
questions and the protectability of individual features question
did not differ in a statistically significant way from the
responses of the respondent group as a whole.  The average
predicted industry effect among the managers, for example, was
1.74.  The average responses by job function closest to the
manager predictions were those of faculty (1.76), students (1.73)
and user interface designers (1.72).  All were still well on the
side of a 2.0 rating, which itself was a negative rating.  Not a
single category of respondents measured by job function predicted
even a neutral, let alone a positive, effect on the industry if
the look and feel lawsuits established the kind of copyright
protection being sought.

     When we compared the answers respondents gave concerning
their predictions of effect on their own work with their
predictions about the industry effect, we noted that while there
was some shifting in both directions, people who gave different
ratings on the two questions were 3.5 times more likely to move
in a more negative direction when predicting the industry effect.
Particularly striking was the finding that of the nineteen per
cent of respondents who expected to be unaffected in their own
work if the current lawsuits established strong copyright for
user interfaces, sixty-nine per cent nonetheless expected a
negative effect for the field.

The Extent Of Restriction Perceived At Present

     The survey also inquired about the extent of restriction the
respondents currently felt about the use they could make of the
latest research and design innovations which they saw or learned
about at conferences such as CHI.  Respondents were asked to
select one of four statements that best described their views:
     1)   no restrictions: I can freely use anything I learn
          about or see,
     2)   some restrictions: I can't copy exactly, but I am
          allowed to reimplement or reverse engineer any
          interesting designs,
     3)   significant restrictions: I can copy only general
          concepts or ideas at the research stage, or
     4)   total restrictions: Once I see it at CHI, I know I
          can't copy it in any user interface design of my
          own.
Thirty-one percent of respondents reported feeling "no
restrictions" on use of innovations seen at CHI.  Just under half
of the respondents (49%) felt only "some restriction."  Only one
in five respondents reported feeling "significantly" (19%) or
"totally" (1%) restricted in their use of design innovations seen
at CHI.  (Here, there were some statistically significant
responses among respondents by type of employer.  Government
employees felt least restricted, with 64% assuming no
restrictions.  However, even those who worked for computer
manufacturers felt fewer constraints than one might have guessed,
with 24% perceiving no restrictions, and another 50% reporting
only some restrictions.)

     Not surprisingly, the fewer restrictions that people
currently felt, the more likely they were to expect a negative
effect on their own work if strong copyright protection was
established by the current lawsuits about user interface issues.
But even those who already feel significant restrictions
predicted negative consequences if look and feel protection was
established, with an average of 2.303 on the five-point scale as
to their own work, and 1.70 as to the industry/community effect.
The average response of "no restriction" respondents was 1.464
for the industry/community effect.

How the Legal Debate Affected the Respondents' Views

     Perhaps the most dramatic finding from the survey was that
the more that those in the user interface field learned about how
copyright lawyers thought about copyright protection of user
interface issues, the more likely they were to think that
copyright protection for user interfaces should be weaker, rather
than stronger.

     People came to the CHI legal debate, in general, with some
familiarity with the legal issues involved in the look and feel
cases, for 64% of the respondents rated themselves as "moderately
familiar" with the legal issues before the legal debate, and
another 9% reported being "very familiar" with the issues.  Given
how much press attention the look and feel lawsuits have
received, and given how important this community feels the legal
issues to be to the health of their field, this result in itself
is not surprising.

     What was surprising was how people reacted to copyright
protection for user interfaces after they had heard the legal
debate.  Half of the respondents indicated that attending the
debate had caused them to change their opinion on copyright
protection.  Ten times as many changed their minds to thinking
that copyright protection should be weaker than changed to
thinking it should be stronger.

     It was not the case that Mr. Brown argued less persuasively
than Mr. Hemnes, for a reporter attending the debate found both
sides of the legal debate to be equally persuasive,[5] and a
number of respondents praised Mr. Brown's skill in argumentation.
The lawyers, quite appropriately to a legal debate, presented
arguments based on the issues that copyright law regards as
relevant to deciding legal disputes.

     What then explains the strong shift toward thinking
copyright protection should be weaker?  The authors believe that
the CHI audience was not so much persuaded to one legal position
or the other, but awakened to the nature of the legal debate and
its implications for how they worked and for the field in which
they worked.  The comment of one survey respondent expresses well
the authors' interpretation of the outcome of the legal debate:
"The arguments and session made me very nervous because the
arguments against strong protection were so compelling based on
my knowledge of the field, but they may not be anywhere near as
obvious to non-practitioners--and the courts are generally non-
practitioners."

     The respondents felt strongly enough about the predicted
harm to their industry that 63% of those who expressed an opinion
wanted SIGCHI to take an official position on the legal issues
based on the results of this survey.  Many who responded "no" to
this question said they did so because they thought the entire
SIGCHI membership should be polled before SIGCHI took an official
position.

Comparing the CHI Survey Results to Law Scholars' Report

     In part because of some novel legal questions presented by
the software copyright lawsuits, a group of ten intellectual
property scholars met last February at Arizona State University
to try to reach consensus on the proper application of copyright
law to the protection of computer programs.  Among the questions
they addressed was whether copyright protection was appropriate
for the look and feel of computer program user interfaces.  The
conferees agreed it was not, saying that use of terms like look
and feel "obscures rather than assists in the application of
copyright principles to software interfaces."[4]  The conferees
recognized that software user interfaces may be highly
functional, and to the extent they are, that copyright protection
is not available for them, nor for functionally optimal
expressions of them.  In addition, the conferees recognized that
user interface design may be constrained by technological
considerations that may limit the range of viable "expressions,"
which would restrict the scope of copyright protection available
to them.

     The conferees found in traditional principles of copyright
law an affirmation of the right to study and take unprotected
elements from copyrighted programs and reimplement them in other
products.[4]  While accepting that intellectual property
protection for computer programs should balance the needs of
innovators and competitors so as to promote the health of
industries such as that for software, the conferees regarded
their aim to be a limited one of articulating how traditional
copyright principles might be applied to computer programs, not
to offer their judgment about whether the larger goal of
intellectual property law can best be served by use of copyright
law to protect computer programs.

     The conferees, while in agreement with the user interface
survey respondents that individual commands or even groups of
commands should not be protected by copyright, were not able to
reach consensus on whether a less than functionally optimal
arrangement of commands (or icons) in a user interface would be
protectable by copyright.  Some conferees, like many in the CHI
survey population, thought that the benefits to users that would
flow from standardization of such things as command names and
command groupings in software user interfaces, as well as
functional reasons that might exist for grouping certain kinds of
commands together, made it inappropriate, in general, for
copyright protection to attach to arrangements of commands; other
conferees thought that in view of the protection copyright law
had traditionally afforded to compilations, the organization of a
set of commands from one program, if original, might protected by
copyright from exact duplication in a competing program.

     Though the conferees were in agreement with those in the
user interface field about some issues related to the protection
of icons--that "adoption of a functional general purpose metaphor
might limit the range of copyrightable expression" and that where
"the choice of icons is logically based upon the choice of an
overarching metaphor (such as a desktop) or the icon itself has
no fanciful characteristics, application of traditional
principles would preclude copyright protection for the particular
representation of the icon"--the conferees nonetheless thought
that traditional principles of copyright law might provide
protection for some more fanciful pictorial representation of
icons.  The CHI survey respondents, however, did not support
copyright protection for icons, perhaps because they perceive
user interface icons as not really having a fanciful character.

     The CHI survey data suggests that the user interface field
has developed because people in the field come to conferences
such as CHI to share their new user interface design ideas with
others.  When attendees see good design ideas and the research
that stands behind them, they feel they can incorporate these
designs into new products of their own, blending the ideas they
have gotten from others with the ideas they have come up with
themselves.  And they do not consider themselves thieves,
plagiarists, or copyright infringers when they do so, but rather
they consider themselves scientists and engineers who are
innovating on top of others' ideas in the kind of evolutionary
fashion which has exemplified development in this field.

     This evolutionary development seems to have brought about a
considerable amount of innovation, improved designs which have
made computers and software more accessible and usable by those
with minimal or no technical training, and competition about
performance, enhanced features, and price.  If each software firm
had to develop a different style of user interface than all the
others to comply with copyright law, there is concern that
copyright might impede how those in the user interface field do
their work, might harm the health of the industry, and might make
more difficult the achievement of the goal of making computers
usable by ordinary people.

Conclusion

     It is oft-stated, but nonetheless true, that a fundamental
purpose of the intellectual property laws is to provide
protection for innovations in order to give incentives for people
to be creative, thereby promoting progress in various fields of
endeavor.  From this, it follows that where legal protection of
some kinds of innovations is not needed to promote innovation in
a field, no protective legislation is needed.  Thus, it must
surely be the case that where intellectual property protection
would have a detrimental effect on innovation in a field, it
should be withheld, especially in an era where the law does not
as yet dictate the protection being sought.

     In the case of user interfaces, the CHI survey clearly
demonstrates that a significant segment of the leading designers
and researchers in the user interface field are overwhelmingly
opposed to strong copyright protection for user interfaces and
regard such protection as very likely to be harmful to the field,
rather than helpful to it.  These are the very people whom the
copyright law is supposed to be encouraging to be creative.
Given that copyright law has, as yet, not formed a firm position
about protection of various aspects of user interfaces discussed
in this article, and can be construed to support either side in
the legal debate, judges in the current round of copyright look
and feel cases should be receptive to considering the effect
strong protection would have on the industry.

     That the views of the technical community are largely
consistent with views expressed by legal scholars [4] as to the
application of copyright law to computer programs suggests that
judges in look and feel cases could find an adequate basis in
copyright doctrine to do what the user interface field thinks
would be in the field's best interest.


References

1.  Samuelson, P.  Why the Look and Feel of Software User
Interfaces Should Not Be Protected by Copyright Law,
Communications of the ACM, 32, 563 (May 1989).

2.  Samuelson, P.  Protecting User Interfaces Through Copyright:
The Debate.  Proceedings of Conference on Human Factors in
Computing Systems 97 (1989).

3.  Samuelson, P., and Glushko, R.  Comparing the Views of
Lawyers and User Interface Designers on the Software Copyright
"Look and Feel" Lawsuits.  Jurimetrics Journal, 30, 121 (Fall
1989).

4.  LaST Frontier Conference Report on Copyright Protection for
Computer Software, Jurimetrics Journal, 30, 13 (Fall 1989).


5.  Lewis, P. Legal Constraints on Sharing Ideas, New York Times,
5/7/89, F10.


Acknowledgments

     The authors wish to thank:  Gary Perlman and Jonathan Grudin
for reviewing a draft of the survey; Jonathan Grudin for his help
in production of the survey; Jack Brown and Thomas Hemnes for
their excellent legal presentations at the CHI legal debate; John
Leggett and his army of student volunteers who distributed and
collected the surveys at the CHI debate; Connie McFarland and
Andrea Lynn for data entry and preliminary data analysis; Dan
Sewell for statistical advice; Bill Curtis for encouraging the
idea of the legal debate at CHI and the survey; and Mark Hall,
editor-in-chief of the Jurimetrics Journal, for his editorial
assistance with the article in note 3 above, on which this column
is based.


About the Authors

     Pamela Samuelson is a Professor of Law at the University of
Pittsburgh School of Law.

     Robert J. Glushko is a Senior Staff Scientist, Search
Technology, Norcross, Georgia.

5.

