The text below represents commentary, not legal advice.

Return to: Critical Copyright, and more.

Return to: Paul Edward Geller, attorney.

 

As published in the Journal of the Copyright Society of the U.S.A., vol. 46 (1998), p. 39, and reprinted in Dear Images: Art, Copyright and Culture (Ridinghouse ICA, 2002), edited by Daniel McClean and Karsten Schubert.

 

Hiroshige vs. Van Gogh:
resolving the Dilemma of Copyright Scope IN remedying Infringement  

by Paul Edward Geller*

To say that copyright is "property," although a fundamentally unhistorical statement, would not be baldly misdescriptive if one were prepared to acknowledge that ... in practice the lively questions are likely to be whether certain consequences ought to attach to a given piece of so-called property in given circumstances.[1]

 

    This Article addresses the question: What should be the scope of copyright protection?  Part I will consider a hypothetical case to dramatize this question.  Part II will explain how attempts to determine the scope of protection lead into a dilemma basic to all copyright laws.  Part III will reframe infringement analysis with an eye to resolving the dilemma.  Part IV will apply the newly reframed analysis to the task of fashioning remedies.

 

i.  A Thought-Experiment

 

    In the latter half of the nineteenth century, European literati and artists became captivated by Japanese art, especially prints.  When in Paris, Vincent van Gogh, a collector of Japanese prints himself, painted the Portrait of Père Tanguy.  Tanguy had an art-supply shop in which he often displayed the works of avant-garde painters.  The background of the Portrait of Père Tanguy is covered with Japanese prints.[2]

    In the Fall of 1887, Van Gogh painted his Japonaiseries.[3]  The most notable are based on a pair of works by Utagawa Hiroshige from the series One Hundred Views of Famous Places in Edo.[4]  Hiroshige, in his wood-block prints, composed in clearly delineated levels of space and subtly gradated colors; Van Gogh, in his oil paintings, pushes and pulls our vision with agitated brush-strokes and boldly opposed colors.  In the first exhibit to this Article, we see one of Hiroshige's pair of prints, The Plum Garden at Kameido; in the second, Van Gogh's study of that print, The Tree.[5]

    How would Hiroshige have responded had he known what this foreign enthusiast, Van Gogh, had made of his works?  It might be a mistake to attribute to Hiroshige our litigious impulse to dispute Van Gogh's decision to make studies of his graceful prints in viscous oils.[6]  While traditional Oriental culture might have demanded that a student acknowledge a master for creating models worthy of imitation, it did not threaten the student with legal sanctions for copying such models, even badly.  Hiroshige might have thought, like the Chinese scholar Shen Zhou before him, "if my poems and paintings, which are only small efforts to me, should prove to be of some aid to the forgers, what is there for me to grudge about?"[7]

    Assume that Hiroshige's prints were protected by copyright in France toward the end of the nineteenth century.[8]  Suppose, more hypothetically, that one of Utagawa Hiroshige's heirs, holding his rights, then sued Vincent van Gogh for making studies of the prints and Vincent's brother Theo, or another of his heirs, for offering these studies to the public.  This hypothetical suit is not posited to test the state of French copyright law over a century ago, but rather as a thought-experiment to elucidate the dilemma which arises with the basic question: What is the proper scope of copyright?  Since this dilemma is common to all copyright laws, the attempt to resolve it here will make use of comparative legal analysis.[9]

 

II.      The Dilemma: Copyright Scope

 

    The field of copyright has distinct dimensions.  On one dimension, authors create works.  Here a court has to ask:  Is copyright in one work infringed by another work?  On the other dimension, works are disseminated to the public.  Here the question becomes:  What remedies should the court grant against the dissemination of an infringing work?  Our dilemma will be considered on both dimensions.

 

A. The Rate of Cultural Feedback

 

    These dimensions, though distinct in theory, are intertwined in practice.  Creators elaborate works in communicating with colleagues and critics, patrons and promoters, and larger publics.[10]  Technology has progressively globalized the communication networks into which creators have fed, and feed, their works.  To assess this progress, start with the fact that, two millennia ago, the Roman Empire and the Han Dynasty of China did no more than sporadically trade with each other.  Only later did Greek art forms, previously brought by Alexander to the frontiers of India, filter into the Far East over the Great Silk Road.[11]  Sill later yet, Chinese inventions essential to globalization, such as the magnetic compass and paper, reached the West.[12]

    Creations are thus fed back into communication networks at varying rates.  To take an example closer to our time, the art worlds of modern Japan and Europe initially learned of each other's rapidly developing styles with significant delays.  The Japanese art of the "floating world" had already produced such greats as Utamaro in the eighteenth century, but it most notably synthesized Oriental and Western models of landscape art in the work of Hokusai and Hiroshige in the early nineteenth century.[13]  While Europe had only fragmentary contacts with the full range of artistic developments in Japan before the middle of the nineteenth century, European access and response to this art had expanded greatly by the time Van Gogh reached Paris in 1886.[14]

    This rate of feedback is critical to the rationale of intellectual property.  Farmers would be reluctant to market produce that was promptly stolen once it went on sale.  But the law recognizes property in the fruits of the farmers' labor and polices the marketplace for thieves.  Authors and inventors would lose incentives to release their products of mind to the world if, in the process, they found such public goods pirated by free-riders.[15]  Instead, armed with intellectual property, they are supposedly induced to feed more works and inventions back into communication networks, in turn further stimulating culture and technology.  As the rate of feedback increases, so do the chances that new ways to create, for example, new plot lines, musical harmonics, artistic techniques, etc., come to light.[16]

    Such feedback helps to achieve policies that motivate intellectual property.  The Constitution of the United States speaks of promoting "the Progress of Science and useful Arts."[17]  Starting in the French Revolution, Continental European doctrine posited authors' rights as "sacred" natural rights, but the Revolutionary legislators understood full well that such rights had to be codified with an eye to enhancing culture.[18]

 

B. Infringement Analysis or Exceptions?

 

    Our hypothetical case squarely raises questions of copyright scope.  Should the law entitle Hiroshige or his heirs to stop Van Gogh from making studies of Hiroshige's prints?  To stop Van Gogh's heirs from feeding resulting new works back into world culture?  To impose liability for making or exploiting these works?  Such questions lead into the following dilemma:

    If, on the one hand, creators had no legal control over the fate of their products of mind, they would lose reasons to input such public goods into communication networks.  On the other, with too much control, they could stop others from elaborating on their creations and from releasing still further products of mind to the world.  If the rules on point are too lax, copyright becomes ineffectual for accelerating the feedback of new works.  If the rules are too stringent, copyright becomes counter-productive by burdening such feedback.[19]

    This dilemma may be reconceptualized in doctrinal terms.  Copyright arises out of creative acts: Hiroshige has copyrights in the prints he creates, and Van Gogh has them in the studies he creates.  In principle, copyrights, so-called properties, entitle creators to dispose of their respective works at will; however, in our hypothetical suit, such rights could well conflict.  That is, Hiroshige's right in a print, if given force without regard to other creators' rights, could lead to stopping Van Gogh or his heirs from exercising this later creator's right to dispose of his study of that print.  But the fact that Hiroshige's work comes first ought not be decisive for the simple reason that, unlike patent law, copyright law does not favor works or authors because of priority in time.  The follower, Van Gogh, may further invoke freedoms of creation and expression.[20]

    Courts, when faced with such dilemmas or conflicts, want to find equitable and just solutions.  They are like Goldilocks who, in the home of the three bears, wanted porridge that was neither "too hot" nor "too cold."  While acknowledging Hiroshige's work as calling for protection, a court might sympathetically hear counter-arguments.  Van Gogh could claim that his works are only private studies, or his heir could argue that his works ought not be kept from the public.  The court may well ask:  Does any equitable or policy-based defense, sometimes a provision setting forth an apparent exception, excuse arguable infringement?  In the United States there is the doctrine of fair use, and Germany allows transformative uses under the narrower doctrine of free utilization (freie Benutzung).[21]  Other jurisdictions permit such transformative uses as parody under specific exceptions, while still others subject such uses to general infringement analysis.[22]  But in hard cases, it often becomes difficult to distinguish fair uses, or even specifically exempted uses, from those which are arguably infringing.[23]

    True dilemmas are never fully resolved by Goldilockean compromises.  Rather, courts have to resolve our dilemma consistently with basic copyright rationales.  Unfortunately, the judicial maneuver of seeking ad hoc compromises can scramble the lines between true exceptions and infringement analysis and, at the same time, make a muddle of underlying rationales.  In a seminal case in the United States, suit was brought for abridging George Washington's writings, and the Supreme Court asked whether this arguably transformative use was "a justifiable use of the original materials, such as the law recognizes as no infringement of the copyright."[24]  It is symptomatic of subsequent confusion that the Court's decision is now seen as significant for the evolution of the so-called exception of fair use, even though the decision merely upheld a finding of infringement.[25]  One of the four criteria set out in the statutory provision for fair use continues to invite courts in the United States to re-test infringement by asking: Does the copying in question lack "substantiality"?[26]  It also remains unclear in German and Swiss law whether the doctrine of free utilization serves as an exception or does anything other than limit infringement analysis.[27]

    Most importantly, infringement analysis is common to all copyright laws.  By contrast, such doctrines as fair use and free utilization are idiosyncratic to particular copyright laws, and specific exceptions vary considerably between all these laws.  In approaching the common dilemma of copyright scope, only infringement analyses common to the major copyright systems will here be considered: no further analytic tools will be used.[28]  It should be stated upfront: it will not suffice to clarify criteria for determining when copyright infringement does or does not take place.  For purposes of resolving the dilemma of copyright scope, such all-or-nothing determinations will prove to be too crude.  Rather, as explained in the next step of the argument here, courts need to reach more finely differentiated findings of infringement.  After that, it will be a matter of calibrating such findings with the different remedies available for infringement.[29]

 

iii.     Creating Works: Infringement?

 

    Asking when infringement takes place engages analysis along the dimension of creation.  It focuses on pairs of works: on the one hand, the work of a prior creator, in our case Hiroshige; on the other hand, the work of a later alleged copier, in our case Van Gogh.  In such a case, infringement analysis compares the prior and later works.

 

A. Complementary Doctrines for Hard Cases

 

    Our hypothetical suit poses a hard case of infringement to dramatize our dilemma.  Let us start by sorting such hard cases out from easy cases and by broaching the complementary doctrines that help courts to decide hard cases.  Such doctrines, developed in different copyright systems, guide the analysis of protectability and, after that threshold issue, of infringement itself. 

 

1. Sorting Out the Hard Cases

 

    Courts avoid our dilemma in easy cases, where they are faced with only literal or close copies.  Such copies, displaying little or nothing new above and beyond the works copied, cannot in themselves enrich the creativity that feeds culture.  The dilemma becomes acute, however, when prior works are admitted to be the bases of later works that are therefore alleged to be infringing.[30]

    In the eighteenth century, copyright was instituted to deal only with easy cases, the pirate reprinting of books or restaging of plays.  At the start of the nineteenth century, courts typically found no infringement in what leading French commentary called "[t]he transmutation of form that the translator causes the original to undergo."[31]  But in the course of that century, as trade in books became increasingly globalized, authors and publishers started to claim rights to stop translations in foreign markets.  Ultimately, the right of translation was subsumed under the more general right to control the making and exploitation of derivative works.  The scope of copyright was effectively expanded beyond protecting prior works against substitution by later works in the markets that the prior works targeted.  Copyright reached new markets in new media, for example, as literary works were adapted to the stage or film.[32]

    In hard cases, plaintiffs allege that defendants have derived works from their own.  In such cases, courts often resort to complementary doctrines to limit copyright scope.  These doctrines operate on theoretically distinct levels of analysis that tend to come together in practice.[33]  On the level of determining what is protectible, there is the principle that courts may not protect "ideas" or "facts," but rather only "expression" or "forms."  On the level of finding infringement, courts ask whether plaintiff's work is copied in defendant's "substantially" similar work or whether "essential" or "characteristic traits" of one work are taken in the other.  But such notions as "ideas" defy ready definition, and equally metaphysical notions of "substance" suggest that works of the mind are things like tables and chairs, consistently perceived by all audiences, but none of these doctrines by itself guides courts to consistent decisions.[34] 

    For example, in the United States, the holder of copyright in the novel Gone with the Wind brought suit for infringement by the novel The Blue Bicycle.  But the suit went no further than the court's refusal to grant a preliminary injunction once it found similarities only between idea-determined materials such as stock "characters" and "scenes and the sequence of events."[35]  In the French case involving the same novels, the court of first instance found infringement in similarities between just such "intrinsic elements," while the intermediate trial court held that such materials, notably characters and plot, could not be protected.[36]  Although the highest French court of appeal overturned that intermediate judgment for insufficiently specified findings, the court on remand found that the materials in question had been so creatively transformed that no actionable similarities remained.[37]

    The right to control derivative works and any limiting doctrine, such as the idea-expression distinction, tend to struggle with each other, like a pair of overly oiled wrestlers.  Unfortunately, whether the right or the limiting doctrine wins all too often depends on a momentary false step rather than any consistently balanced judgment from case to case.  It is relatively easy to ascertain when copyright may be invoked to prevent copies from substituting for a specific work in a given market, but it becomes more difficult to determine when to stop protecting a prior work against later works in other markets that the prior work does not address.[38]  Courts and commentators have refined doctrines limiting copyright scope in this regard, but they have done so largely in case-by-case analyses.  Basic refinements in major jurisdictions will be touched on schematically here. 

 

2. Creative Options and Protectability

 

    In the case of the play Abie's Irish Rose, Judge Learned Hand formulated these classic, cautionary words concerning the distinction between ideas and expression:

Upon any work, and especially upon a play, a great number of patterns of increasing generality will fit equally well, as more and more of the incident is left out. The last may perhaps be no more than the most general statement of what the play is about, and at times might consist only of its title; but there is a point in this series of abstractions where they are no longer protected, since otherwise the playwright could prevent the use of his "ideas," to which, apart from their expression, his property is never extended.[39]

    In hard cases, courts sort out "ideas" and "expressions" with diverse tests and criteria.  All these judicial devices will be generically considered here under the rubric of creative options.  This term will indicate the range of expressions potentially available to elaborate a given work.  Courts in the United States apply the so-called merger test to measure whether sufficient creative options are present.  To the extent that an idea can be realized in only one or a small set of expressions, there is merger between the idea and the expressions that, accordingly, are not protected.[40]  Continental European doctrines comparably lead courts to grant or withhold protection to the extent that underlying ideas do or do not allow for sufficient creative options to elaborate original works.  French jurisprudence refuses protection when "an automated or constraining logic" dictates how the author's mental input (apport intellectuel) is given form, while German doctrine speaks of a more or less open "room for the play" (Spielraum) of creativity.[41]

    As a corollary, courts may be said to "filter" some expressions, such as scènes à faire, functionally determined software modules, etc., out of the scope of protection.[42]  For example, in one French case, a researcher sued a writer for taking expressions from his scholarly works on Cajun culture to use in a novel set in a Cajun locale.  But the trial court found no infringement in the use of such culturally typical expressions that it considered to be indispensable to giving the novel "a minimum of verisimilitude."[43] Moreover, to the extent that there is only one way, or very few ways, to convey facts accurately in compilations such as lists, charts, etc., resulting expressions may not be protected.  Obviously, there can be just one correct spelling and numbering for each listing in a telephone directory and only a few conventional schema or signals for situating sites on maps.[44]  To put the matter more generally, whether idea-determined materials may be protected against copying will turn on how they fare under tests or criteria to which doctrines of creative options lead in the cases.  For example, in Romeo and Juliet or Abie's Irish Rose, there were only so many ways that a family feud might have served as background for the particular tragedy or comedy into which each play cast its young lovers.[45]

    All these doctrines do not always dictate precise tests or criteria.  Courts often glibly assume that we all know ideas when we see them.  However, the term "idea" often evokes only evanescent Platonic flashes of insight or mental will-of-the-wisps.[46]  As a result, courts often have difficulty specifying at what point a set of creative options is too restricted to allow for protection, that is, how coarsely or finely to filter out idea-determined materials.  For example, in one case in the United States, copyright was asserted in a so-called pitching form into which such baseball statistics as game times, betting odds, pitchers' performances, etc., were compiled.  But one judge on appeal went to the heart of the matter, effectively asking: What is the "proper level of abstraction" for formulating such categories of data, more generally, for grasping "the ideas which underlay" the expression in the work at issue?[47]  Logically, without knowing just how to focus on any idea against which merger or creative options might be measured, it becomes impossible to say when there are too few ways to express that idea to protect corresponding expressions.  The ambiguity of the term "idea" also obscures the borderline between copyright law, in which ideas are not protected independently of expression, and patent law, in which they are not protected independently of utility.[48]

    In response to these definitional vagaries, consider the sense of "idea" in the statutory terms of any "procedure, process, system, [or] method" for generating works and thus "embodied" in works.[49]  Classically, Aristotle explained the idea of tragedy in such operational terms: have a hero commit an act of hubris, and have him fall and discover his own fall, to accomplish a catharsis of pity and fear in the audience.[50]  Contemporary aesthetic analyses have unpacked entire systems of comparable formulae for elaborating different plot lines that appear in genres starting with folk tales and extending to modern stories.[51]  Of course, works differ in how they incorporate procedures or methods: for example, a musical score instructs performers on how to play the musical work it represents, in what key, with what timing, etc.  In videogames, audiovisual displays are run by interactive software procedures, while paintings embody methods of composition, coloration, etc.[52]

 

3. The Sliding-Scale Analysis of Infringement

 

    In easy cases, courts find infringement in the full reconstruction of plaintiff's work in a literal or close copy.  In hard cases, the question remains open: how much of the creative fabric of a work must be re-generated before there is infringement?  Judge Hand warned against responding to this question once and for all: "Nobody has ever been able to fix that boundary, and nobody ever can."[53]  Hard cases are endemically subject to such line-drawing exercises that may well start at the threshold stage of analyzing protectability, but that all too often slip unnoticed into the stage of analyzing infringement itself.  Before tightening up analysis, it will be useful to survey how it tends to shift from stage to stage.

    Since creativity is protean, standards of protection, such as "originality" or "creativity," are satisfied to varying degrees from case to case.  In the United States, the Supreme Court indicated that a minimally creative compilation of facts would attract "thin" protection against literal or close copies.[54]  By implication, a more creative work would call for "thicker" protection, not merely against literal or close copies, but against translations, adaptations, or other derivative works.  Continental European doctrines, typically predicating degrees of creativity that can vary from work to work, comparably allow for the sliding-scale analysis of infringement.[55]  Courts in diverse jurisdictions may then adjust the scope of protection in the light of the respective creativity of plaintiff's and defendant's works.[56]

    Unfortunately, judicial penchants can at times push and pull such sliding-scale analysis ambivalently.  Consider, for example, a pair of cases in Germany where copyright law, in principle, has traditionally set the ostensibly high standard of protecting only "personal intellectual creations."[57]  In one case, the courts, generous to a plaintiff, seemed to short-circuit that standard by indulging the premise that industrial-drafting techniques of perspective, shadowing, etc., allowed for introducing sufficient personal creativity into a simple technical drawing to protect it against close copying.[58]  In another case, judicial sympathies favored the defendant: the image of a female nude ingeniously posed by the innovative fashion photographer Helmut Newton was at issue, but a painting which creatively transformed that image, albeit recognizably, was found not to infringe Newton's copyright.[59]

    Realistically viewed, courts can at times be generous toward plaintiffs who face close or literal copies and, at other times, tolerant toward defendants who transform prior works.  Professor François Dessemontet notes the opening for possibly inconsistent results:  "The outcome in a given case might well be that, while the plaintiff's work could be protected against outright piracy by a third party, it does not receive protection as against defendant's work that incorporates parts of it less obviously or more creatively than any slavish copy."[60] 

 

B. Reframing the Doctrines into More Coherent Analysis

 

    To reconcile such judicial approaches, doctrines of creative options will here be more tightly tied into the sliding-scale analysis of copyright infringement.  Hopefully, as it is made more coherent, such analysis will guide courts more consistently with underlying copyright rationales.  But analysis is here reframed only experimentally; other attempts to pull it together might work as well or better.

 

1. The Spectrum from Copying to Creation

   

    Our operational definition of "ideas" encompasses a wide range of procedures and methods for generating works.[61]  At one end of this range, distinguish as routines those information-processing procedures so fixed and detailed that they leave little or no room for creative options in themselves.  Moving forward from that starting point, distinguish the following processes along a spectrum from copying to creation.[62]

a.   At the start of the spectrum, rote copying results from applying a single set of identified routines to a work.  For example, to copy a text literally, with the same wording, we might change fonts, using as many routines as there are printing symbols, plus some others to reformat the text.  Or, to copy an image closely, we might trace it out on transparent paper, or we might photocopy it, laying it into the photocopy machine, setting the desired number of copies and other parameters, and pressing the button.  Either way, a given set of identified routines suffices to obtain our copy.

b.   There are many different processes in the middle of the spectrum.  The rubric of knowledgeable reworking seems to cover most of them.  Such processes use routines that are generally known, but not necessarily all identifiable.  Nor need the set of such routines used in a given case fully suffice to determine the entire work generated in that case.  Consider translating a French cookbook into English: it is necessary to rely on lexical and syntactic routines known, sometimes only implicitly, by bilingual speakers.  For example, while French usage regularly places adverbs between verbs and direct objects, English-speakers most often relocate the adverbs, usually to the beginning or end of clauses.  In any event, a straightforward cookbook, if not almost every work-a-day text, is susceptible of only a slightly variable set of translations likely to be acceptable to bilingual speakers.  As other examples of mid-range processes, consider selecting and organizing facts into a compilation, or excerpts into an anthology, or recontextualizing a work, as in appropriation art.[63]

c.   At the far end of the spectrum, there is innovative recasting.  Here, no known set of routines suffices for moving from one work to another or even to comparable works.  Consider the enterprise of translating James Joyce's Ulysses into French: at hundreds of thousands of points, the translators had to make choices that linguistic rules alone could not have dictated.[64]  The saxophonist Charlie Parker provides another example: starting from the chords of the tune Cherokee, which an accompanying guitarist had inverted just "to keep a beat going," he moved into an entirely new mode of jazz improvisation.[65]

    At the start of the spectrum, rote copying can be automated.  When a photocopy machine copies an image, a microprocessor in the machine executes a complex set of routines.  Rote copying comes about for the simple reason that, before formalizing these procedures in any computer algorithm, the programmer must have identified them all.  Imagine a more complex case: a computer, digitally recording audio input, for example, a musical work, converts all the sounds of that work into video output, which is screened as moving visual patterns.  The computer program effectively reprocesses the musical work, its rhythms, melodies, chords, timbres, etc., visually by playing these materials out as lines, light and dark, colors and values, etc., all moving in space and time, apparently as an abstract cinematic work.  Assume, too, that the algorithm used here is in an off-the-shelf program, which the party making the conversion does not creatively supplement, say, through interactive input.  That is, without such input, this party makes a rote copy.[66]

    Moving further along the spectrum, mid-range processes such as knowledgeable reworking can only be partially automated.  For example, an expert system can be used to help with the translation of a text, but the routines of such systems succeed with variable confidence factors in translating all but the simplest texts.  These factors represent the probabilities that the expert system will translate given passages to the satisfaction of bilingual speakers, and they can vary depending on the difficulty of any given passage and the languages in question.  Such variability will arise even in using an expert system to translate a work-a-day text like our hypothetical cookbook from French to English, both closely related languages that share largely common roots.  It will every so often prove necessary to call upon the implicit knowledge of bilingual speakers to decide between proposed translations that are arguably acceptable or equally debatable.[67]

    At the far end of the spectrum, innovative recasting largely defies automation.  Suppose that a text to be translated, like our example of James Joyce's Ulysses, breaks systematically with most cultural precedents.  As Umberto Eco emphasizes, such a work can challenge us "to rethink the whole language, the entire inheritance of what has been said, can be said, and could or should be said."[68]  Innovative recasting is needed to translate Joyce's novel into French, since no bilingual consensus provides rules to guide translation at most points in this seminal English text.  Of course, run-of-the-mill works can also be innovatively recast, as when popular tunes became fodder for modern jazz.[69]

    A word of caution is called for at this point.  Infringement analysis need not look to the actual genesis of the works in question.  It is rather a matter of ascertaining to what extent there were routines available to move from plaintiff's to defendant's work at the time of the later work.[70]  To the extent that available routines did not suffice, it may be inferred that the allegedly infringing work arose thanks to processes that ranged from knowledgeable reworking to innovative recasting.  The computer metaphor, just used, provides merely one way to think about such a determination.  Expert testimony will often be helpful in making the requisite findings.[71]

 

2. Application to Literary and Artistic Works

 

    Our framework of analysis ostensibly applies no matter what type of work is at issue.  But, once again, there is a difficulty that arises from protean creativity, which manifests itself in the ever-proliferating diversity of media and genres.  Indeed, some commentators suggest that infringement analysis changes radically from one type of work to another, for example, from literary to artistic works.[72]  Let us then test our framework of analysis by asking whether it can be cogently applied to diverse types of works.

    On the one hand, literary works are coded in discrete terms, such as words and phrases, arranged pursuant to grammatical rules.  On the other, artistic works are embodied in continuously variable materials such as line, space, light, and color, configured in space.  In computer terms, a literary work can be generated using a lexicon like the ASCII code and a sequencing or syntactic program, while an artistic work can be more or less faithfully bit-mapped.  The same distinction can be reconceptualized as applying between works repeatably readable from texts and works uniquely embodied in objects.[73]

    This distinction admits of both border-zone and hybrid cases.[74]  For example, a musical work is more like a literary work when it is written down in a score, but it moves into a border zone, where it becomes more like an artistic work, when it is performed.[75]  Chinese poetry forms a hybrid case when written in calligraphic form: no translation into an alphabetic language, no matter how creative, can render the resonance of such a poetic text with its visual form.[76]  How can a court assess where works fall along the spectrum of creative processes, given that works tend to be so diversely situated relative to the basic distinction between literary and artistic works?

    In a literary work, like a novel, it is a matter of analyzing, most notably, how the text is sequentially translated into another language or sequentially adapted into another medium or genre.  For example, Dashiell Hammett's detective story The Maltese Falcon was made into John Huston's film by dramatizing it into dialogue, acting instructions, etc., and by visually elaborating it into sets, action, camera work, etc.  Making Hammett's text into a screenplay required knowledgeable cuts and adjustments to fit the story into the time, censorship, and cinematic constraints of the motion picture.[77]  Huston's filming of the story moved up the spectrum toward innovative recasting in the ways the motion picture was shot, paced, and edited.

    By contrast, it is often a subtle matter to analyze how one artistic work is transformed into another.  The materials of such a work, for example, graphics, light and shade, colors and values, etc., can be so packed together, its visual signals so saturated, that they become inextricable to the untrained eye.  Nelson Goodman, whose philosophical analysis is seminal here, gives the example of a print by Hokusai, in which "[a]ny thickening or thinning of the line, its color, its contrast with the background, its size," would change the entire impact of the work.[78]

    Look at the exhibits to this Article, reproducing a pair of works in question in our hypothetical suit.   Of necessity, Hiroshige's Plum Garden reappears in Van Gogh's Tree, since the prior print is the subject matter of the later study.  Furthermore, the bare similarity of their patterns neither proves rote copying nor disproves knowledgeable reworking or innovative recasting.[79]  To compare the earlier and later works in this regard, it is necessary to disentangle differences far more tightly woven into their expressive fabrics than overall similarities.

 

3. Illustration in our Hypothetical Case

 

    Recall the test for sorting out creative processes:  Can any one set of known routines suffice to move from a prior to a later work?[80]  To the extent that routines do not suffice, knowledgeable reworking or even innovative recasting might be at work.  Now apply this approach to Hiroshige's Plum Garden and Van Gogh's Tree, which we already contrasted briefly at the start.[81]  Here is how Klaus Berger, a leading expert on "Japonism" in Western art, focuses on the movement from one work to the other:

... the bright orange framing strip and the added, entirely random Japanese characters point in a new direction that lies beyond both Western  and Eastern traditions.  The exaggeration of the graphic arabesque, and the dense, ungradated masses of paint, combine to destroy the Japanese equilibrium between drawn framework and rhythmic colour.  ... The gnarled black bough screams aloud; the scattered blossoms are trapped between those fateful black tracks and the red wall of the sky.  Whereas in the Japanese work everything expands into space, here it is confined, shut off, dramatically exaggerated.  The curves and verticals seem to fight each other, and the green configuration below thrusts against the upper red quarter, conveying the suspense of a conflict that still hangs in the balance.[82]

    Starting from Oriental models, Hiroshige's space hints at classic Western perspective and anticipates Impressionism as well.[83]  His print The Plum Garden moves the eye back from more intensely colored foregrounds to paler backgrounds, while Van Gogh's study The Tree scrambles all such inherited visual signals, for example, drawing our eye under the red sky and bursting yellow blossoms into a constricted space punctuated by angular tree trunks.  Our expert is categorical about the revolutionary step Van Gogh is taking beyond the Impressionism which formed the background for his mature work: "The result is a kind of explosion that opens the way to Expressionism."[84]

    Contrast another pair of works not in the exhibits: Hiroshige's print Sudden Shower over the Great Bridge near Atake with Van Gogh's study The Bridge.[85]  The older artist clearly blocks out a simple Spring shower which falls on passers-by who are walking across a solid bridge planted in a placid river.  The younger artist brushes in driving rain, a bridge uneasily suspended on moody pilings, and a turbulent river, all with the tensions of a raging storm.  Our expert wonders "what the unfortunate people in the picture have done to deserve such an elemental onslaught."[86]

    Is Hiroshige's copyright infringed by Van Gogh?  On the spectrum starting at rote copying and including knowledgeable reworking, Van Gogh ultimately reaches the stage of innovative recasting.  It would nonetheless oversimplify matters to declare in the abstract which of these processes infringes copyright and which does not.  To resolve the dilemma of copyright scope, it will prove necessary to distinguish more finely between the consequences that a court may give to these processes.

 

iv.     Disseminating Works: Remedies?

 

    It is time to change from the dimension of creation to that of dissemination.  If the court orders Van Gogh to stop painting his studies, it nips dissemination in the bud.  If it orders Van Gogh's heir not to display these studies, it stops dissemination in its tracks.  The impact of awarding money to Hiroshige's heir will depend on the amount of the award.

 

A. Property versus Liability Remedies

 

    It is helpful at this point to distinguish between property and liability regimes.[87]  When a court is ready to order you off someone else's land, whenever you set foot on it, a property remedy is available.  When the court merely holds you liable to pay money after you have already trespassed, a liability remedy is imposed.  As Professor Jerome Reichman explains, the optimum mix of property and liability remedies will vary according to the branch of intellectual property and the case.[88]

    Our dilemma arises precisely because Hiroshige may claim property, not merely in the prints he created, but in the later studies Van Gogh created of those prints.  Hiroshige's property claim would be most decisively implemented in an injunction to preclude Van Gogh or his heirs from exercising any further property claim to dispose of his later work freely.  As already pointed out,[89] however, not only does copyright law not contain any principle, as does patent law, favoring any initial creator because of priority in time, but it is subject to freedoms of creation and expression.  In reaching an appropriate mix of remedies, it will be argued, a court may resolve our dilemma in the case before it, while avoiding conflicts between prior and later creators' claims.[90]

    Typically, under copyright law, the mix of remedies has erratically varied from case to case.[91]  This variation has in part arisen because the doctrines limiting copyright scope, such as the idea-expression distinction and substantiality criteria of infringement, have been neither uniform in meaning nor coherently articulated.[92]  Furthermore, in considering remedies, courts have to take account of equitable considerations that change from case to case, for example, the force of the parties' initial showings, the eventual impact of proposed remedies on the parties' respective positions, etc.[93]  It is submitted that courts should fashion and coordinate property and liability remedies within a framework of analysis such as that  proposed here.  In particular, they would do well to situate where works fall along the spectrum from rote copying to innovative recasting.[94]

    The following guidelines are proposed to tighten up present copyright law, with due regard for equitable considerations present in any given case:

1.   No injunction or other coercive remedies should be issued against whoever makes a solitary copy exclusively for private enjoyment or study.

2.   Courts should (a) always be ready to enjoin and otherwise provide coercive relief against rote copying for the public, (b) exercise discretion in granting injunctions and other coercive remedies concerning works generated by mid-range processes such as knowledgeable reworking, and (c) refrain from enjoining innovative recasting, as well as the dissemination of the new works thus generated, absent strong equitable reasons for stopping them.

3.   Courts may (a) impose the full range of monetary awards, including statutory or other special damages used for punitive or deterrent purposes, in cases of rote copies, especially when these are marketed with scienter, but (b) adjust actual or statutory damages, or reasonable royalties or profit shares, to the market interests at stake in any other case.

    How would these guidelines apply to our hypothetical suit?  Assume that the court finds that Hiroshige's prints are innovatively recast in Van Gogh's Japonaiseries.[95]  First, no order would be issued to stop the making of these studies privately; second, no order would lead to seizing these studies or enjoin publicly displaying or marketing them.  Third, a money judgment could still be obtained for such marketing, and it would be measured by considering how Hiroshige's prints contributed to the appeal that Van Gogh's studies would hold for relevant audiences.  Our dilemma would be resolved: neither creativity nor feedback into cultural networks would be blocked, while both prior and later creators' rights would be protected on the marketplace.[96]  Having glanced at how our guidelines come together in our hypothetical case, let us now look at how they may apply more particularly in other hard cases.

 

B. Applications to Different Remedies

 

    Works of authorship display different creative processes in endlessly variable combinations and permutations.  As a result, the guidelines just proposed would still lead to different mixes of remedies from case to case, although hopefully with more rhyme and reason than before.  Some comments are in order on how to reach optimum mixes in complex cases.

 

1. Injunctive and Related Relief

 

    At the level of injunctions and other coercive remedies, there is an obvious difficulty.  A framework has been proposed for the sliding-scale analysis of infringement.  But how does it help in making all-or-nothing decisions, notably whether to enjoin or not?  Of course, courts should try to tailor remedies to the cases before them as closely as possible.[97]  But all too often, judges are confronted by hard choices, with no Solomonic compromise possible.  The guidelines proposed here are clear and bright in some of these cases, but call for refinement in others.

    Regarding private copies, a French commentator already explained at the turn of the century:  "A copy made as a [private] study is exempt from remedies for infringement."[98]  Courts and legislators have to avoid invading privacy, whether of creators or of end-users of works.  Refraining from coercive relief against private copying avoids the risks both of intimidating the timid muse that inspires creation in private and of violating constitutionally protected privacy interests.  For example, the law does not stop computer programmers from making private copies of programs to discover methods and codes underlying these programs.[99]  Privately made studies are also indispensable for younger artists who conduct research into artistic methods in copying older works.  They need to get the knack of such methods manually as well as in inspecting the works visually.[100]

    Faced with public dissemination, courts have to decide whether or not to grant injunctions or other coercive remedies.  Courts protect copyright owners' markets by stopping copying, by seizing hard copies, or by preventing sales or showings.  It has been proposed that discretion to issue such orders be exercised with an eye to where plaintiff's and defendant's works fall along the spectrum of creative processes.[101]  Courts are most ready to enjoin or seize rote copies, that most obviously threaten markets the copied work targets but that bring nothing new into the marketplace.  In mid-range cases, where knowledgeable reworking generates translations, adaptations, etc., discretion has to be exercised to balance the risk of allowing harm to protected markets against the risk of obstructing the feedback of new content.[102]  For example, the Ninth Circuit thought it inappropriate to enjoin the exploitation of the film Rear Window when that work turned out to be subject to renewal rights in the short story on which it had been based.  Such an injunction would not have advanced the right-holder's market interests, and it made little sense to deny "the public the opportunity to view a classic film for many years to come."[103]

    Of course, how discretion is exercised may vary according to procedural options available in a given legal system.  For example, a judge may be more ready to grant a preliminary injunction pending an expedited determination on the merits than the long preparation of a jury trial.[104]  At the far end of the spectrum, where a prior work is innovatively recast into a truly different work, there tends to be less risk that markets properly reserved to the prior work might be prejudiced by the later work.  At the same time, the argument for declining to enjoin the different and later work becomes all the stronger to the extent it represents new expression to release to the world.[105]

 

2. Damages and Other Awards

 

    The difficulty at the level of monetary awards is not so obvious.  It would seem that such awards can be easily adjusted within our framework of sliding-scale analysis.  After all, the court can award more or less money to plaintiff depending on how massive or marginal defendant's taking is found to be.  It is not so simple, since the measures of monetary awards are diverse: some are granted for punitive or deterrent purposes; some, to compensate plaintiff for usurped markets or other damages; some, to recover defendant's undue gains.[106]  The proposed guidelines suffice to coordinate such measures in simple cases, but they call for refinement in more complex cases.

    For punitive or deterrent purposes, many courts may assess monetary awards, especially statutory or other special damages, at levels beyond the market interests at stake in the cases before them.[107]  There is no risk that the prospect of paying such extraordinary damages might discourage the release of new creations to the world when they are levied against rote copies, for the simple reason that such copies feed nothing creative into communication networks.  Quite the contrary, incentives to release original works run the risk of being blunted unless the unauthorized marketing of rote copies as substitutes for original works is discouraged.[108] But there is the risk that awards out of proportion with market stakes might burden the feedback of new content in cases where reworked or recast works are found to be infringing.  Such cases call for purely compensatory awards, notably damages, royalties, or profits, or else statutory damages granted in lieu of such awards.[109]

    The amount of any compensatory award has to be gauged in the light of the relative impacts of plaintiff's and defendant's works on the marketplace.  It is easy enough to measure damages in a case where plaintiff's work finds itself forsaken because defendant's work merely serves as a substitute for it on the markets it addresses.  It is then a matter of measuring losses, such as lost sales of copies or of tickets, other lost payments to access or show the work at issue, lost royalties accruing from such uses, or even arguably lost good will.[110]  In other cases, plaintiff might not yet have put the work at issue on the market or might not have had any market success, or the work might be transformed by defendant into a derivative work that addresses different markets, for example, as when a novel is adapted to film.  In neither event has plaintiff had damages in any actual market, so that reasonable royalties or shares of defendant's profits seem to be the only workable forms of monetary compensation.  But it cannot be assumed, in conclusory fashion, that plaintiff is entitled to royalties or profits from all the new markets to which defendant's derivative work appeals.  That issue turns on the very scope of the copyright which is here in question.[111]

    The key here lies in apportioning compensatory awards to the extent to which plaintiff's work contributes to the appeal of defendant's on the marketplace.[112]  To start, in terms of our spectrum, the court can ascertain the extent to which plaintiff's work is routinely copied, knowledgeably reworked, or innovatively recast in defendant's work.[113]  Suppose, for example, quite hypothetically, that no license had been obtained from Dashiell Hammett to film The Maltese Falcon, as John Huston did knowledgeably and to some extent innovatively.[114]  Not only would it not be equitable, but it would undermine incentives to make and release the film, if the writer were accorded all the film profits that would be otherwise attributable to talented stars, cinematic direction and production, promotion, etc.  But the court may award Hammett reasonable royalties or shares of profits for Huston's rather thorough-going use of his story, which partially accounts for the abiding appeal of the resulting classic film noir, even while the court factors out creative filming, etc., also contributing to that appeal.[115]  It is not only a matter of the quantity of the materials taken from plaintiff's work, but of the qualities of these materials that contribute to their market appeal, even after defendant reworks and recasts them.[116]

    Finally, there remains the issue of including court costs and attorney's fees in monetary awards.  For example, judges might well impose such litigation expenses on pirates selling rote copies without consent but excuse them for parties who sought to obtain licenses for transforming the works at issue.  Such options are not easily coordinated with our guidelines, for one thing because legal systems differ considerably in their approaches to litigation expenses.  Some might deal with them in their copyright statutes, and others in purely procedural laws.[117]

 

3. Proof: the Audience Test

 

    In the United States, juries may find facts, while other systems largely rely on judges for that task.  Of course, juries do not decide on injunctive and other equitable remedies that remain the judges' responsibilities.  If called in the United States, juries find whether infringement takes place and assess monetary awards if they do so find.[118]  Judge Jerome Frank left us with the following conventional wisdom on point:

The proper criterion on that issue is not an analytic or other comparison of the respective musical compositions as they appear on paper or in the judgment of trained musicians.  The plaintiff's legally protected interest is ... in the potential financial returns from his compositions which derive from the lay public's approbation of his efforts.  The question, therefore, is whether defendant took from plaintiff's works so much of what is pleasing to the ears of lay listeners, who comprise the audience for whom such popular music is composed ....[119]

    Perhaps, all triers of fact are tempted to act as test audiences when hearing infringement cases.  Judges in non-jury systems, given their discretion to choose the best methods for finding facts, may at times consciously apply audience tests in such cases.[120]  Nonetheless, whether applied by judges or juries, audience tests give rise to a pair of related problems: first, not all triers of fact are sure to emulate accurately the responses of the relevant audiences; second, it is not always obvious who should constitute the relevant audience.  To take an example of the first problem, imagine a case in which a poem is allegedly translated: to ascertain whether there is translation, much less how the appeal of the original poem finds its way into the translation, the trier of fact must itself be bilingual or hear testimony from an expert who knows the languages of both texts.  As to the second problem, it brings us back to the issue of copyright scope: the extent of the markets reserved to plaintiff, along with the markets open to defendant,  will depend on just this issue.  But collapsing all criteria of relevant audiences into some notion of undifferentiated audience response only obfuscates the issue.[121]

    An imaginary case, introduced above, dramatizes the problem of emulating audience response: a computer converts audio input into video output.[122]  The computer reprocesses plaintiff's musical work, its rhythms, melodies, etc., into what appears to be an abstract cinematic work, namely moving and colored patterns played out in real time on screen.  Since the algorithm running the conversion is assumed to be pre-fixed without regard to the work at issue, the defendant who mechanically uses it to reprocess that work engages in rote copying that should trigger a full panoply of remedies.[123]  Nonetheless, the average judge or jury might not hear, and in turn not see, much similarity in appeal between plaintiff's musical work and defendant's ostensible cinematic work, even though the algorithm maps one work onto the other, virtually point by point.  As a result, a test merely asking the trier of fact to find infringement on the basis of some lay listener's and ordinary observer's response, but without the benefit of any analytic or expert comparison, might not lead to any finding of infringement at all, much less damages.[124]  Return, for a moment, to the example of the allegedly translated poem: there a bilingual expert could explain to the trier of fact how linguistic rules do, or do not, account for moving from plaintiff's text in one language to defendant's in another.[125]  Similarly, in this imaginary case, an expert could focus the trier of fact on how the conversion algorithm generates, from plaintiff's musical work, the full fabric and potential appeal of defendant's ostensible cinematic work.[126]

    Consider a real case which illustrates the problem of selecting relevant audiences.  The artist Jeff Koons instructed an art studio to make a sculpture, in a number of copies, of a photograph of a couple holding a string of cute puppies.  The sculpture, Koons claimed, critiqued kitsch sentimentality typified in the photograph, but no defense of fair use was allowed on that basis, and infringement was found on summary judgment.[127]  But from whose perspective would Koons' sculpture so obviously constitute an actionable copy of the photograph, and whose markets would it usurp, if any, for purposes of assessing damages?  In the eyes of other artists, Koons' colored, three-dimensional sculpture slyly recontextualized, that is, in our terms, knowledgeably reworked, the black-and-white, two-dimensional photograph at issue.[128]  Still, the court, anticipating on the jury's emulation of an ordinary observer's response, felt compelled to make a "black-and-white" finding of infringement, thus setting the stage for sweeping remedies.[129]  However, as Robert Rotstein acutely points out, such an audience test unavoidably involves a host of tacit premises about the markets that plaintiff and defendant might be respectively addressing.[130]  In the Koons case, such premises seemed to blind the courts to differences between the photograph and the sculpture possibly significant for the art circles constituting one of these markets.[131]

    In the imaginary case of the audio-to-video conversion, expert analysis of the conversion algorithm could explain how plaintiff's musical work was reprocessed into defendant's ostensible cinematic work.[132]  In the Koons case, did the instructions for "copying" the photograph as a sculpture amount to such a fully determinative algorithm, or did they leave open enough creative choices to put infringement, or at least the measure of damages, into question?[133]  Even if "enough" must ultimately be construed in this context with reference to some audience, it remains to be seen which audience was relevant in the Koons case: plaintiff's popular audience or defendant's artistically sophisticated audience?  Plaintiff's popular audience was not relevant to the extent that no significant evidence indicated it to be lost as a market, but defendant’s sophisticated audience could be, according to the appellate court, a source for an apportioned award.[134]  Following the reasoning outlined above, it would then have made sense in the Koons case to ignore plaintiff's photograph to the extent that it did not account for the appeal of defendant’s sculpture on the art market.[135]  An expert might well testify that Koons could have appropriated any equally banal image for that purpose.  On that basis, only his audience would be relevantly plugged into the compensatory equation.[136]

 

V. A CAVEAT AND Conclusion

 

    We started with a hypothetical suit.  At issue were Hiroshige's wood-block prints.  Relief was sought against Van Gogh's studies in oils of these prints.  This case dramatizes the dilemma implicit in all cases where prior works form bases for later works.  On the one hand, denying relief would ignore copyright rationales altogether; on the other, overly stringent remedies could betray these rationales.  This dilemma has become all the more critical as copyright has been expanded, for example, with rights to control derivative works and with ever-longer terms.

    We have here reframed infringement analysis with an eye to resolving this dilemma.  But our framework is not definitive, but merely an attempt to pull together various strands of infringement analysis in diverse copyright systems.   Furthermore, all infringement analysis has to look to current culture for its premises concerning creative processes, aesthetic perceptions, and economic realities.  Indeed, as a creature of our culture, copyright law can only proceed critically from just such premises in the light of its own basic rationales.[137]  No doubt, future aesthetic and economic insights will help to reconceptualize infringement analysis in new and better forms.

    With that caveat in mind, a pair of points should be stressed in concluding.  On the level of theory, courts need better analytic tools than shifting sets of vague and disparate tests and criteria of protectability and infringement.  The attempt to bring underlying doctrines together into some coherent framework of analysis at least helps to debug corresponding tests and criteria and to optimize infringement findings and remedies in the cases.  On the level of practice, courts ought not content themselves with simply finding infringement vel non but, in hard cases, would do well to discern more finely how plaintiffs' works are reworked or recast into defendants' works.  In that light, it is submitted, they could grant remedies more consistently with copyright rationales.