Copyright and Digital Typography

Copyright © 1995 Paul F. Schaffner
ALL RIGHTS RESERVED

It may be supposed that digital fonts pose for the librarian the same kinds of problems, and that they should be treated according to the same guidelines, as other software. It may also be supposed that librarians and academic users fall outside the category of wholesale software- and typeface- pirates at which the majority of litigation is likely to be directed. To a large extent, these suppositions are true. Nevertheless, type, because of both its ubiquity and its unusual legal status, raises some distinct and thorny issues. Consider the following scenarios, all of which have occurred in my own experience:

  1. You are writing for publication and need to prepare camera-ready copy. Commonly available digital fonts do not include the full set of characters that you need to use. May you purchase a professionally made font outline, rasterize a bitmapped soft font from it, load it into an editor, and add characters to it? Or add diacritics to, or otherwise modify, existing characters? May you merge characters from another commercial font into it?
  2. Suppose, legally or otherwise, you do all of these things. Then a colleague in your field learns of your modified font and asks for a copy. May you send her one? It seems unreasonable and uncivil not to. May you go further and post the font file for anonymous ftp, announcing its existence to the members of a listserve-based mailing list?
  3. A class project on which you are a consultant is engaged in transcribing some early modern English texts. They need to represent both on screen and in print some of the odd abbreviations that the texts contain: may you modify a common TrueType or Postscript font (e.g., Times Roman or Arial) to help them out? Does it make any difference if you are paid for your efforts?
  4. You have designed an unusual font and would like to distribute it free of charge, but would like to ensure that no altered versions are distributed under your name. Or would like to distribute it as shareware, and would like to ensure that no altered versions are distributed at all. Or would like to be paid for your font, and ensure that no one distributes it but you. Is there any protection available?
  5. You glance through an old Monotype sample book and spot a particularly attractive Irish typeface called Columcille. Fortunately, it has been digitized. Unfortunately, Monotype charges $250 for the digital version. You decide to recreate it by scanning the printed image, and creating a TrueType outline font from the scanned image. When that proves difficult, you borrow a friend's Monotype font CD and print out Monotype's digitized form of Columcille at very large point sizes, one letter per page, then create a TrueType font from the images. Still dissatisfied, you load the Monotype PostScript outline directly into Fontographer, then use Fontographer to generate a TrueType outline and re-generate a PostScript outline. Unhappy even with this, you simply copy the PostScipt outline from the CD. You keep the fonts that you have copied and created and return the disk to your friend. Do any of these methods of obtaining Columcille infringe Monotype's copyright? Would it make any difference if the desired typeface were ITC Stone?
  6. You wish to display texts on the Web that contain occasional Greek and Hebrew words, as well as including pronunciation information using a subset of the International Phonetic Alphabet (IPA). Your solution is to obtain commercial Greek, Hebrew, and IPA fonts, display them on your screen, capture the screen image of each letter in an individual graphic file, then use these letters as inline graphics within your HTML-tagged texts. May you do so without permission?
  7. Copyright Law and Type

    A good number of long-standing controversies on the scope of copyright entangle the copyright status of digital typography almost inextricably. The amount of protection that may be afforded to industrial designs; the distinction, if any, between useful and decorative articles; the separability of functional and aesthetic components of design; the kind of protection that may be afforded to software and to digital data; the status of common typeface names as trademarks, and the legitimacy of software licensing (and "shrinkwrap licenses" in particular) all involve themselves in the question of digital type. The last issue is a particularly vexing one, since software manufacturers have resorted to "licensing" their software precisely because they wish to evade the rights granted to purchasers (but not licensees) under copyright law. The question of software licensing is therefore quite distinct from questions of copyright and I will largely ignore it.

    Locating Typeface

    To determine the place of typeface designs among the works offered protection by copyright, or among those denied it, requires consideration of statute, legislative history, case law, and federal administrative rule and practice. Historically, these are far from independent sources. For example, language dealing with design was added to the Copyright Office regulations in the 1950s in an effort to implement the Supreme Court decision in the Mazer case (see below); this regulatory attempt to interpret the Court's interpretation of the 1909 Copyright Act itself became the basis for changes introduced by the 1976 Copyright Act; these changes, together with the House Judiciary Committe Report that commented on them, became the basis in the 1980s for new changes in the Copyright Office Regulations and practices; the Copyright Office rules have since then formed the basis for judicial decisions.

    To begin in the middle, in the language of the 1976 act: no one doubts that a typeface design meets the fundamental test of eligibility for copyright protection, that of being an original work of authorship fixed in a tangible mode of expression;N1 nor that it is in some sense a "pictorial, graphic, [or] sculptural work." The vital question, however, is whether it qualifies as such within the definitions supplied by the statute, or whether it is excluded by virtue of being a "useful article":

    The design of a useful article, as defined in this section, shall be considered a pictorial, graphic, or sculptural work only if, and only to the extent that, such design incorporates pictorial, graphic, or sculptural features that can be identified separately from, and are capable of existing independently of, the utilitarian aspects of the article.N2
    A "useful article," moreover, is
    an article having an intrinsic utilitarian function that is not merely to portray the appearance of the article or to convey information. An article that is normally part of a useful article is considered a "useful article."N3

    Two points about this language should be noted, both relevant to the standing of typeface designs under the act. The first is that it attempts to codify in statute a broad de facto definition of useful object (an object having "an intrinsic utilitarian function") long employed by the Register of Copyrights in defiance of the Copyright Office's own regulation (which defined a useful object as one the sole intrinsic function of which is its utility).N4 Typefaces, in so far as they serve the ordinary purposes of printed communication, would seem to qualify as useful articles by this broad definition. More significantly, the definition of a PSG work attempts to codify the separability test introduced by the Supreme Court in Mazer v. Stein,N5 when the court held that the sculptural design of a lamp base was separable from the lamp that it supported and hence copyrightable. As the House Judicary Committee Report accompanying the 1976 act declares in commentary on this clause:

    The Committee is seeking to draw as clear a line as possible between copyrightable works of applied art and uncopyrightable works of industrial design.... Although the shape of an industrial product may be aesthetically satisfying and valuable, the Committee's intention is not to offer it copyright protection under the bill. Unless the shape of an automobile, airplane, ladies' dress..., or any other...product contains some element that, physically or conceptually, can be identified as separable from the utilitarian aspects of that article, the design would not be copyrighted under the bill.... Even if the appearance of an article is determined by esthetic (as opposed to functional) considerations, only elements, if any, which can be identified separately from the useful article as such are copyrightable.N6

    Unfortunately, despite the legislative attempt to codify and clarify the definition of a copyrightable PSG work, judicial effort "to apply the 'separability' standard has not resulted in a clear or consistent pattern of decisions."N7 With regard to typefaces, the most important aspect of the case law on the separability criterion is the extent to which courts have accepted the "conceptual" (as opposed to the physical) separability of aesthetic design elements (mentioned, note, not in the Copyright Act but in the Committee Report) from functional elements as sufficient grounds to award copyright protection. Assuming that typefaces are "useful articles" insofar as they are legible representations of letters, are the slab serifs of an "Egyptian"-style typeface conceptually separable from the letters? If so, is this separability sufficient grounds to award copyright protection to the design? Terrence Carroll has observed that the form that the separability test has taken in some recent decisions opens the door to the copyrightability of at least some typeface designs.N8 In particular, the Second Circuit Court of Appeals ruled in Carol Barnhart v. Economy Cover Corporation that a conceptually separable expressive element was sufficient to establish copyright protection for that element:N9 many decorative typefaces, if not the usual run of text typefaces, certainly contain such expressive elements. The court went further in Brandir International v. Cascade Pacific Lumber, adapting the "Denicola"N10 test and allowing protection to such elements as "can be identified as reflecting the designer's artistic judgment exercised independently of functional influences."N11 Under this test, almost all typeface variations are potentially eligible for protection.

    Standing in the way of any such result are the long-standing resistance of the Copyright Office to registering typeface designs; the refusal of Congress to pass any of the dozens of design-protection bills that have been proposed, including some that explicitly protect typeface design;N12 and the apparently express desire of Congress to exclude typefaces from protection by the Copyright Act. The same passage of the Committee Report that discusses the separability test (and the protection afforded to architectural designs) goes on to declare:

    The Committee has considered, but chosen to defer, the possibility of protecting the design of typefaces. A "typeface" can be defined as a set of letter, numbers, or other symbolic characters, whose forms are related by repeating design elements consistently applied in a notational system and are intended to be embodied in articles whose intrinsic utilitarian function is for use in composing text or other cognizable combinations of characters. the Committee does not regard the design of typeface, as thus defined, to be a copyrightable "pictorial, graphic, or sculptural work" within the meaning of this bill and the application of the dividing line in section 101.
    Again, this statement appears in the report of a single House Committee, and does not have the force of statute. The corresponding Senate Judiciary Committee report, in many places parallel to the House report, omits this section; but it should be noted that the original Senate bill contained a Title II devoted to giving limited protection to industrial designs, and that this title was omitted in the bill as passed, partly because it might have had the tendency to offer protection to typeface designs.N13 Congress has, in fact, been consistent in opposing protection for typeface designs, and the legislative history cannot reasonably be construed otherwise.

    The single Circuit Court decision (and that, unusually, in the fourth circuit) explicitly to consider the copyright status of typefaces, Eltra Corp. v. Ringer,N14 was decided under the 1909 Copyright Act. Eltra Corporation sued in order to compel the Register of Copyrights to grant copyright registration to one of its typeface designs. The district court had dismissed the suit on the grounds that Congress, by failing to overrule the policies of the Copyright Office legislatively, had effectively accepted that Office's administrative interpretation of the 1909 Act.N15 The Appeals Court upheld the decision, on two rather different grounds: that typefaces failed the "separability" test; and that the legislative history of the 1976 act (at the time of the decision very recent history indeed), in particular the passage quoted above from the House Report, could be safely used as a guide to the interpretation of the 1909 Act. Though the principle is sound enough, based on the express intent of the 1976 Act to codify and clarify, not alter, the eligibility requirements of the 1909 Act,N16 and one widely employed elsewhere, its effect in this case is to make the House Report extremely influential, even retroactively, and to create an almost insuperable presumption against the copyrightability of typefaces. In February of 1992, this presumption was reinforced by the Copyright Office's Regulation 202.1(e),N17 which lists as "Material not subject to copyright," "Typeface as typeface" (Further on this regulation below). Under the so-called "Chevron doctrine," named after a case in which the Court held that any reasonable agency ruling ought to be upheld if it represented a reasonable interpretation of the statute being administered, the Copyright Office's regulation has exercised all the power of case law precedent or statutory enactment, and then some.

    One potential shortcut to typeface copyright is to argue that typefaces designs are not designs for useful articles at all, thus obviating all the "separability" arguments based on industrial design. Carroll, for example, suggests that the useful item in question is not a typeface but a font, and that the relationship of font to typeface design is not that of article to design (like that of house to blueprint) but that of medium to art work (like that of book cover to photograph reproduced thereon). A font is something that allows a design to be seen; a design is not something that enables a font to exist in a particular form. An alternative argument, on the same lines, that seems plausible, though it is not one that I have seen made, is to argue that the utilitarian purpose of letterforms is precisely "to convey information"--the one function that in section 101 excludes an item from being considered a "useful article." I have not been able to ascertain what sort of article this clause was intended to cover, or why it should not be thought to cover letterforms--instruments of communication if anything is. While the current position of the Copyright Office obtains, of course, all such arguments remain moot.

    Locating Digital Type

    To the perplexing problems posed by consideration of typefaces as useful articles, and of typographic style as conceptually separable from typographic functionality yet additional complications have been added by the development and widespread use of digital type. Could a computer program, normally accredited as a copyrightable work of literary authorship, continue to be regarded as such when its whole purpose was to generate unprotectible typefaces? On the other hand, if a computer font could be protected, why not also the typeface that it embodied?

    The most important rulings bearing on these questions have been administrative rather than judicial or legislative. In 1986, the Copyright Office published a Notice of Inquiry in response to "applications to register claims to copyright in digitized information or material that represents typeface designs and is used to print texts," seeking "public comment about the copyrightability of digitized typeface apart from the uncopyrightable typeface design."N18 After receipt of considerable public comment, the "Policy Decision on Copyrightability of Digitized Typefaces," was issued two years later. Though it summarizes and discusses an interesting variety of opinions on the subject, its conclusions and policy recommendations are based heavily on the House Report and the Eltra case:

    Although the master computer program used to control the generic digitization process is protectible and may be registered, if original, this protection does not extend to the data fixing or depicting a particular typeface or typefont or to any algorithms created as an alternative means of fixing the data. The Office will register a program that can be used to create digitized versions of various typefaces but will not register the data used to depict a particular typeface or individual letterforms. If the computer program submitted for registration includes data that fixes or depicts a particular typeface, typefont, or letterform, the Office requires an appropriate disclaimer of copyright on the application to exclude uncopyrightable data.N19
    The effect of this ruling was, at least on the face of it, to disallow copyright registration for digital fonts, however they might be implemented. It explicitly disallowed registration of bitmapped fonts, point-defined scalable fonts, and Bezier-curve-based outline fonts, as well as any comparable method of representing typographic data. Only generic font-generation programs (like Bitstream's "Fontware") could be registered--not the font outlines (like Bitstream's .BCO-format fonts) that they employed.

    Though the 1988 ruling directed type designers to Congress as the only source of remedy for those seeking protection, in fact the Copyright Office itself reconsidered its ruling only three years later, "concerned that [recent] claims represented a significant technological advance."N20 It is not entirely clear, either from the notice of enquiry or from the comments received, just what the technological advances consisted of: hinted font outlines had been explicitly mentioned, and rejected as uncopyrightable, in the 1988 ruling. The one advance mentioned, the creation of scalable fonts from already digitized typeface data as opposed to the digitization of pictorial data, does not seem sufficient to justify changes. Nevertheless, the 1992 ruling that emerged from the enquiry process reversed to a significant degree the policies laid down by the 1988 decision:N21

    The Copyright Office is persuaded that creating scalable typefonts using already-digitized typeface represents a significant change in the industry since our previous Policy Decision. We are also persuaded that computer programs designed for generating typeface in conjunction with low-resolution and other printing devices may involve original computer instructions entitled to protection under the Copyright Act. For example, the creation of scalable font output programs to produce harmonious fonts consisting of hundreds of characters typically involves many decisions in drafting the instructions that drive the printer. The expression of these decisions is neither limited by the unprotectible shape of the letters nor functionally mandated. This expression, assuming it meets the usual standards of authorship, is thus registrable as a computer program.
    The Copyright Office further emended its practices so as no longer to require a disclaimer of copyright in typographic elements contained in software programs. These changes resulted in the amended regulation 37 CFR 202.1(e) (forbidding the registration of "Typeface as typeface"--implying that typeface as software could be registered). The change of practice with regard to the disclaimer was regarded as just that, not a change of regulation, and was recorded as such; manufacturers whose programs had already been registered with the office would not be allowed to re-register with the intention of removing the disclaimer requirement. And the change of policy with regard to registering font-generating software was described as a clarification. Indeed, it appears that it reflected and did not create changes in practice, for the Copyright Office had accepted typeface outlines for registration as early as 1990, when Adobe's Garamond Light was granted copyright registration number TX 396-004.N22 Adobe had, as contemporary practice dictated, specifically disclaimed all copyright in the "underlying typeface characters" and the overall design, but successfully registered the font anyway.

    The Present Situation in a Nutshell

    Absent any case law on the subject, the following seems to represent a conservative view of the present situation in the United States:

    Tentative Answers to the Six Scenarios

    Counterarguments and clarifications are welcomed.

    1. "May you purchase a professionally made font outline, rasterize a bitmapped soft font from it, load it into an editor, and add characters to it? Or add diacritics to, or otherwise modify, existing characters? May you merge characters from another commercial font into it?"
    2. "May you send [a copy to a colleague]? May you go further and post the font file for anonymous ftp, announcing its existence to the members of a listserve-based mailing list?"
    3. "A class...is engaged in transcribing some...texts."
      1. "May you modify a common TrueType or Postscript font (e.g., Times Roman or Arial) to help..out [a class project]?"
        • The answer would seem to depend on the copyright status of such widely distributed fonts as TrueType Times Roman (distributed with every copy of MS Windows and every MacO/S). Such fonts have been described as "effectively in the public domain," but how effectively is that? I do not know the answer. Nor do I know if a good fair use case could be made on educational grounds.
      2. "Does it make any difference if you are paid for your efforts?"
        • No.
    4. "You have designed an unusual font.... Is there any protection available?"
    5. "You spot...Columcille...."
      1. "...Do any of these methods of obtaining Columcille infringe Monotype's copyright?"
        • Creating an outline from a scanned image of a book printed in a font does not infringe, assuming that the scan itself qualifies as fair use.
        • Creating an outline from a scanned image of a laser print of a font does not infringe.
        • Generating a new font by passing an existing font through a font editor's proprietary format is a dubious case. Fair use would not seem to apply, since your intent is to deprive the copyright-holder of a sale. It certainly appears as though this procedure infringes the right to create a derivative work, and perhaps also the right to reproduce the work. Though the new font would certainly lack the hinting (and therefore the quality) of the original, I believe that it could be argued persuasively that the making of a copy is not protected simply because it is a bad copy, any more than a streaky photocopy thereby becomes a legal one. The critical case bearing on this question, the long-standing Bitstream, Adobe, et al. v. SWFTE case,N29 has been expected to clarify the legality of such practices, some variations of which have been regularly employed by SWFTE to generate its library of cloned fonts.N30 Unfortunately, the case has now been settled out of court, apparently at Swfte's expense, but without creating judicial precedent.N31
        • Copying the font itself is clearly an infringement.
      2. "Would it make any difference if the desired typeface were ITC Stone instead of Columcille?
        • Perhaps. ITC Stone is among several faces (including also Adobe Garamond and Adobe Minion)N32 which are protected not only by copyright but by design patents. The protection offered by patent is considerably greater than that offered by copyright.
    6. You wish to...capture the screen image of each letter...."

    Neglected Issues

    In the interests of space, of I have omitted consideration of several issues that influence or seem likely to influence the protection granted to digital typography. International practice, increasingly codified by multilateral treaties, increasingly grants protection to digital type. Interactive type creation, of the sort pioneered by Adobe's Multiple Master technology, will increasingly leave the authorship and fixedness of type designs in question. And, most influential of all, the institutional environment in which digital type is created, organized chiefly around elaborate licensing agreements (in default of copyright protection), will continue to dictate for the forseeable future the terms under which high-quality professional typefaces can be obtained and used.


    Notes

    1. 17 USC 102(a).RETURN
    2. 17 USC 101.RETURN
    3. Ibid.RETURN
    4. "If the sole intrinsic function of an article is its utility, the fact that the article is unique and attractively shaped will not qualify it as a work of art. However, if the shape of a utilitarian article incorporates features, such as artistic sculpture, carving, or pictorial representation, which can be identified separately and are capable of existing independently as a work of art, such features will be eligible for registration." 37 CFR 202.10(c), adopted in response to Mazer v. Stein; revoked 1 Jan. 1978.RETURN
    5. 347 U.S. 201 (1954).RETURN
    6. House Report 94-1476 (Report of the Judiciary Committee), 3 September 1976, CR 122 (1976). Reprinted in United States Code, Congressional and Administrative News, 94th Congress, 2nd Session, 1976 (St. Paul: West), 5:5668-5669.RETURN
    7. Robert A. Gorman, Copyright Law (1991), 31. On the general history of the separability criterion, see Erica Lehrer, "The Design of Design Law Today," Copyright Law Symposium Number Thirty-Five... (NY: Columbia University Press, 1988), 145-181.RETURN
    8. Terrence J. Carroll, "Protection for Typeface Designs: A Copyright Proposal," Santa Clara Computer and High Technology Law Journal 10 (1994): 139-194, especially 151-4.RETURN
    9. 773 F.2d 411 (2d Cir. 1985).RETURN
    10. See Robert C. Denicola, "Applied Art and Industrial Design: A Suggested Approach to Copyright in Useful Articles," Minnesota Law Review 67.4 (April 1983): 707-48.RETURN
    11. 834 F.2d 1142 (2d Cir. 1987) at 1145; cited by Carroll, 153.RETURN
    12. E.g., in 1986, S.791; in 1990, HR 902 and HR 3017; see "Copyright Office and PTO Back Design-Protection Legislation," BNA Patent, Trademark, & Copyright Law Daily, Oct. 10, 1990 [via LEXIS].RETURN
    13. Reasons for this resistance to protecting typefaces included fear of the possibility that owners of rights in typefaces would "attempt to enforce tying arrangements with typesetting equipment also owned by them" and fear of the possibility of injunctions against printers and publishers, with a consequent restriction on the freedom of the press: Lehrer, 174, n124. For a fuller discussion of political and practical objections to copyrighting typeface designs, see Carroll, 172-182.RETURN
    14. 579 F.2d 294 (4th Cir. 1978), espec. at 297-8.RETURN
    15. Eltra v. Ringer, 194 USPQ (BNA) 198, 201, cited by Carroll, 155.RETURN
    16. House Report, p. 5664 and 17 USC 113(b), both cited by Carroll, 158-9.RETURN
    17. 57 FR 6201 (1992) at 6201-2 = 37 CFR Ch.11, 202.1(e).RETURN
    18. 51 FR 36410 (10 October 1986) [via LEXIS].RETURN
    19. 53 FR 38110 (29 Sept. 1988) [via LEXIS].RETURN
    20. 56 FR 42073 (16 August 1991), "Notice of Public Hearing: Reconsideration of 1988 Policy Decision on Copyrightability of Digitized Typefaces."RETURN
    21. 57 FR 6201 (21 February 1992), "Registrability of Computer Programs that Generate Typefaces" [via LEXIS].RETURN
    22. "Adobe Successfully Registers Copyright Claim for Font Program," Current Developments, The Computer Lawyer 7.2 (February, 1990): 26.RETURN
    23. Raymond T. Nimmer, The Law of Computer Technology: Rights, Licenses, Liabilities, 2d ed. (Boston: Warren, Gorham, Lamont, 1995), 1/105 (Sect.1.18[1][b]), citing Foresight Resources v. Pfortmiller, 719 F. Supp. 1006 (D. Kan. 1989) and Vault Corp. v. Quaid Software, Ltd., 847 F2d 255 (5th Cir. 1988).RETURN
    24. R.Nimmer, 1/103 (Sect.1.18[1][a]).RETURN
    25. David Bonser, "Preemption of 'Shrink-Wrap' Legislation by the Copyright Act," 127-172.RETURN
    26. R.Nimmer, 5/40 (Sect.5.10[4]), noting that Australia, Great Britain, and Singapore all allow waiver of this right by express contractual arrangement.RETURN
    27. Stephen Fishman, Copyright Your Software (Berkeley: Nolo Press, 1994), 8/2.RETURN
    28. R.Nimmer, 5/41 (Sect.5.10[5]).RETURN
    29. Civil Action No. 93-11068H, US District Court for the District of Massachusetts.RETURN
    30. William M. Bulkeley, "Font War: That's My Type," Wall Street Journal, Eastern Edition, 19 Nov. 1993, p.B1.RETURN
    31. Adobe Systems, Inc., Press Release, 18 Sept. 1995, posted to news:comp.fonts.RETURN
    32. Fishman, 7/15.RETURN
    33. 53 FR 21817 (10 June 1988), cited by Curtis E.A. Karnow, "Data Morphing, Ownership, Copyright, and Creation," Leonardo 27.2 (1994): 122.RETURN

    (c) bird Click on the © symbol to return to the ILS604 class page.
    Click on the bird to return to the Index of Schaffner's class essays.

    Paul Schaffner / pfs@umich.edu / 11 October 1995; rev.4.11.95