Typefaces and the copyrights

magnus_gaarde's picture

I just found this article which you may or may not find interesting.



raph's picture

Here's my take. The law is clear that typefaces themselves cannot be copyrighted in the US. I have heard suggestions that some European countries differ. Bitmap images of fonts are absolutely not copyrightable, and representing a font as a data structure containing bezier control points is almost certainly not copyrightable. This state of affairs seems to be the desire of the typesetter industry. If one were cynical, one might suggest that they simply want to avoid having to pay any sorts of royalties to font designers.

The trick that font distributors have used to get around this quirk of copyright law is to treat the digital representation of fonts not as pure data, but as a computer program which, when executed, draws the actual glyphs.

For most popular font formats today, this legal fiction has some basis in reality. Metafont fonts and TrueType instructions unarguably have many of the characteristics of computer programs. I consider PostScript Type1 to be a bit cloudier. Originally, this font format was mostly treated as a (somewhat trivial) PostScript program and typically rendered by executing it in a PostScript interpreter. However, in practice the programming features are so restricted that it is arguably a data structure rather than a true programming language, and many Type1 renderers simply extract the charstrings and skip over any real PostScript code that happens to be embedded.

But, for legal purposes, good enough. Type1 fonts are programs, sure. And, fast-forwarding to today, the "features" in OpenType (even the PS outline variants) are arguably complex and sophisticated enough to be software rather than just data.

So how did Adobe and Emigre win? First, keep in mind that they settled, so there's still no real judicial decision. Second, from what I can see the "infringing" fonts were not new outlines of the same basic shapes, but were generated by extracting the Bezier outlines. This gets you into the murky waters of what constitutes a "derivative work".

Quick summary, then: typefaces are not copyrightable, but for all practical purposes large corporations with teams of highly-paid lawyers can pretend they are.

John Hudson's picture

...representing a font as a data structure containing bezier control points is almost certainly not copyrightable.

Raph, if you are going to hold forth on such matters, at least do your homework. Some of us have spent many long hours studying the decisions of the US Copyright Office and specific cases.

There was indeed a judicial decision in the Adobe, Emigre etc. vs. SSi case: a summary judgement on the basis of which SSi decided to settle since it was clear that Judge White had determined that the plaintiffs' arguments had merit. In the summary judgement, Judge White specifically determined that the placing of outline points was non-arbitrary and constituted an authorial act that was properly subject to copyright. He made this decision within the framework of the 1992 Copyright Office decision on the protectability of font software, noting that the interaction of outline definition and hinting to produce a specific result met the requirements of software copyright.

raph's picture

Raph, if you are going to hold forth on such matters, at least do your homework.

But isn't people spouting off about things they know little about the primary discoursive mode of this board? If not, maybe I've misunderstood.

Seriously, thanks for the clarification on the Adobe, Emigre vs. SSi case. Most of my info comes from this excellent review by Paul Schaffner. Now that I look more closely, that review is almost 10 years old, and while I don't believe the law has changed much since then, I can certainly see how more recent decisions such as SSi can change its interpretation.

Syndicate content Syndicate content