This article is from the Copyright Law FAQ, by Terry Carroll with numerous contributions by others.
Section 2.8, above, describes fair use in a nutshell. This follow-on
entry provides a more detailed description of the doctrine for those
interested in the nuts and bolts.
There are four factors used to decide whether a particular use of a
copyrighted work is a fair use:
(1) the purpose and character of the use, including whether
such use is of a commercial nature or is for nonprofit
(2) the nature of the copyrighted work;
(3) the amount and substantiality of the portion used in
relation to the copyrighted work as a whole; and
(4) the effect of the use upon the potential market for or
value of the copyrighted work.
17 U.S.C. 107.
The remainder of this answer discusses how each of these factors has been
(1) The purpose and character of the use: In considering the purpose and
character of the use, courts have looked to two characteristics of the
use: whether the use is commercial and, somewhat less frequently, whether
the use is a "productive" one.
If the copyrighted work is being used commercially, e.g., all or part of
a copyrighted drawing being used in a commercially published book on
drawing techniques, that's a strike against it being fair use. On the
other hand, if the same drawing were used in a non-profit school to teach
children to draw, then this factor would be in favor of finding a fair
use. Most situations are somewhere in between. That is, a use might not
be commercial, but it's not necessarily non-profit educational, either.
Note, though, that the statute does not command this "commercial/non-
profit educational" balance, and not all courts use it, at least not by
itself. Commercial use might be forgiven if the use is characterized as
a "productive" or "transformative" use, i.e., a use of the material that
interprets or otherwise adds value to the material taken from the
copyrighted work. See Consumers Union v. General Signal Corp., 724 F.2d
1044, 1047 (2nd Cir. 1983) (noting that the use of one of Consumer Report
magazine's reviews of a vacuum cleaner in an advertisement was a fair
use, in part because the purpose and character of the advertisement was
in part to educate consumers). The Supreme Court has noted that the
distinction between "productive" and "unproductive" uses is not wholly
determinative, but is helpful in balancing the interests. Sony Corp. v.
Universal City Studios, 464 U.S. 417, 451 n.40 (1983).
(2) The nature of the copyrighted work: If the work being used is one
that is factual or functional in nature, then that's a point in favor of
use of that work being a fair use. That's because copyright isn't
available for facts themselves, and the courts recognize that it's kind
of dumb to force someone with a newspaper clipping to completely rewrite
it to avoid infringement (besides, a paraphrase is still an infringement,
because it qualifies as creating a derivative work, even if it's not a
direct copy). If the work is a fictional or artistic one, though, taking
the work is taking much more than any underlying facts. A fictional or
artistic work is more expressive than a factual one, so the copyright
(which is designed to protect expression) is stronger. Even in factual
works, however, where the portion used includes subjective descriptions
whose power lies in the author's individualized expression, this factor
might go against a finding of fair use, if the use exceeds that necessary
to disseminate the facts. See Harper and Row v. Nation Enterprises, 471
U.S. 539 (1985) (finding no fair use for infringement of former U.S.
President Ford's memoirs despite its factual content).
Another point that's often examined in looking at the nature of the
copyrighted work is whether the work has been published. Courts will
generally consider a use of an unpublished work as more likely to
infringe than a similar use of a published work. Harper and Row v.
Nation Enterprises, 471 U.S. 539, 564 (1985). This is for two reasons.
First, the first publication is often the most valuable to the copyright
holder. Second, it affects the copyright holder's ability to choose not
to publish the work at all. See Salinger v. Random House, 811 F.2d 90,
97 (2nd Cir. 1987).
As with the first factor, while the "fact/fiction" balance and
"published/unpublished" balance are two of the most common, they are not
commanded by the statute, which only requires considering the "nature of
the copyrighted work." For example, in Sega v. Accolade, 977 F.2d 1510,
1525 (9th Cir., 1993), the Court of Appeals noted that the nature of most
computer programs distributed in object code is that the functional (and
therefore unprotected) elements cannot be discerned without disassembly.
This supported the court's opinion that, in certain limited instances,
disassembling of a competitor's product to find interface information
that cannot be obtained in any other way is a fair use of the work,
despite the fact that code disassembly necessarily involves making a copy
of the copyrighted program.
(3) The amount and substantiality of the portion used in relation to the
copyrighted work as a whole: This appears simpler than it really is. On
the face of it, it means that if you incorporate 95% of a copyrighted
work into another work, it's a lot less likely to be a fair use than if
you take only a small portion, say, 5%. And that's true. However,
assessing this factor is a bit more complex than that simple statement.
Even if only a small portion of the work is used, if that portion is
"qualitatively substantial," e.g., if the portion used is essentially the
heart of the work, that use will be deemed to have been "substantial,"
and could go against a finding of fair use. See Harper and Row v. Nation
Enterprises, 471 U.S. 539 (1985) (finding no fair use for infringement of
former U.S. President Ford's memoirs, where the portion used (which
described Ford's decision to pardon former President Nixon) included "the
most interesting and moving parts of the entire manuscript"), and Roy
Export Co. v. Columbia Broadcasting System, 503 F.Supp. 1137 (S.D.N.Y.
1980) (taking of 55 seconds out of 89-minute film deemed "qualitatively
To confuse matters further, some courts have (probably erroneously)
interpreted this factor by looking at what percentage of the work _using_
the material is composed of material from the copyrighted work, rather
than what percentage of the copyrighted work was used. See, e.g.,
Association of American Medical Colleges v. Mikaelian, 571 F.Supp. 144
(E.D. Pa, 1983), aff'd 734 F.2d 3 (3rd Cir., 1984), aff'd 734 F.2d 6 (3rd
Cir., 1984). While this interpretation is probably erroneous, it's worth
bearing in mind that, at least in one judge's courtroom in the Eastern
District of Pennsylvania, that's how the statute will be interpreted.
(4) The effect of the use upon the potential market for or value of the
copyrighted work: The U.S. Supreme Court has stated that this factor is
"undoubtedly the single most important element of fair use." Harper and
Row v. Nation Enterprises, 471 U.S. 539 (1985). The late Professor
Melville Nimmer, in his treatise on copyright law, paraphrased it, "Fair
use, when properly applied, is limited to copying by others which does
not materially impair the marketability of the work which is copied."
Nimmer on Copyright, section 1.10[D]. If the use impacts the market for
the work, the use is less likely to be held to be a fair use.
Note also that the weighing is of the impact on the potential market, not
on the actual market. For example, although Playboy magazine does not
distribute its pictures in machine-readable form, it may choose to do so
in the future. One might argue that digitizing a picture and posting it
on the net does not impact the current market for the magazine originals.
However, it impacts the potential (but currently non-existent) market for
machine-readable copies. Because there is an impact on the potential
market, an analysis of this factor in such a situation would not support
a finding of fair use.
If all this sounds like hopeless confusion, you're not too far off.
Often, whether a use is a fair use is a very subjective conclusion. In
the Harper and Row case cited above, for example, the Supreme Court was
split 6-3. In the famous "Betamax case," Sony v. Universal City Studios,
464 U.S. 417 (1984) (in which the Supreme Court found that off-air non-
archival videotaping of broadcast television was a fair use), the split
was 5-4. In both of these cases, the District Court ruled one way (no
fair use in Harper and Row, fair use in Sony) and was reversed by the
Court of Appeals, which was then itself reversed by the Supreme Court.
This goes to show that even well-educated jurists are capable of
disagreeing on the application of this doctrine.