As I wrote in yesterday's post, my book, Copyright's Paradox, explains that copyright serves both as an "engine of free expression" and silencer of free expression. Yesterday, I focused on how copyright still serves as an engine of free expression in the digital age, albeit an engine of more modest proportions than the Supreme Court's moniker suggests.

Today I look at the other side of the equation — how copyright law burdens speech. I will emphasize at the outset that the fact that copyright burdens some speech is NOT reason in and of itself to abolish copyright or to find that copyright law violates the First Amendment. Rather we must tailor copyright to minimize its speech burdens while still enabling copyright law to serve as engine of free expression. (I do argue in my book that copyright's duration and scope have expanded so much that, in its current dimensions and configuration, copyright burdens too much speech.)

Very basically, if your movie, song, graphic, book, or blog is protected by copyright, I can't copy from it in my own speech unless you give me permission (express or implied) or unless my copying falls within an exception to your exclusive rights, like fair use. So copyright law can effectively prevent me from speaking using the words, images, or sounds of my choosing.

If the government were to do that directly and because it didn't like my choice of locution, say by forbidding me from distributing copies of the Koran or the Communist Manifesto on a street corner, it almost certainly abridge my First Amendment right to free speech. As the Supreme Court has held repeatedly, the government abridges speech in such cases even if the speaker could convey his message using other words.

Yet, like copyright's role as engine of free expression, the question of when and how copyright law can truly be said to burden speech (let alone violate the First Amendment) is far more complex than might initially meet the eye. Perhaps the easiest cases -- and the closest to what we usually think of as censorship — are those in which the copyright owner withholds permission and sues or threatens to sue for infringement because he wants to suppress the speaker's message.

Here's an example given to me by a reader of this blog: During the 2004 presidential campaign, opponents of John Kerry sought to highlight what they viewed as Kerry's radical Left views by offering the public free PDF copies of the book "The New Soldier," which Kerry co-authored in 1971 together with the Vietnam Veterans Against the War. According to a post on Free Republic, the book's co-editor, a close friend of Kerry, used the notice-and-takedown procedures of the Digital Millennium Copyright Act to force Yahoo to shutdown the web site that offered the free copies as part of a concerted effort to suppress circulation of the book for political reasons.

But sometimes the copyright holder's reasons for suppressing the allegedly infringing speech are mixed or entirely commercial. The Margaret Mitchell Estate sought to enjoin publication of Alice Randall's racy sequel to Gone With the Wind from the viewpoint of a slave both because it objected to her message (perhaps) and because it wished to license only those sequels that furthered the approved image and economic value of the original work.

And often copyright holders are, in principle, quite willing to grant permission but insist on a license fee that exceeds the speaker's ability to pay. Speakers may be unable to afford various goods, ranging from a computer to a printing press, needed for effective speech. Yet we don't normally limit property rights in those to goods in order to give speakers an entitlement to own or use them. Why should copyright be viewed differently? Why, for example, should we support an interpretation of fair use that heavily favors noncommercial uses, like educational documentary films, when we don't require owners of top-of-the-line movie cameras to make them available to documentary film makers?

Many commentators favor treating copyright differently because copyright law creates an artificial scarcity in a good -- original expression -- that, but for copyright law, would be available at the average cost of production. That argument gets us part of the way, but ultimately, I think, the answer lies in distributive free speech policy. Fostering expressive diversity (in the sense of speech from diverse and antagonistic voices, not product differentiation) is a cornerstone of American information, telecommunications, and free speech policy. The balance between copyrights and fair use should be struck to promote that goal.

The same is true when we examine media markets as a whole. As I have detailed in a recent post on Balkinization, major media markets are highly concentrated, at levels that are deleterious both to competition and expressive diversity. Copyright is relevant to that untoward consolidation because incumbent media regularly use copyright as a vertical restraint to stifle competition from new media. New media, like today's YouTube, peer-to-peer file trading networks, Internet radio, and Google News and yesterday's then-new recording industry, radio, and cable television, often seek to establish themselves in part by distributing incumbent media industries' copyrighted works. But new media also are vehicles for greater expressive diversity, for breaking incumbent media's stranglehold on the market and public discourse. So to the extent copyright gives incumbents a proprietary veto over new media's use of copyrighted expression, copyright law tends to be an obstacle to expressive diversity and thus is rightly seen to impose a burden on speech.