Copyright Notice on Websites: Too Little, Too Much

Copyright under United States law is often misapplied to websites. Notices are often inadequate. Notices that are adequate often overclaim. Both issues afflict most websites I’ve seen. I’m not a lawyer, but I copyright my websites, and I see issues. Here’s what I do. It likely fits your websites.

Why At All

If a work is protected by copyright but someone copies it without permission, they are infringing, and that can go into a lawsuit, which can lead to money. Notice helps protect. The law does not require notice; the copyright on the work exists even without it. But if someone infringes the copyright and the copyright owner is entitled to compensation, a court may award less where a copyright notice was absent. The lack of notice gives the infringer the right to defend on the ground of innocent infringement, which roughly means that the legal violation occurred but wasn’t as bad, since notice was lacking. So, you likely want to provide notice.

Basic Notice

The notice must be legally adequate. The Federal statute says this: “The notice shall be affixed to the copies in such manner and location as to give reasonable notice of the claim of copyright.” In other words, it must be placed where it is calculated to inform someone who otherwise might innocently infringe the copyright. In a book, that’s usually in the front.

Websites are different. While you can pick up a book and open it in the middle, at least you had the front of the book in your hand when you skipped past it. But search engines have been linking deep into websites, we follow links directly to the content we want, viewing the home page is relatively rare, and many users wouldn’t even know how. That would explain why many websites now put copyright notices on every page, not just on the home page. One stand is that one notice on the home page suffices for the whole website, but I’m more cautious.

Notice in More Places

I go beyond that.

Notice Specific to Source Code

Count me as overcautious, but I think more notices are legally necessary. I add three more kinds of notice.

Here’s why. Every page of a website is fetched and delivered to your browser as source code. That’s how programmers write it. (In the HTML world, they’re called authors, but they’re programmers.) Your browser interprets what comes in, but you can see the raw form, and that’s the source code. Most browsers now let you examine the source code. (You might work for an institution that doesn’t allow you to use that command, but you likely can do this at home.) There’s usually a menu command, something like “View Source” or “Page Source”, that lets you see the source code without interpretation. Many programmers work with that.

But notice on a page as viewed by a user in a browser might not apply to what cannot be viewed in the way most users look. (Possibly courts have already settled the point and favored applicability, but I haven’t read that.) The source code is itself a copyrightable work. You may need notice where someone examining the source code per se will see it, so they can’t claim innocence if they infringe.

This leads to an issue of legibility. The copyright notice for public display is written in a programmer’s form. That could make it harder for other programmers to read, and much harder for nonprogrammers. That’s because programmers write Web pages with lots of markup, various tags that go in places people who don’t program wouldn’t expect. In a copyright notice, the year and the name might get tags that an untrained person might not know how to read. Not every HTML programmer knows all of the HTML elements, since there are so many. The result is that the general public and even a programmer could easily miss the notice.

With the markup mixed in, the notice might not fit the legal format for a notice, and so a court might accept a defense that an infringement was innocent. An example: Copyright: Revisions: Copyright <time datetime="2020">2020</time> Nick Levinson. Balance: Copyright <time datetime="2017">2017</time>–<time datetime="2019">2019</time> Nick Levinson. All rights reserved except as otherwise provided.

In this case, a notice that was 78 characters long became 1,286 characters long, 16 times as long. The legal part of the notice was buried in tags and hard to read. That means that programmers could reasonably see source code without being aware of a copyright notice on that page. Then, they might copy from it and might have the defense of innocent infringement. Especially so, if a judge doesn’t see it.

Solution For Source

Part of my solution is to write a plain form of the copyright notice in the source code, in a way compatible with the purpose of source code but humanly-readable by anyone seeing the source code itself with their bare eyes. How to do that depends on the computer language used for writing the source code, but it’s likely that almost every computer language has a comment facility. It is a way of writing a human-to-human comment so that it will be ignored by the program that has to interpret, compile, or translate the source code into object code. The comment is meant to be read by humans with or without programming skill.

The years in the copyright notice for source code may be more or fewer than those for renderable content alone, and they may be the same. The renderable content is often in the source code but often replacement content or a script generates renderable content not in the source code, so years for a copyright notice may differ.

Here’s an example for the computer language most often used for writing Web pages, which is HTML: . Your browser won’t get confused by that. Your browser will skip past it, but a human will have an easier time understanding it.

Some website hosters, who put websites up onto the Internet so other people can see them, strip those comments out. The process is minification and it speeds up page loading, which users and Google like. But removing the comment means that the copyright notice is not seen by anyone. The traditional notice is left intact, but not the source code comment. One possible solution is to mark the comment as permanent, but HTML does not have an agreed way of doing that. I try to indicate it in HTML by adding an exclamation point after the opening string (“<!--! ” instead of “<!-- ”) because some parsers recognize that, but it’s not generally agreed to. Another solution, which I also use, is not to minify. If your website hosting service minifies by default, turn minification off before uploading and check they complied; or change hosting providers.

You’d still write the notice in the traditional way, as well as in a comment. The traditional way is how your browser will display it for users who don’t want the raw form. Almost no one wants to see the raw form. But, in the raw form, the traditional way is buried in tags. So, you also write it in a comment, so anyone seeing the source code will see it.

Finding Where It Is

To make sure they see it, put a pointer near the top of the source code.

In HTML, the file must begin with a declaration telling the browser how to interpret the file. Nothing can come before that, or the browser won’t know how to read it. Anything after the declaration on the same line could confuse human readers. So, I insert a second line. On that line, my source code says this:

Writing For Inhuman Robots

I go farther still. I talk to bots.

Websites are read by humans, but not only by humans. Parsers read them, too, ultimately on behalf of humans but in the meantime for the parsers to use. Bots are parsers. Bots swarm all over the place, in large numbers. Some are helpful. Parsers, therefore, are part of Internet life. And there are many kinds of parsers.

Some parsers apparently check the copyright status of a work. This can be surmised because there are special codes for telling parsers about the copyright status of a Web page, and one would hardly waste time creating the codes without a demand through parsers to have the codes. Parsers cannot be counted on to reliably understand a copyright notice written for humans, either as rendered in a browser window or in an HTML comment in the source code, because none of them have enough consistency, so our only hope is in a computer language parsers will parse. The most widely used language for the Web being HTML, and HTML having those codes, the need for parsers and the humans behind them to be bound by copyright may require that we who program or create the Web pages include the copyright notice in a form that parsers will recognize. What form that will be depends on the computer language in which that Web page is written. Most often, that’s HTML. If a given page is in HTML, we need to use HTML meta tags that signify the copyright status to parsers. I use them. Those are the rights, dcterms.rights, dc.dateCopyrighted, dcterms.dateCopyrighted, and dcterms.rightsHolder metatags. In addition to the metatags, I use Microdata to identify the legal rights holder, initially the copyright owner and maybe a licensee: . . .. Using all those tags and the Microdata may seem redundant of each other, but it’s not. They’re all needed.

Overclaims Damage Your Legal Position

The other copyright notice problem is in claiming too much.

It’s convenient for programmers to overclaim. And it probably won’t cause a legal problem for many years to come. This, in practical terms for almost all websites, means never. There could be a legal credibility issue affecting some other legal right, but I don’t know if that would arise.

The overclaim is to assert copyright as of the latest year only, as if everything is new. Often, the text content, the graphic design, the layout, the style, and what copyright lawyers call enhancement and arrangement have been developed and under copyright for years. So, the copyright notice has to list the years when various portions first came under copyright protection.

For example, in some cases this would be overclaiming:

Instead, this, depending on the years, could be accurate:

Renewal dates are optional. They’re not needed.

If a portion of your source code comes from someone else and is under their copyright, it likely needs a separate copyright notice, probably at the beginning of that portion of the source code or somewhere before it, somewhere calculated to be seen by someone viewing that portion. For this, you can write, “Portions: © 2020 Pat Doe”. If you can clearly specify which portion is covered by that notice, you should. For example, if all of the JavaScript is by someone else and it is clear which parts of the page are in JavaScript, you could write, “JavaScript: © 2020 Pat Doe”. If only some of the JavaScript is to be covered by the notice, you could write, “Portions of JavaScript: © 2020 Pat Doe”.

If your website is reproducing offline work, like a book, and you've added original portions, such as a new arrangement, it’s likely acceptable to assert all relevant years for copyright, because the new online content includes enough originality by you to qualify. If the old work is in the public domain, you likely can still assert copyright on your online arrangement; if so, that’s not an overclaim. If you provide a new layout for plays by Shakespeare, scores by Mozart, or paintings by Rembrandt, the work by Shakespeare, Mozart, or Rembrandt is in the public domain because of its age, but your new layout of their work is under your copyright, so you can add a notice for the year you made the different layout. In that case, you might write, “Portions: © 2020 Chris Doe”.

Anything Else?

Licenses, often including copyleft, open-source licenses, dedications of otherwise-copyrightable works to the public domain, individually-negotiated licenses, music licenses from ASCAP, BMI, and SESAC, and other means by which someone may copy, distribute, use, modify, and/or do other activities protected by copyright, whether with substantial, little, or no limitation, are premised on copyright pre-existing any such license. Therefore, the copyright notice must still be provided. Typically, according to the terms of the license, the license must also be cited, but the license citation would generally come after the copyright notice. Generally, each license states what is required for compliance with its terms. Modifying the work or creating a derivative work generally requires continuing compliance with the terms of the license, depending on the terms of the license.

State copyright law may apply, but websites generally appear to qualify as works in a fixed medium and therefore Federal law pre-empts State law on point, and so owners of most websites need not worry about complying with State law on point.

Patent, trademark, trade secret, and other intellectual property rights are separate from copyright. Copyright may or may not apply to a work subject to other intellectual property rights. An example is that application software may be subject to patent while its look and feel may be subject to copyright. If both copyright and other intellectual property rights apply to the work, the copyright notice is required regardless of what is required for other intellectual property rights. Otherwise, other intellectual property rights are beyond the scope of this article.

Sources

(Sources: The URLs from the links at “innocent infringement” and “separate legal background” are as accessed Apr. 21, 2020. The statutory language on notice is from <https://www.copyright.gov/title17/92chap4.html>, as accessed Jan. 11, 2019 (17 U.S.C. 401(c)). The URL from the link at “one notice on the home page suffices for the whole website” is as accessed Apr. 21, 2020. The URLs from the links at “fetched”, “source code”, “called authors” (author and programmer used interchangeably), “Most browsers now let you examine the source code.”, “source code is itself a copyrightable work”, “harder for other programmers to read”, “much harder for nonprogrammers”, “likely that almost every computer language has a comment facility”, and “minification” are as accessed Apr. 22, 2020. The URL from the link at “HTML does not have an agreed way of doing that” is as accessed Jan. 12, 2019. The URL from the link at “Bots swarm all over the place, in large numbers.” is as accessed Jan. 11, 2019. The URLs from the links at “Some are helpful.”, “meta tags”, “rights, dcterms.rights, dc.dateCopyrighted, dcterms.dateCopyrighted, and dcterms.rightsHolder metatags”, “Microdata”, “overclaim” (the document at the URL is on copyright issues other than the initial copyright year), “new arrangement” (whether this is the latest law on point is unknown), “copyleft”, “open-source licenses”, “dedications of otherwise-copyrightable works to the public domain”, “ASCAP”, “BMI”, “SESAC”, “State copyright law may apply”, “Patent”, “trademark”, “trade secret”, “intellectual property rights”, and “look and feel”, are as accessed Apr. 22, 2020.)