MacStansbury

One of the most infuriating things about the New York Times website is how it gets so stingy with links. If there is a link, it will go either to itself, or another, high–authority website. For example, if the story is about Digg, they may or may not link to Digg, but they will surely link to other coverage from the New York Times in the story. This bugs me.

At first, I figured the reason it caused me consternation was that the editors at the Times were just a bunch of stuck–up sticky–beaks who didn’t want to spread the linky love. Then I came to realize that this very site is an excellent example of the New York Times policy in play. After many revisions (this being version 8), the prior five years of archives just disappeared. Any story linked from the New York Times website would then end up at an error page.

Thus, the reason for my new linking policy. Before I get to the details of it, I’ll need to explain what is meant by ‘deep linking,’ why big authority sites avoid it, and why what they do a good idea.

Deep linking

Years ago, it was common for a website to only link to the home page of any other cited source. A site’s citation wouldn’t link to Yahoo.com/subdomain/source.html — a site would just link to Yahoo.com. As more people without formal training in writing etiquette started creating content on the internet, multiple versions of the rules appeared. The first formality to fall in this new era of citation was the courtesy of deep linking.

To ‘deep link,’ you need to have a target page, such as the above example. Instead of linking to the home page of Yahoo.com, you bypass that to get to the pertinent page. This is the same as going to a virtual back door to people you’ve never met. For some people, this is an intrusion of privacy, while other welcome it as a way to encourage interchanges.

Encouraging or discouraging deep–linking is more of a philosophical exercise than anything else. It is now such a standard practice that people think that discouraging deep–linking is the problem. While that (again) is more of a philosophical argument, the editorial argument is rather clear.

Stingy linkers

Going back to the New York Times, while I’ve a strong, negative opinion of their editorial policy, I can understand the logic. First of all, it is the newspaper of record, and what they print sure better be right or about 5 bajillion other sources will gleefully step up to take their shots. This is what happens when you are the market leader.

Besides the accuracy issue, there’s the stability issue. When nytimes.com sends a link your way, you’ve got to be sure the linked server can take it. It’s just not acceptable to send people somewhere that’s…not there. While I’m pretty sure the home page of Google can take a few links from me, the same could not be said from there to here.

You’ve ensured the accuracy of the link, and the strength of the linked server, are you sure you have the right page? That is the question that goes through the linker’s mind. Not to over–think this, but there’s a good possibility the objects of links may move. There’s no guarantee the linked will redirect your link to the correct place, or even tell you the page moved. Considering the scale of publishing at a national daily paper, they would need a full–time employee (at least) to stave off dead links.

Nuke ‘n Pave

Much like the accuracy, stability, and location, what about when a page just…disappears? Since there’s nothing old to look at, how can you ascertain what the linker was linking to? This is the problem for anyone linking to other sites.

A perfect example of a when a website completely clears its old content and starts anew is this site. It wasn’t a hard decision to clear five years of content, as the only remaining qualm I would have would be that I’d loses the residual search engine traffic. This approach brought in one–time readers, yes. But those one–time readers generally leave after they get the one thing they were looking for — like locust.

There’s many explanations of this phenomena, from the ‘long tail,’ ‘sith traffic,’ to my locust example. Since most of the search engine data will be out–of–date, they will continue to send people to dead links. That’s the same for old archived stories, linking to nothing. If I were an editor from a high–authority source (such as Salon.com, for instance), I would be hesitant to link to something I’m not perfectly sure will exist in a year, or even five–year’s time.

So my linking policy is made up of a few, good ideas. First, the idea is only link to things as general as possible. If it’s to another site, the link goes as high as possible. That means that if I can send you to the home page, I will. If it takes any deeper, I’ll try to keep it as high as possible.

Thus, a link that would go to a page like othersite.com/journal/index.php/page-name/#more would be worded in such a way as to send you to either that other site’s home page, or to as general a page as possible. In this example, since I needed to specify an individual, deep–linked post, I would have use the link to othersite.com/journal/index.php/page-name/. This should be a rare case, however.

I’m aware that people tend to read blogs and websites because they want to know what other people think. That’s why I watch television, why I read blogs, and why I like a good argument. So making the editorial policy here to keep it as much about what I think is really the best way to not have to worry about where I link. Thus, the linking policy would be that I am relatively sure the site I link to will exist, those links will be rare, and most of the content should be about what I think rather than profuse linking to what other people say.