Wording is important in good human-computer interaction design. Wording of commands, messages, status and other interface elements can make a big difference. So let’s think for a minute about the word “safe”.

Safe is good, right? Is there anything bad about “safe”?

Firefox has a “Safe Mode”. And it’s a really bad idea. Why, you ask? “Safe” sounds really reassuring. An army of marketing experts could hardly come up with a better word, you would think.

So what’s not to like?

Well, it’s very simple really. Safe may sound good, but the problem is in its implication of the opposite. When Firefox is not in Safe Mode, what mode is it in then? Unsafe Mode?

The IT section of the Guardian has a technology advice column, where readers can ask questions. There, the following letter appeared recently:

I’ve just installed Firefox and see that it has a “safe mode”. If I use unsafe mode, what effect will it have on Windows and Internet Explorer?

The real problem is that technology experts, and that pretty much by definition includes all developers of these systems, tend to smile about these examples, if they think about them at all. With a deep enough understanding of what Firefox Safe Mode does, the thought of “Unsafe Mode” does not even occur to you.

However, the question from this user is not particularly naïve or unusual. It is probably pretty representative of the general user population. The plain fact is that interface developers have to be aware that the level of understanding in the general population is not as deep as their own, and that their words project models and expectations.

And in that respect, using the term “Safe Mode” is a really bad idea, since it indeed suggests (implicitly) that the normal mode is “unsafe”.

Should the Firefox designers have known that? Maybe, maybe not. But if you look over the shoulder of some regular users you can usually find out pretty quickly.