Hypertext isn’t a word we use much in 2002. It sounds out of date: the archetypal old-aged hi-tech word that sounds as much simpleton as it sounds faux-futuristic. We don’t think in terms of ‘hypertext’ any more. We don’t worry about those insidious links lurking almost unnoticed amongst the text, links that could take us to the next page or to a web site with absolutely no discernable relationship to the text it’s linked to. The links that make the text ‘hyper’. The myriad uses to which the Web has been put have taken text, originally the only information carrier available online, and put it in the place where we find it in magazines and adverts: subservient, a worthy and rather dowdy adjunct to the much more attention-grabbing images which frame it on a thousand web pages. Where text is still the primary feaure of web pages, it rarely reaches a high typographic standard. It may stretch from edge to edge, be multicoloured, go beyond even the most frivolous printed document in its styling. So despite its innate capability to whisk us away, it can be presented as dowdy and shambolic. Decidedly low-tech. Why should this be?

The Web is a new medium, and it seems right that the new medium should have a new typography. But does it need to be different from the old (ink) typography, different in that it seems to be a cramped parody of what we’re used to seeing in ink? Refer to the debate about the ‘new typography’ of the 1920s if you want to see people with typographical training arguing over a revolutionary (or appalling) new way to work with type, but I don’t think you’ll find anyone defending the way hypertext looks today. It’s a hostage of fortune, the typographic capabilities of newer browsers mired in the need to accommodate much cruder and more idiosyncratic capabilities of the older browsers.

It’s important to remember that we don’t only use our faithful browsers to read: some people never do. We search, we buy, and some people can even configure or control their computers through their browser window. Nobody would ask a magazine layout to accomodate this, let alone demand it done rigorously. The browser window’s typography has to be very flexible. Modern browsers supporting Cascading Stylesheets (CSS) give us this capability.

The inheritance

To start with a clean slate is not going to lead immediately to attractive web typography, or even to serviceable typography. We know that it will be used for very diverse purposes, so we want to make sure the type can work hard for its living. Modern browsers give us the ability to position and scale, and to choose from a limited set of typefaces. The rest of our options must surely come from print typography, the medium which gave hypertext its literal and metaphorical inheritance.

Taking our cues from this source saves us having to make it clear what we’re up to: like the desktop metaphor of user-friendly operating systems, we use reference to a familiar thing (type in print) as a way to make sure that users know what kind of thing they’re working with: sure, it’s made of coloured blobs of light, but it looks a bit like a real floppy disk, so perhaps that’s what it represents in the two-dimensional computer universe. As far as typography is concerned, it’s a case of establishing which styles and sizes of text represent feature text, such as headlines and advertising copy, and which represent the meat of the writing. That way, we can do what’s done in print: use the type to silently convey those distinctions without the need for any reading at all.

Some pointers to a web type norm

So has this been done? Sure. There are really good examples of straightforward typography on the web. Of course, it’s not everywhere: it never will be. Slowly, though, the people who work with text are coming to recognise the sizes and type styles which are the most comfortable and accommodating when the peculiarities of designing for the web are taken into account.

No properties are guaranteed online ...

The ‘peculiarities’ of designing for the web come down to one factor: the exact size and shape of a browser window can’t be guaranteed. Users have different computers with screens of various sizes; they use different browsers on different operating systems, and these systems have different ideas of how large a given size of a given typeface should be. Therefore a short line of smallish type on system A can seem to be the longest practical line of rather large type on system B. A typographer working on the web must find the best compromise between the extremes.

The unpredictable variety in appearance of the page that is eventually displayed can be a crippling handicap to its use, or it can be merely a fact. There are some tricks which can (almost) be relied on. The use of contrast is the key: contrast in relative type size, between italic and roman, between bold and regular weight, between sans and serif type, and between light and dark shades. Of course, undisciplined use of these qualities leaves you exactly where you started: with an unintelligible mess of variety. But used intelligently, with reference to what we see in print, they provide many of the cues readers need.

Colour is ‘free’ online: it’s safe to assume that almost everybody who can discern colour will be able to see distinctions in colour, although not necessarily the colours you specify. Users of Lynx might feel put out when I offer roman and italic (let alone serif and sanserif) as viable ways to create distinction: I don’t think they would be correct in imagining that Lynx offers them the full potential of hypertext, though. And I write as a Lynx enthusiast.

...except type.

Many designers are bitterly disappointed to discover that they have a range of six typefaces (give or take half a dozen) at their disposal: a far cry from what XPress, Suitcase and a copy of the Creative Alliance CD-ROM can provide. But there’s always a worthwhile experience to be had working within the tight confines of a limited typecase, and of course there’s the fact that in book typography there are still only about half a dozen typefaces in the running. The faces we can rely on are distributed by Microsoft (and are thus practically universal). They were designed for reading from the screen, and they do a very good job: much superior to ‘book’ types when displayed on a monitor. So for the time being, we bury the hatchet and accept these faces. We learn how to make the most of limited means. It’s a moral tale.


To round up, I’d like to support my assertion that web typography (the formatting of hypertext) is now something designers are learning to master, using a few clear examples. So without ranting about the use of graphics to display type, or tables to lay out pages, or about page length, I’ll let you be the judge. These examples are not necessarily ‘mainstream', nor do they have to answer to the design demands of a high-traffic web portal, but they indicate that the expertise is available, ready to be tapped into.