[rohrpost] Hypertext pre1.4
Henning Ziegler
henning@fotomat.org
Tue, 5 Nov 2002 23:26:13 +0100
(...der vierte Teil meines Hypermedia-Aufsatzes...)
4 Interface Politics
The Macintosh interface is designed to provide a computer environment
that is understandable, familiar, and predictable.
-- Apple Computers, Inc. - Macintosh Human Interface Guidelines
Computer programs and hypermedia objects today typically 'run' on a
computer workstation that consists of the computer, a monitor, a mouse,
and a keyboard. The visual window into digital data is the human
computer interface (HCI, in most instances simply a computer monitor
screen), so the formal aspects of interface design should substantiate
my thesis that the interface is a site where absent cultural and social
contradictions clash and meaning is being dialogically produced for a
cultural community, in that the interface makes an unstable, messy, or
liquid data world coherent to the user (and the programmer). In what
follows, to set the general backdrop for my analysis of hypermedia in
the next section, I'll try to substantiate this thesis by looking at the
interface designs of two prominent operating systems (Mac OS and Windows
XP), along with the computer programs AOL 8.0/Internet Explorer and
Mozilla/Netscape 7 that are often used to access hypermedia.
Most graphical computer environments of today (such as Windows XP,
Gnome, or KDE) derive their interface design from the first Macintosh
computer that Apple introduced in January 1984. Arguably, we're now
moving toward the invisible computer that is integrated, for instance,
into human clothing, but Apple's initial desktop metaphor still serves
as the most widely used interface, since Apple's idea of a relatively
inexpensive personal computer was apparently years ahead of the computer
development at the time of its design (expensive multi-user systems
prevailed that were securely stored away in some computer lab on a
university campus). Interestingly, although the company had the
possibilities of designing a product almost entirely from scratch,
Apple's strategy turned out to be quite 'conservative.' The company
puts its strategy this way: "The 80 percent solution means that your
design meets the needs of at least 80 percent of your users. If you try
to design for the 20 percent of your target audience who are power
users, your design will not be usable by the majority of users" (Apple
Inc., 1992, p. 35). Out of marketing considerations, then, Apple
apparently reduced the relative complexity of text-based computer
interfaces from the start to familiar, graphical metaphors of everyday
U.S. life, thus heeding the warning of Ben Shneiderman's /Designing the
User Interface: "Surprising system actions, tedious sequences of data
entries, inability or difficulty in obtaining necessary information, and
inability to produce the action desired all build anxiety and
dissatisfaction" (Shneiderman 75). Generally, the Apple solution became
'point and click' rather than type a command, and to design and
communicate an ease and simplicity in the use of a Macintosh computer
has ever since the design of the first Mac been Apple's major
promotional strategy. In June 2002, for instance, Apple launched the
famous switcher campaign (ads that feature 'real life people' who
switched from PC to Mac) that again makes use of such terminology:
"More people are interested in switching from PCs to Macs than ever
before. See why they made the change and how easy it was. (...) And
understand how Macs can make your life easier and your possibilities
endless" (http://www.apple.com/switch). In its official design handbook
from 1992, the /Apple Human Interface Guidelines, Apple developed eleven
principles of interface design by which the company tried to incorporate
ease and simplicity into its products: Metaphors, direct manipulation,
see-and-point, consistency, WYSIWYG, user control, feedback and dialog,
forgiveness, perceived stability, modelessness, and aesthetic integrity.
The Apple Macintosh interface thus ideally presents the user throughout
with nicely designed, familiar metaphors (such as the trash can or the
desktop itself) which one can interact with in close-to-real time by
using a point-and-click device while immediately seeing results or
getting a failure feedback and a possibility to 'undo' the action.
Since Microsoft, having copied the Mac interface and its principles
almost one to one, dominates the personal computer market with its
Windows operating system, the socio-political purpose of virtually every
user interface today can be said to be to "create safety nets for
people" (10).
>From the principles in the official Apple design handbook, I think it's
already clear that, at its inception in 1984, the personal computer as
we know it in 2002 has not been arbitrarily designed into "the friendly
computer" (Microsoft on its XP website). Rather, the eleven principles
seem to rest on specific socio-political assumptions that Apple
expresses in the /Interface Guidelines' discourse on the power user and
stability. First, the /Guideline strictly separates the features that
the so-called 'power user' needs from those that 'the rest of us' play
with. The /Guidelines explicitly advises the interface designer not to
"hide features in your application by using abstract commands" (8) or
not to "use technical jargon or computer science terminology" (307).
Whereas the power user of an Emacs editor (http://www.emacs.org) might
have extensive keyboard shortcuts that call up the many program
functions, the 'rest of us' is better left with a few directly visible
choices in plain English, or so Apple says. After all, we might even do
harm to the computer, or, as the Guidelines nicely put it, the goal of
today's 'safety net' interfaces is to achieve "a balance between
providing (the power) user with the capabilities they need to get their
work done and preventing (the rest of us) from destroying data" (9).
Such expert politics are incorporated into today's interface design in
many ways, most notably perhaps in Windows XP, Microsoft's most recent
computer operating system, in the so-called principle of gradual
disclosure. Gradual disclosure means that, for instance, in Office
applications such as Microsoft Word, the menus only show a small number
of commands by default (such as 'format paragraph'), but hide more
specific commands (such as 'format styles') from the user - those
commands are only available to the 'power user' who moves his
point-and-click device over a little arrow at the bottom of the opened
menu. Another, more general instance of gradual disclosure in the
Windows XP interface is data visibility in the file manager program
Explorer. By default, the Explorer starts with a subfolder that only
contains user data ('My files') and hides the content of subfolders from
view that contain programs or system files, such as C:/Progam Files or
C:/Windows/System. On opening such a subfolder, the Windows XP
interface warns: "This folder contains files that are important for your
systems stability. You should not modify the content of this folder."
For expert use, below the large letter warning is a link that reads in
small font "Show folder content." Expert interface politics is closely
intertwined with the discourse of stability that serves as the other
bottom line of Apple's eleven design principles. Not regarding Apple's
claim to sell the more stable operating system, both Mac OS and
Microsoft Windows exhibit design features that back up the user
perception of stability and continuity of a HCI. The most prominent
example of stability design is the taskbar that is located at the bottom
of Windows XP and at the top in most Mac OS. It doesn't matter how many
programs you are running, in Windows XP the taskbar is always visible
and by default tells you the local date and time (thus locating your
physical body), always shows you a pop-up start menu (or gives power to
execute commands), and usually iconically represents all programs that
are running in small form. The taskbar visually remains 'always on
top,' meaning that you cannot move another window over it, so it's the
most prominent stability feature in every interface (even alternative
shells such as LightStep have one). Other stability features include
the feedback that the interface gives when the computer calculates for a
longer time - usually this is done in the form of a growing bar at the
bottom of a window (such as Internet Explorer) or a 'remaining time'
pop-up. Familiar icons such as the Recycle Bin or Trash that have a
specific place in many different interfaces, the upper left corner in
Windows XP, make novice users feel 'at home' instantly. Also, the
continuity of design throughout all applications is important for the
stable look-and-feel: Every program has menu bar and it's first and
second entries are usually 'File' and 'Edit.' Now, all these stability
features are linked to the expert discourse in the following way: They
actually cover up an /instable system to 'the rest of us' (Apple has
even taken this into the design of the iMac and the iBook themselves:
closed, shining entities). If you run an alternative shell such as
LightStep, you'll see on a little monitor the data writing and deletion
that constantly takes place on your computer's virtual an physical
memory; thus, much of the data representation you have in, say, the
Explorer file manager is actually a reference to a quite unstable heap
of data. As even the /Apple Guidelines remind the designer-expert, "it
is the /perception/ of stability that you want to preserve, not
stability in any strict physical sense" (11).
Given the inherent instability of new media objects, it is perhaps
little surprising that Jacob Nielsen and Don Gentner came up with their
Antimac interface that turns most of the Apple principles upside down
(although the authors, for whatever reason, emphasize that they do not
think the Mac is bad). The basic assumption of Nielsen and Gentner in
their "The Antimac: Violating the Apple Human Interface Guidelines" is
that while the Mac/Microsoft interface that we're using today might be
appropriate to teach what a computer can do to novice users, today's
computer users are "people with extensive computer experience who want
to manipulate huge numbers of complex information objects while being
connected to a network shared by immense numbers of other users and
computers." In other words: Apple's 80 percent solution doesn't work if
your users are the "Post-Nintendo Generation" (Nielsen and Gentner
1995). In the /Antimac, Gentner and Nielsen detect a number of
problematic aspects of what they call the "WIMP model" (windows, icons,
menus, pointer) of the original Apple Macintosh HCI. Although the
principle of metaphor usage, for instance, may help the novice user
("Oh, the trash can is where my deleted file went!"), metaphors usually
hide those computer capabilities that go beyond the actual metaphor: The
trash can, for instance, saves deleted files on every physical drive
separately, but if you 'empty' it, all data are gone from all drives -
the possibility to only delete files on, let's say, the C: drive, is
undermined by the very use of the metaphor. The problem with the
principle of consistency according to Gentner and Nielsen is that,
although learning might be reduced if new media objects look the same,
new possibilities are overlooked. As the /Antimac puts it, for Apple
and Microsoft we're "still children in the computer age, and children
like stability. They want to hear the same bedtime story or watch the
same video again and again. But as we grow more capable and are better
able to cope with a changing world, we become more comfortable with
changes and even seek novelty for its own sake." Finally, WYSIWYG is
inappropriate to computer usage in 2002, since the "What you see is what
you get"-principle "assumes there is only one useful representation of
the information: that of the final printed report." The principle thus
overlooks that "it may be useful to have a different representation when
preparing the document. For example, we may want to see formatting
symbols or margin outlines, or it may be useful to see index terms
assembled in the margin while we are composing." Against this interface
politics of the beginning new media age, Gentner and Nielsen set an
interface that features the central role of language, richer internal
representation of objects, expert users , and shared control. Instead
of a poor office imitation, then, they see the computer today as an
ubiquitous tool to work, communicate, and play. The computer, so to
say, introduced the new ethic of "You won't always have to work that
hard" instead of giving you the early digital capitalist "Power to Be
Your Best" (Nielsen). Of course, when recalling Althusser's statement
that "men represent their real conditions of existence to themselves in
an imaginary form," even the Antimac becomes another imaginary, cultural
representation of absent socio-political reality. For instance, to
invoke a gendered reading of both Mac and Antimac: One could say with
Robert Milthorp that the Antimac is a result of men's "fascination with
technology (that) is linked to the masculine need to be in control of
the material world, to know how to extend that control, to be able to
act, and to be independent of reliance on others" is expressed in the
Antimac (Milthorp, Rob 137). However, I do not want to elaborate on
this criticism here but instead underscore that the very Antimac
alternative points to the existence of the space of inherent instability
of new media objects that is so central to my argument.
Now, in the coming age of the highly networked computer, Web browsers
are the other important new media objects that frame and politically
shape data within HCI. I'll look at the AOL 8.0 software and
Mozilla/Netscape 7 in order to show how this might be so. America
Online, a subdivision of AOL Time Warner, claims to be the single
largest access provider to the Internet - although relative in size in
relation to the US population, the AOL member figures actually do
suggest that many people (at least in the U.S.) access the Web through
the company's software. According to the company's website, AOL has
"more than 35 million members of its flagship AOL service along with
more than 3 million CompuServe members, 120 million registered users of
ICQ, and 48 million registered users of the Netscape.com service" (AOL,
2002). In addition to those 206 million users, AOL also operates
popular services on the World Wide Web such as AOL Instant Messenger and
Winamp, a music file sharing tool. No doubt then, that "America Online
has played a major role in creating the consumer online experience
worldwide," as the AOL website claims. Now, similar to Apple's 80
percent solution, at the heart of the AOL marketing strategy stands
"providing convenient, easy-to-use services for mass-market consumers."
But as we have seen above, the internet remains a new media object that
is unstable and multilayered, so in reaction to that, AOL has developed
its own version of an Apple Macintosh desktop for internet access,
namely, the AOL software suite. This software contains the AOL Mosaic
Web browser, an email client, chat programs and other tools conveniently
compiled in one package that is downloadable from the company's website
or freely available in many stores or on CD-ROMs that come with digital
lifestyle magazines. Similar to Apple's Macintosh, the installation
instructions of AOL's Website address a novice computer user: They
highlight the ease and simplicity of the software usage ("Follow the
easy online instructions to install your FREE America Online Software!")
and the speed with which the novice user gets accustomed to the new
programs ("You'll be enjoying the benefits of America Online in no
time!"). AOL is also quick to warn the novice user of the dangers that
might lurk in her computer file system: "Important: Be sure to write
down the filename and path to the directory where you save the file."
After the user has installed the program ("Look for a file that starts
with 'setup'!"), she is ready to experience "the Internet in an
instant!" The mission statement page of the German AOL subdivision
nicely summarizes the general AOL strategy: "In the infinite world of
the internet, AOL offers you a home. An intimate place where you meet
friends, where you feel cared-for and safe."
Similar to Apple, AOL is aware that the company is painting a neat
surface/interface over something that is messy, fragile and unstable.
And similar to Apple, AOL silences the space for difference on the
internet; a strategy that can be deduced from what AOL holds its term
'netwise' to mean. "Like the rest of the world, the Internet may
contain some material that is inappropriate for young audiences" is one
of the conclusions of the instability factor for AOL (one might imagine
"for any audience" as an alternative ending). To answer this problem,
AOL has coined the term 'netwise' for what it likes its member to be:
informed about those countermeasures against messiness and instability
which AOL gives you: "Working together, we can make the online
environment a safe and rewarding experience." Not surprisingly, in
stark contradiction to what one might imagine a term 'netwise' to
signify, AOL safety strategies heavily rely on automatic filtering
software ("Filtering software like CyberPatrol, NetNanny, and SurfWatch
can help keep children from inappropriate online areas") and
AOL-controlled on-the-surface security settings (similar to Microsoft
Internet Explorer). AOL devised software improvements for the 'netwise'
user such as "giving members greater control over their incoming mail
with a mail sorting feature that lets them choose to view only messages
from people they know," alongside with new "self-expression features,
including animated Buddy Icons, a choice of new Buddy Sounds, colorful
backgrounds and stationary that let members tailor the appearance of
their e-mail and instant messages to reflect their personalities and
moods", and a "choice of (...) six different versions of the Welcome
Screen offering distinct programming tailored to member interests and
updated through the day" that include "Headlines, Business News and
Sports," "Headlines, Latest Music, Games and Homework Help," and
"Headlines Nightlife and Great Discoveries." The 'netwise' person,
then, uses MatchChat, Music Share, and Buddy Share technologies to let
their computer find out their likings. Generally, then, "AOL seeks to
build a global medium as central to people's lives as the telephone or
television," and this is where AOL does not grip the new logic of the
net. German hacker Dragan Espenschied has talked about this in his "How
AOL influences its users," observing that within the AOL software
environment, basic principles of new media do not work anymore, namely,
"there does not exist an option to save or edit the content of a viewed
site" and "one cannot rely on the principle of copy and paste." His
analysis culminates in the assertion that "the totality of the Web turns
into one single page out of AOL's content."
In contrast to the restrictive interface politics of America Online
stand the capabilities of the Mozilla software suite which derived from
the Netscape source code. Since the date of Microsoft's coupling of its
fast-and-easy browser program Internet Explorer to the operating system
Windows in 1995, the percentage of users that surf the Web with the
Netscape browser has been steadily on the decline. In fact, Netscape is
only used by an estimated 10% of internet surfers to access Websites,
and as a result of that many Websites such as amazon.com target their
content only to the capabilities of Internet Explorer (you can't click
on "Buy this book" for example as a Netscape 7 user - there's no button
there to click on). Notwithstanding the low usage, Netscape, now in
version 7, still does stand as an alternative approach to the politics
of Web browsing of AOL or the Internet Explorer, possibly due to the
fact that at one point of usage decline, the company thought the browser
battle against the Internet Explorer lost and decided to lay open the
program source code of the Netscape browser and to make it freely
available for modification and redistribution. With one limitation:
Netscape used a restricted general public licence (GPL) that allowed the
company to later integrate the free, changed code into its proprietary
program again (more restrictive free software licences such as the GPL
only allow the free, open distribution of the program). The result of
this act is the Mozilla 1.0 browser which again serves as the code basis
for Netscape 7. Again, this is not to suggest that Netscape 7 is the
'better' choice - it's just to highlight formal differences that do
suggest a different user politics of AOL or the Internet Explorer and
Netscape 7/Mozilla. Now, both Netscape 7 and Mozilla are freely
available for download from their respective websites (www.netscape.com
and www.mozilla.org). Since the programs are software suites, modules
such as the web browser, the email client, or the usenet mailer have a
similar look and feel, and, in contrast to AOL, they allow copying and
pasting in between each module. In addition, the similar feeling of,
for example, writing an email or writing a usenet post, gives a feeling
of equality of the action. Similarly, the equal look of an FTP address
and an http address 'reminds' the user of the fact that email and http
are only the most prominent, but not all, services of the internet. As
Dragan Espenschied has argued, Netscape/Mozilla's "view source" menu
that makes the HTML source code of a website accessible to the user has
pushed the HTML prominence to a large extent. Microsoft Internet
Explorer has only later integrated this feature and it's still less
prominently placed. The view source menu virtually gives the user power
in that it shows how the website was done - if you copy the source code
you have an identical website. In addition to this, Netscape/Mozilla
comes with an HTML composer, which allows you to copy and edit the text
of any Website you access or to program your own. Furthermore, in the
browser 'file' menu, "edit page" becomes an entry just like "open page,"
so there is a whole emphasis on the creation of content that is
non-existent in AOL or the Internet Explorer that does not come with a
composer program. Mozilla even goes beyond such features: The browser
is able to suppress advertising pop-ups, and many script behaviors (most
of which are potentially dangerous to your computer) can be individually
configured.
Seemingly, data are "collections of individual items, with every item
possessing the same significance as any other," as Lev Manovich has
argued in his article on the concept of the database (Manovich 2001:
218). But although such 'data equality' might hold true on the level of
a code stream or on a strictly physical level, I have suggested in this
chapter that the modules in a database have a specific, socio-political
organization within the HCI. It seems that in today's interfaces,
programmers and interface designers (while trying to get closer to the
machine and actually admiring its instability), seek to /sell the
computer as a nice looking, reliable work tool (which it is not) to 'the
rest of us.' Apart from interface design strategies, this dichotomy
also figures in the fact that many companies that engage in discourses
of stability and constancy make a large amount of revenues from
so-called 'second level services' such as bug fixing, anti-virus
security, and installation support. Now, recall that, for Ernesto
Laclau, "self-determination can only proceed through processes of
identification" (Laclau 1996, 55) and that a subject/user emerges in the
distance between the undecidability of the structure and the decision.
Recall also that a subject/user is necessarily hegemonically represented
as a part of a larger socio-political group. If this is so, a
programmer and a designer of a user interface deals with several, in
part contradictory, aspects: First, the decision with what to identify
is largely taken on the level of the interface designer, not on the user
level; the subject/user is only left to decide about pre-selected items.
Ironically, if subjectivation in Laclau's sense then becomes limited in
the context of user interface design (the undecidability of the
structure becomes silenced through the interface design process), the
very goal of interface design to foster identification with a specific
company or interface is undermined: An identification with a Microsoft
interface might as well turn into a passion for the Apple Macintosh
desktop. So here's a gap in new media that, in my mind, might open up a
space for resignification (in a traditional Marxist framework, one could
even say that someone gave away the means of production here) in the
sense that interface logic then becomes essentially a logic of making
something unstable repeatedly look nice (each time a start Windows XP -
until it crashes). Notwithstanding coherent interface designs, however,
I think that users 'feel' that the computer is essentially unstable, and
that and how designers try to paint over this fact - "the presence of
the 'Other' (the crash) prevents me from being totally myself," as
Laclau says (125). Within such an framework, hypermedia artists and
everyday users become part of a group that tries to bridge unstable
media - the imagined, shared experience of, for instance, the Apple
Macintosh 'user community might serve the desire for intimacy of each
user. In the next chapter, I'll try to interpret hypermedia authoring
as trying to attain (empty) intimacy in the space that instability opens
up.
--
Henning Ziegler
http://www.henningziegler.de
http://www.micromusic.net