Originally published in 'post.pic: imageboards, tagging, tool images, visual
studies - a primer by practitioners', a publication from the research group
'Communication in a Digital Age' of the Piet Zwart Institute, Willem de Kooning
Academy, Rotterdam University.

---

Tagging
Florian Cramer, Aymeric Mansoux


The most familiar form of tagging is probably still the graffiti tag:
a hybrid of writing, calligraphy and images, typically used as a kind of
signature by graffiti artists, mainly on buildings and trains. Graffiti
signatures have existed since the late 1960s, when graffiti writers
such as Cornbread and TAKI 183 first received media attention. The
phenomenon was already widespread by the late 1970s and early 1980s,
when the term “graffiti tag” first appeared in American subculture. In
computing, meta-information and tagging are standard features of file
systems and databases; on the most general level, tagging is simply
the labelling of an information object (typically a file) using a name or
other keywords. Meta-tagging entered the popular consciousness when
it became a central feature of the so-called “Web 2.0”. On websites such
as YouTube and Flickr, meta-tagging transformed digital image culture:
visual elements could now be identified through the sets of tags attri-
buted to them, such as “apple”, “green” and “photograph” in the case
of a photo of a green apple. Meta tagging is also a key feature of web
sites such as del.icio.us, a social bookmarking site where people assign
keyword tags to web links (in the context of computers and the Internet,
the terms keywords and meta tags are, for all intents and purposes,
interchangeable).

But is there any actual resemblance between an urban graffiti wall, and
our green apple in Flickr? Or is it merely some linguistic coincidence
that makes us use the same word for both image cultures? Basic semio-
tics might help to shed some light on this question. Semiotics (literally
the study of signs) originated in the late nineteenth century through the
pioneering work of the American philosopher Charles Sanders Peirce;
its main concern is the analysis of signs and signification processes.
Peirce divided signs into three basic categories: icon, index and symbol.
If we consider a photograph of the smoke caused by a forest fire, we
identify the image as smoke because of its iconic resemblance to actual
smoke. But if we are standing outside, looking at actual smoke, we inter-
pret this as an “indexical sign” (a trace and an indicator) of fire. Finally,
there is the symbolic sign, the simplest example of which would be the
five-letter word “smoke” (which bears no iconic resemblance to smoke,
nor is it an indexical trace of smoke).

If we apply this classification to tagging, then the most intuitive asso-
ciation would be to define tags as symbolic signs: whether they be
calligraphic graffiti, or meta tags on digital files, they do not iconically
depict anything, but rather resemble writing, as signatures and mark-
ings. Or perhaps graffiti aims to be iconic, since it emphasizes visual
form and visual associations with other elements of the signature.
Yet such explanations are not entirely satisfactory. A historical analysis
of graffiti tags, and of their evolution within electronic media, may
perhaps give us more insight.

In the opening scene of “Wild Style”, an early 1980s cult film about the
old-school hip hop and graffiti scenes, we see a graffiti artist spraying
a tag based on his own name (“Ray”) on a subway train in a tunnel.
The tag is not merely a symbolic inscription of the name – nor is the
juxtaposition of the symbolic sign “Ray” with the iconic footage in the
film of Ray’s heads and hands. What the scene (and indeed most of
the film) is about, is the marking of a site, using the subway train as a
mobile territory which expands the inscription throughout the urban
space. Another 1980s film, Dennis Hopper’s “Colors”, tells the story
of two cops in gangland Los Angeles. In one key scene, they catch a
gang member spraying his tag on another gang’s wall. They immedi-
ately move him out of “enemy” territory and back into his own gang’s,
even showing him a place where he can spray without getting himself
killed. And indeed, his tag functions as a declaration of gang war, which
further escalates as the film progresses. Tagging becomes an elaborate
form of insult, “dissing” or verbal abuse, which is of course risky in any
situation, besides gangland tagging conventions. Here the graffiti tag is
not a symbolic or iconic representation, but an indexical territorial mark-
ing. Likewise, early graffiti signatures often consisted of the tagger’s
nickname and street number.

Dennis Hopper’s “Colors” can help us understand one major cultural
difference between Europe and America, at least regarding the way
graffiti tags are perceived. In a city like Rotterdam, graffiti is seen as
a mostly harmless expression of rebellious youth culture; whereas
in contemporary America, graffiti tags are habitually associated with
criminal gangs: the territorial marking of no-go areas. This is an actual
claim (and not merely a symbolic one) that a group operating outside
of state authority is in control of a territory; not unlike the way dogs will
mark their territory with their urine. And so, beyond the symbolic and
iconic elements of graffiti tagging, its most powerful aspect is clearly an
indexical one.

In the underground computer scene of the 1980s and 1990s, graffiti
culture was more or less seamlessly expanded into an electronic graf-
fiti culture among hackers and crackers. This manifested itself most
prominently in the “cracktros” for illegally copied computer games:
besides removing the copy protection of games on floppy disks, cracker
groups (identifying themselves using pseudonyms) would add their
own intro screen to the game. Not only did the intro symbolically tag
the game with the name of the cracker group; the visual aesthetics of
these screens quickly evolved from using plain text and display hacks,
to visually emulating actual graffiti writing. Here the territory is shifted,
from the city to the computer game and the distribution of media. The
practice later branched out into several other activities, such as the
“demo” scene, where cracktros grew into complex, computer-generative
audiovisual animations.

Another similar subculture was FTP tagging. For a brief period in the
early 2000s, the Internet had become a vast jungle of poorly maintain-
ed servers. These machines were leftovers from the late dotcom boom,
when many of the new companies providing web hosting and server
administration had little or no understanding of network security.
For approximately two years this provided an extraordinary new terri-
tory for amateur pirates, who made public FTP servers their playground.
FTP taggers, once they had located the “pub” folder of these servers,
used them for sharing their own files, including illegal material, cracked
software, music, and videos. Once a “pub” folder had been found, it was
tagged to mark it as the sole property of the individual or the group who
had discovered it. This tag was simply a file directory path stating the
name of the “pub” owner. For example:

/tmp/.test/=-=/-/=-=Tagged by GT!!!!!=-=/-/=-=/Filled.by.S/c/a/r/f/a/c/e/for/(^.^)Y0FXP(^.^)/

Such tags were not plainly visible, but could be read by looking at all the
subfolders contained in the “pub” folder. In the example above, the tag
indicates that the folder was discovered by “GT” and that the files in the
last folder were uploaded by “Scarface”, both of whom are members of
the group “Y0FXP”. To prevent “pubstealing” and in-group vandalism,
and to make it harder for anyone to get rid of the original “squatters”, a
technique known as “dirlocking” was developed to make the tag impos-
sible to remove (analogous to using a permanent marker for signing on
a graffiti wall). To make things even trickier, it became common practice
to upload thousands of variations of the same tag all at once, thus creat-
ing a complete file-system maze on top of the lock.

Eventually, pubstealing evolved from stealing storage space to simply
deleting any uploaded content – just for the sake of it, or for the thrill
and satisfaction of solving a tag maze or breaking a locked tag, in order
to re-appropriate the territory and mark it with one’s own tag. In this
subculture, tagging was not merely a matter of marking or symbolically
describing a territory, but of actually creating it for a group of peers. The
tag thus became a means of granting or preventing access and informa-
tion retrieval – not unlike the function of tags in systems such as Flickr
today.

Tagging has become one of the core features of the so-called Web 2.0.
Vodafone even advertises a mobile phone service with the slogan
“Tagging, posting, chatting, surfing. And making phone calls” – thus
recounting a history of media in reverse, in which tagging has become
the most contemporary (and most important) form of telecommunica-
tion media usage.

Tagging the image of the apple with attributes such as “green” and
“Granny Smith” could of course be considered redundant, given that we
can already see the apple in the image. However, we need such words
in order to be able to find the image at all. Without the tags, the image
cannot be retrieved from any database, search engine or web site. Con-
sequently, the so-called “Semantic Web” is nothing more (or less) than a
standardized, comprehensive meta-tagging system for the World Wide
Web, allowing for better and easier retrieval of information. The way in
which we use computers to access images is predominantly linguistic.
Even before the Web, before Google, YouTube and Flickr, we were al-
ready well acquainted with linguistic tagging – giving images filenames
such as “apple.jpg”, which are, of course, nothing but tags: in fact, the
oldest system of meta tagging used in computers. We cannot use a
picture of one apple in order to “google” other images of other kinds of
apples – at least not without some prior human and computer tagging:
a set of numerical pixel patterns common to several digital photographs
of apples, and a human programmer who has identified this set as cor-
responding to the English word “apple”.

This should put into perspective any overblown claims of a “pictorial
turn” in our culture, at least as far as the Internet is concerned. At the
very least, we must reconsider the notion of images being in opposition
to text, or being an entirely different medium than text. In our contem-
porary visual culture, we can no longer separate one from the other.
The three semiotic properties of images – symbolic, iconic and indexical
– converge in these systems; much in the same way that indexical mark-
ing of territory, symbolic writing, and iconic pictorial representation con-
verged in the older visual medium of the graffiti. And so the leap from
the graffiti wall to Flickr may not be so much of a leap at all.