<nettime> Ippolita Collective: The Dark Side of Google, Chapter 6 (part 2)

["Patrice Riemens" <patrice@xs4all.nl>]

NB this book and translation are published under Creative Commons
license 2.0 (Attribution, Non Commercial, Share Alike).
Commercial distribution requires the authorisation of the copyright
holders: Ippolita Collective and Feltrinelli Editore, Milano (.it)


Ippolita Collective

The Dark Side of Google (continued)

--------------------------------------------------------------------
Hiya,

To my regret I must confess that Chapter 6 was up to now the one I found
least rewarding to translate. A possible cause is the French version from
which I translate in the first instance, where a flurry of semantic
(over-)sophistication and grammatical contortionism might be there in order
to enshroud a - sometimes far-reaching - unfamiliarity with the concepts.
But I am afraid the authors are also to 'blame'. Their basic message is
fairly simple: Google causes passivity, and present results in an
artificially simplified and monolithic fashion, greatly reducing the joy,
amazement, and serendipity of (re)search in the process.

But whereas I tend to agree with the first statement - while at the same
time holding the victims of 'Google palsy' as primarily responsible
themselves, I do have some doubts, out of personal experience, about the
second part. Yes, I admit to seldom go beyond the first page of returns
(but on the other hand I very often do not go for the first link
suggested), unless my query was very vague. But I usually quite happily
jump from one link to the other, which in its turn, and especially if the
link was to something like Wikipedia - (un)surprisingly almost always very
high on the list, in my queries at least - can lead very far from the
original 'search intention' indeed. To put it simply, the argument that
Google does not - contrarily to what it might suggest - covers 'all'
information potentially available is OK, yet Google does provide enough of
the stuff to drown in already. Besides one ought to be aware of the fact
that the Internet itself, does not contain 'all' information, and is
sometimes (e.g. if you delve in historical (re)search) shockingly
deficient. But that is also a valuable information in itself.

Regarding Google's (in)famous "I am Feeling Lucky" button, which the
authors appear to consider a supreme  indictment of Mountain View's bogus
claim to digital age infallibility, it should be noted that it would seem
more to apply to the French variant ("J'ai de la chance", I _am_ lucky)
than to the English, let alone the Dutch one, for instance ("I waag een
gok" - 'I try my luck').

I do hope that the next chapter ("Technocracy") will be more enlightening.

Cheers from Pune, patrizio and Diiiinooos!
(came here from Bangalore by First Class AC on the Udyan Express, all
alone in a huge 4-berths compartment (3 no-shows), but return Sleeper
Class (aka rail-sardine-can, but one can lie down - like the sardines ;-) 
in the Sampak Kranti Express - admittedly a crack train. Do pray for us!
Oh yeah, day temps: 40 oC and counting up...)
------------------------------------------------------------------



Chapter 6 (part 2)

The Myth of instantaneous search

Since it is clear that Google's data 'capital' , gigantic as it is, will
never correspond to the totality of {the information present on} the Web,
presenting oneself as an 'instantaneous' interface, bridging the gap
between the users search intentions and the so-called 'exact' result
smacks of naivety - or of deceit.

Since the Web consists of nodes (pages) and arcs (links), every time one
browses it by visiting pages, one is follows up links constituting a
trajectory analysable through the mathematical models of the graph theory.

The pre-set orientations search engines will propose us will always lead
us to the 'right' object, indifferent of the dimensions the Web might have
or get {in future}. By applying efficiency and efficaciousness criteria, a
search engine will chart out of query the 'optimised'  trajectory, meaning
that the number of nodes hit will be low, and {hence} the time taken by
the search will look nearly instantaneous. Google actually pushes in the
direction of one single trajectory, something illustrated by the "I'm
feeling lucky" button on its main page.

This 'optimisation' squeezes search into a three pronged sequential
scheme: user-algorithm-goal. On the long term, this dynamic leads to
'digital passivity', a stage where we simply wait till results are brought
to us, for us to choose among them.

Moreover, this efficiency/ efficaciousness is paradoxically grounded not
on an increase in the size of the data pool where searches are conducted,
but on its opposite, on a limitation of the access to the information
'capital' , since no trajectory proposed by the search engine will ever
take place in real time [French 'the moment t'] on the network, but will
be calculated first according to what has actually been archived, and the
user personalisation obtained through filters and cookies.

The access to the information offered by Google is fast, very fast, and
looks even immediate, to the point of suggesting the annihilation of time,
and to imply the existence of an immensity of data that have been perused
for the purpose. The mediation of technology (through interfaces,
algorithms, pre-set searches, etc.) makes this temporal 'annihilation'
possible as well as {the feeling of} practically immediate access [*N4].
The rapidity of results return, however, has a detrimental [Indians would
say:'deliterious ;-)] effect on the quality of the search. As everyone is
aware who has conducted (re)search herself, the time one spend on
(re)searching is a determinant element of the experience: to map out one's
own {research} path, to make choices according to the moment, all this
generate a feeling of being into it and is {deeply} satisfying. Google
allows us to 'localise' in space (that is its own multidimensional space)
what we want, but, however brief the time spent waiting for the result, we
always adopt a passive attitude in front of the technological oracle.

In an active (re)search drive, the aim is no longer about 'access' to the
data, but to accomplish a rich and variegated journey, and to use the
(re)search endeavour for mapping out complex trajectories. Efficiency as a
concept vanishes. The larger the number of visited nodes, the greater the
complexity of the interlinkages we conceive, the more numerous occasions
will be to trigger significant choices, and to refine our (re)search. This
approach allows for a cognitive enrichment going well beyond the immediate
performance. For instance, when we visit links offered to us by a site we
are visiting, and then continue our navigation on sites that have been
marked as congenial, we create {every time} a unique trajectory; maybe
we'll even resort to bookmarking them. Such a procedure is {starkly}  at
variance with a coherent user-algorithm-result sequence, but it does
create a rich path full of sidelines, of branches, of {cognitive} jumps
and winding detours, all catering to a non-linear cognitive desire [*N5].

To conclude, search engines are perfect tools for fulfilling the
quantitative aspects of a (re)search taking place within an already fully
structured resource pool, such as are lexicons, cyclopedias, etc.
{Here,}The quantity is directly in proportion to the accumulation and
computing potential: Google's reach obviously dwarfs that of all its
competitors, but in order to retain its position, Google needs to
constantly expand in terms of algorithms, machines, users, etc.

Conversely, quality needs not necessarily to reside with technological
prowess or economic might. Nobody {in her right mind} believes that the
results returned correspond to the full spectrum of available information:
the emergence of the best possible path cannot be foreseen, cannot be
computed, but can only be arrived at step by step.

Under the veil of the myth

 The positioning values of Google's ranking do not correspond to any clear
evaluation criterion: yet, in the majority of cases results returned are
[look?] exhaustive, that is, we can in no way tell whether something has
escaped the spider, unless one is an expert in the issue at stake and
knows a resource that has not been indexed {by Google}.

The capillary distribution of its search tools has made Google a 'de
facto' standard. The white space ('blank box') where we type the keywords
of our (re)search functions for the user as 'Weltanschaung' of sorts,
promoting a very particular world-view, that of the idea of 'total
service': the search engine will answer any question, and will satisfy all
requests made in the realm of the Internet.

Epistemologically speaking, the 'blank box' represents a cognitive model
of the organisation of knowledge: We request {through} the white space an
answer to all the search intentions we have put forward: indifferent
whether  we wanted documents, or further information, or data, or that we
simply wanted to 'navigate'. The (re)search activity becomes completely
merged with the entity that provides the service, Google, [of which we
have an invading perception (?)].

The habit of using this tool becomes ingrained behaviour, a repetitive
activity: it becomes very difficult for users to imagine a different way
to satisfy their need for 'input'. They have become tied up to the
reassuring efficiency/ efficaciousness of the 'blank box'.

To be active on the Web, and hence to need access interfaces and tools for
unearthing information and setting out paths is a is a profoundly
contextual and diversified occupation. (Re)search is everything but
homogenous and cannot be reduced to the use of the 'blank box'. What we
request and what we desire does not solely stem from a desire that can be
expressed in the analytical terms of quantitative information, but is
something that also hinges upon the way we approach (re)search, the
context in which we undertake that (re)search, our own cultural background
and {last but not least} on our aptitude to confront novelty, explore new
territories, and face diversity {in general}. It is impossible to satisfy
the quest for information through a one size fits all solution.

Since the indexation of {web}pages is {by definition only} incomplete, in
the sense that it is a selection obtained through {the} ranking {system},
what Google does offer us is the prosaic possibility to 'encounter
'something'  we might find interesting and/ or useful in its overflowing
amount of data in its collection of subjects [issues]. A (re)search
intention, however, implies a desire to find, or even to discover,
"everything what one  doesn't know but that is possible to learn about".
The {'good'} giant then appears for what he is: enormous, extended,
branching out, but not necessarily adapted to our (re)search purposes.

(Re)search Models.

The ambiguity entertained by search engines, wanting us to 'search in
{their) infinite environment' rather that in a closed, localised world
that conforms to our (re)search intentions, comes from the formal
superimposition of two {distinct} levels, that of the interface [*N6] and
that of the organisation. The interface, in this particular context is the
technological element  through one accesses the information and the search
gets executed; the organisation, on the other hand,  is the architecture,
the technological model through which information is archived and
disposed. The two levels {obviously} influence each other:
organisation-realated choices prescribe the use of specific interfaces,
while the information that are visualised through these interfaces betray
in their form[at?] the way they are archived. [?]

The problem with this superimposition is that such information is
presented in the form of identifiable and unambiguous, single  data. The
user of Google moves in a linear fashion through the results list of the
ranking; in order to move from one result to the next she needs to go back
to the start list, with no cross-over linkages possible at the level of
the interface [?]

With search engines, one retrieves information, but without any
consideration being given to path that have been followed {to obtain it}.
The interface which directs our interactions is the 'blank box' where our
queries are inserted: at this first level of access, all information are
on the same plane [have the same rank(ing) ?] They are homogenous, yet at
the same time separate and fragmented in order to allow the listing of the
results as they have been arranged in order of pertinence by the
algorithm.

However, as far as the (re)searches one does on daily basis are concerned,
the same results can be linked together in all sorts of ways, and it is
not necessary to arrive at the same ordered arrangement every time, and
neither does only a single 'correct' result obtain; on the contrary, a
(re)search which is not about data structured like in a cyclopedia or a
dictionary or any other object of that kind (and that may also change {in
nature} over time), could well remain without an immediate answer, but
would on the contrary require an effort of creativity, of 'mixage', and of
recombination.

When a formal identity is being imposed between the level of the interface
and that of the organisation,  the outcome is {by necessity} a
constraining model. In Google's case, {as} we have to do with what is
perceived as an infinite power of search, the means to arrive at a result
are being substituted for the (re)search activity itself.

Let's take an example: ...
[the example taken is a French word, 'plume', whose English equivalent
('feather') would not yield the same illustrative power. Briefly, the
authors argue that if you 'Google' for that word, the first returns (out
of 6.700.000 !)  will be about everything ( various IT companies, a
circus, etc.) but birds-feathers or ink pens (also 'plume' in French) -
I'll need to sort out a nice equivalent with the collective (or 'invent'
one myself) - but maybe _you_ have an idea?]
...  A more extended perspective of what it means to 'discover'
information, and that would take the cognitive potential underlying every
information resource pool into account in a critical manner, would tend to
see the access-search function as a process of exploration and creation
rather than as one of localisation. The emphasis would then shift from
epistemology towards ontology: it is non longer sufficient to know the
information, but to become aware of our true role as creators of
information [N*7]. Search engines that operate at the access level are
therefore of no use for exploration, as they merely intervene on the first
{and basic} level of the presentation of information.

Browsing is the moment of true dynamism in the linking together of digital
objects, which are then able to express to the highest degree their
heuristic and communicative potential. This is something that is learnt
through experience, and it mutates as we are learning it, during the very
activity of exploring.

There is a major difference between searching and finding. Google makes us
'find' things, causing the satisfaction that goes with the feeling of
accumulation. But far more interesting that 'finding' is the search
itself. And maybe it would be even more rewarding to find, but not
completely, because that would mean that we are sill engaged in the act of
(re)searching.

A search engine is an instrumental model that arranges information into a
certain order. It would be more useful and also more commendable to
imagine models that (re)combine information, and {so} generate knowledge.

END of Chapter 6

(to be continued)


--------------------------
Translated by Patrice Riemens
This translation project is supported and facilitated by:

The Center for Internet and Society, Bangalore
(http://cis-india.org)
The Tactical Technology Collective, Bangalore Office
(http://www.tacticaltech.org)
Visthar, Dodda Gubbi post, Kothanyur-Bangalore (till March 31st, 2009)
(http://www.visthar.org)
The Meyberg-Acosta Household, Pune (from April 2, 2009)


#  distributed via <nettime>: no commercial use without permission
#  <nettime>  is a moderated mailing list for net criticism,
#  collaborative text filtering and cultural politics of the nets
#  more info: http://mail.kein.org/mailman/listinfo/nettime-l
#  archive: http://www.nettime.org contact: nettime@kein.org