The Knowledge-Attention-Gap: Do We Underestimate The Problem
Of Information Overload In Knowledge Management?
Ursula Schneider
(Karl Franzens University, Graz
ursula.schneider@uni-graz.at)
Abstract: The generation of technical knowledge abounds while
the underusage of existing knowledge potential remains a problem in business
as well as in society. Generally speaking value can be extracted from knowledge
in three ways:
- by exclusive use
- by faster access
- by better translation of public knowledge into products that yield
private profit
Each way requires different approaches to KM. But in all cases the problem
of how to deal with abundance arises: It arises at the individual as well
as at the level of interface design in a knowledge dividing society.
First ideas to solve that problem refer to the individual rather than
the interface design level:
- technical solutions
- psychological solutions
- neurological solutions
deal with the growing gap between abunding potential knowledge and scarce
human attention on the one hand and with restricted human capacity to process
information on the other.
For the time being a clear focus on good old virtues, such as will (focus),
modesty (less is more) and courage (to decide under conditions of incomplete
information and uncertainty) seem as trivial intellectually as hard to
implement in practice.
Key Words: Knowledge Management, Information Overload, Document-Explosion,
Intelligent Agents, Positive Ignorance
Categories: A
1 Prologue: A confusing state of the art?
In theory knowledge management remains a patchwork of unrelated approaches,
partly due to its interdisciplinary nature and partly because of a clear
preference to create as opposed to resolve. This state of the art is reflected
in the following quote:
"Knowledge Management (KM) is one now of the most ramified topics
in the business arena. Much of this amplitude can be attributed to the
number of fields that lay claim to the idea, or some part of it, including
computer and information science, business strategy, macro economics and
interpersonal dynamics to mention only a few. Proponents generally claim
that this symphony of schemes is appropriate given the important, transversal
and imminently practicable nature of Knowledge Management. Its critics,
on the other hand, are hearing either a re-mix of older refrains, schizoid
melodies or an outright cacophony. All parties agree that we are nonetheless
witnessing an explosion of interest in the term Knowledge Management and
all that it may or may not imply." [Despres et.
al. 2001]
Practitioners have mostly gone through the typical cycles of enthusiasm
and disillusionment and started to ask tough business questions [Schneider
2001]. Knowledge management activity needs to be aligned with strategy
and needs to be integrated into business operations.
Suppliers of technical solutions have also learnt their lessons at least
at the level of sales' rhetoric: They stress the enabling character of
technology, elaborate on solutions which support every day work and put
effort into integrating existing knowledge-management-related systems.
All three groups of players mentioned above seem to have arrived at
a turning point: Academicians struggle to define those characteristics
of knowing that are relevant to its process design, practitioners start
to accept that the new wonder pill will not solve organizational contradictions
and behavioural shortcomings and software designers as well as system engineers
refer more to context. But there are some caveats peculiar to all three
groups.
Academicians - obviously under the "publish or perish" dictate
- fail to comply with the maxim of their own discipline: not to reinvent
the wheel and to build on existing knowledge*. My - so far unsystematic
- perception of conference papers and journal articles reveals astonishing
evidence[1]: A plenitude of micro studies
are undertaken to research knowledge creation and transfer in most diverse
situations, from face to face to virtual, from national to intercultural,
from not for profit to business organisations, from operations to managerial
tasks, and a plethora of models are constructed by PhD students elaborating
on similar questions - without taking into account their mutual incongruences
and their incompatibilities with conflicting theoretical models.
Why does this paradox just happen to arise in a field that is focused
on the balance between knowledge exploitation and knowledge generation?
Several reasons come to mind:
Incentive structures that prefer generation to exploitation have already
been mentioned. Furthermore, references, instead of covering the existing
knowledge* - in the sense of either building on or challenging its substance
- seem to have degenerated into a pure ritual of name and notion (model)
dropping. But the point to be made here, reaches beyond critisizing awkward
habits within the scientific community and extends to the other players.
My hypothesis is that the growing ease with which we produce, store
and diffuse intellectual products globally via electronic devices will
necessarily result in the following problems:
[1] see for instance the Third European
Conference on Organizational Knowledge, Learning and Capabilities, 5 -
6 April 2002, Athens
- First, the sheer abundance of what could potentially be turned into
knowledge* poses the challenge to deal with volume - or put in more popular
terms - to cope with information overkill.
- Secondly, the principal volume-related incapacity to know it all poses
a challenge to cope with incompleteness, ambiguity and uncertainty while
orientational knowledge tends to assume a shorter life-cycle in itself.
- Thirdly, volume and speed in the shift in relevance, train and reinforce
our capacity to shift attention rapidly but seem to weaken our capacity
to concentrate on substance.
- Finally, the ease with which we store, produce and diffuse knowledge*,
despite of organizing features of office software, seems to result in a
gradual disappearance of the most simple organisational capacities, so
that related incidents, such as correspondence and details pertinent to
meetings and conferences are stored in a dispersed or random way making
it difficult for individuals fighting tough time schedules to retrieve
and to reintegrate them.
Let's take a look at the practical field and see whether we can observe
similar challenges: The first wave of knowledge related activities has
led to what I call the great documentation and externalisation initiative
that in itself has resulted in rather unorganised abundance. Searching
devices have hardly kept pace with an explosive growth of mail and documents
deposited on the world wide web or on intra- and extranets[2].
Content providers as well as chief knowledge officers had to learn that
production was not the bottleneck. There was no shortage of writers or
senders but of readers and receivers, as reflected in the development of
incentives: First everyone was rewarded who made a document available to
others. Then, clicks were counted, so that senders learned to increase
their ability to create attractive key words. Finally procedures of subjective
evaluation were developed to capture the value created by any document
made available on e-devices. In the end what should be rewarded, if at
all, but that is another story, is the value created on markets to customers
which we are still far from capturing.
A loss of office organising capacities seems to apply to the practical
field as well. Attempts to develop "Netiquette" as a disciplined
way to deal with mails, "cc" and attachments seem to deliver
evidence of similar volume and relevance problems than in the academic
field. Jeffrey Pfeffer's and Sutton's work on the talk-do-gap or smart-talk
trap finally seems to point to the same conflict between rapid attention
switching and superficial grasp of diverse issues and their concentrated
and substantial elaboration [Pfeffer and Sutton 1999].
Does the third group deliver support and relief to the abundance of
less structured information* or does it contribute to the problem? The
answer is: Both. Faster searching machines, automated indexing, common
platforms to integrate data* of different format and operating system origin,
filters and intelligent agents can be counted on the relief side.
[2]A study, undertaken at the University
of Berkeley, demonstrates for 1999, that the number of mails and documents
does not only abound in absolute figures but grows at a rate of over 100
% a year (Regents of the University of California 2000).
Proprietary standards (as in e-learning) and the general contribution
to alleviating the difficulty in storing, processing and diffusing data*
contribute to the problem. In conclusion of this prologue I will explore
the following propositions:
- Incentive systems in Western cultures reinforce knowledge creation
(even if faked) more than knowledge exploitation.
- The ease in storing, processing and diffusing data* electronically
has unintentionally led to a deterioration of human capabilities to capture,
structure and understand its immanent knowledge potential and has increased
the sheer volume of potential knowledge to a degree that overburdens decision
makers.
- The contribution of information and communication technology in solving
the challenge of unorganised abundance is constantly counteracted by its
potential to aggravate that challenge.
Together these three propositions form the challenge of information
overkill. Knowledge management thus turns into the search for knowledge
nuggets within an abundance of data-sand.
Humans who are supposed to restrict their knowledge work to those few
fields where they are superior to computers, namely coping with incompleteness,
ambiguity and uncertainty, may lose exactly those capabilities while drowning
in "purposeless information".
2 The information overload challenge
2.1 Effects of information overload on people's time management, self-esteem
and decision making
As demonstrated by the Berkeley study, information abounds globally.
In the US employees surf on average one hour daily on the net, with less
than 20 percent of that time considered to be directly related to work
[Jackson 1998]. In an Austrian study among academicians
respondents claimed to spend about 90 additional minutes a day to read
and write emails, after they had learnt to ignore most of what they receive
[Students' Group 2001].
Let's now put the argument of not reinventing the wheel at its extreme:
If a researcher were to include everything ever published on her subject
historically and globally she would either have to push her subject to
the extremest niche or end up re-appraising for several years what is already
out there. As the latter is neither inspiring nor conforms to the way we
learn the world (that is by rediscovering) procedures have been developed
to rationalise the exercise. So the researcher will only refer to a few
mainstream and recent sources and select only those few aspects that help
him make his point. Still, the official rule remains that researchers,
standing on the shoulders of giants, as Newton said, are supposed to take
existing knowledge into account while contributing invention and new thoughts.
Growing older they learn that they are safe if they comply to referencing
some salient works that become salient exactly because they fulfil the
function as a substitute for profound search and study. In psychology such
a split between formal and informal rules is called a double-bind.
Double-bind situations in early childhood undermine people's self-esteem
which in turn undermines their ability to cope with authority in a sound
way.
We can recognise that - put to the extreme - abundant information undermines
democracy which is very much in contrast to common wisdom and the benefits
promised by proponents of a "better e-society".
Similarly, if in the practical field we became really serious about
the underusage of existing knowledge* and reconsidered every idea, every
concept and every project that have ever popped up, the reassessment of
best practices and benchmarks would last beyond the time any strategic
window were open on the market. Economically we can easily come up with
an equation-based model to solve this double-bind that is based on speedy
action and sound research before action is undertaken: If the cost-benefit
ratio of knowledge generation is better than the cost-benefit ratio of
knowledge exploitation than they should generate, otherwise they should
exploit. The problem is, that in most cases we don't know which numbers
to fill into our equation so that practitioners are left without reliable
clues how to act. Again schizophrenia raises feelings of insecurity and
makes people more prone to manipulation.
One has to think to the extreme to test arguments, but in everyday life
extremes happen rarely. What happens is a general disequilibrium between
the amount of information that is relevant to any task and is accessible,
although at some transaction cost, and the amount of knowledge anyone is
able to process thoroughly. Let's illustrate this point with the matter
of knowledge management. My estimation is that at least two conferences
are offered a day globally, books and articles abound, so that nobody has
the slightest chance to cover the matter fully. How do people react to
such an unbalanced reality? They start to work longer hours [Reich
2001] and doctors report that many suffer from stress or burnout which
are counterproductive to innovative and independent thinking.
As a second effect of information overload we can therefore state its
function to make innovation less probable- again a contra-intuitive finding,
as common wisdom would expect more access to information to result in more
combinatory innovation.
As Dörner demonstrates in his research there is no linear relationship
between the quality of a decision and the quantity of information available
to take it. Economic theory inclines us to accumulate information to take
better decisions if we can gather it at low cost as homo oeconomicus calculates
his best alternative based on complete information. But, Dörner has
found out that the quality of decisions taken by his test persons - all
holding degrees and top leadership positions - deteriorated progressively
as they were exposed to information overload and time pressure. They tended
to fall back into simplifications, black and white picturing of situations
and random filtering of information [Dörner 1998].
The "solutions" preferred in situations of information overload
combined with pressure to act on urgencies tended to be authoritarian,
even violent or childish.
Damasio, in his book on Descartes' Error makes the point that humans
use their emotions and intuitions to survive. If the links between their
emotional and cognitive centres are cut by some accident, the individual
remains able to score high on intelligence tests but must be taken under
custody because (s)he is unable to take any decision out of a nearly infinite
realm of opportunities [Damasio 1997, 108ff].
As a third effect of information overload we can therefore adhere to
the fact that it deteriorates the quality of decision making, processes
as well as outcomes.
2.2 Remedies for information overload
So far our analysis has led us to the conclusion that humans can improve
on their work and lives by having access to information in principle, but
that this positive effect is counteracted by information overload. So what
knowledge management should do is to find remedies for this problem. Overload
can, to be exact, only be defined with reference to a certain context and
a certain individual as abilities to digest differ between people and situations.
Nevertheless, globalisation and digitalisation combined with higher education
of larger parts of world population have led to a situation of structural
abundance: No one can process all the information that is accessible and
relevant to a current task.
Remedies for overload can be divided into three categories: technical,
psychological and neurological.
Technical remedies are devices that filter information to exclude irrelevant
material, that automate retrieval and indexing (intelligent agents) or
that support the recognition of patterns and causal relations in huge amounts
of seemingly unrelated data (data-mining). They build on the superiority
of computers in processing high volumes at high speed. If supplied with
criteria and set rules, computers can take on large parts of knowledge
work, such as structuring, compiling, condensing and comparing data. Office
programs offer classical functions of time and meeting management, of organising
daily tasks and keeping files of addresses, links and task-specific documents.
But, the best system is only as good as the individual using it. Therefore
progress in productivity lags behind its technical potential. But on the
whole we could claim, that we are about to witness a revolution in enhanced
productivity, resulting from the multiplication of growing computing power
with growing connective capacity, growing and cheap storage capacity and
open software standards, such as XML for the time being [Brown
2002]. My guess, though, is that growth in volume of outcome, made
possible by such progress, combined with a decline in human capacity and
discipline due to the ease in processing, storing and diffusing information,
will outgrow the potential technology offers to solve the overload challenge.
Technical progress in the past has usually consumed about one third to
one half of its potential to reduce complexity in order to cope with the
complexity inherent to itself.
Psychological remedies address the way humans cope with structural overload.
Abilities that are less trained nowadays because computers make life easy,
have become even more important: The ability to set priorities and to know
what one wants, the ability to understand the essence of an issue and the
ability to know what one does not know (enlightened ignorance) and to know
what one does not need to know (positive ignorance). To cope with an abundance
of choices, easy to access, requires strong will, however, will is weakened
by the mere existence of abundant choices. Problem solving and pattern
recognition abilities need to be trained from early childhood while our
educational systems still put a premium on reproducing information. Positive
ignorance may be the toughest issue here as we still try to live up to
the vision of the "Renaissance Man" who knows all about her world.
It must not be tackled mechanically as serendipity is to be considered.
Serendipity means openness to seemingly irrelevant pieces of information
that may, by chance, fall into a highly relevant picture later on.
One way to cope with overload from a learning perspective is to train
people to withstand the seduction of volume, another would be to train
their capacity to digest more. The first way focuses on highly effective
use of information by keeping the old virtues of thinking alive. Unfortunately
this is counteracted by industry that distracts consumers by a plenitude
of attention-capturing options that consume time needed to think and to
learn. The second way follows the philosophy of adapting humans to the
new technical environment rather than the other way round. Courses in fast
reading, in attention switching and in memorizing shall increase the ability
to process large volumes of information. As with technical remedies I expect
this type of solution to follow the pattern depicted in the old fairy tale
of the hedgehogs and the rabbit: They will always lag behind.
A third category of remedies targets the interface of human and artificial
intelligence. Human brains will be extended by computing power added to
all kinds of devices, attached to cloths, glasses, watches or even directly
to the cortex. If by 2009 a computer, priced at 1000 $, can process one
trillion operations per second humans will be "relieved" of many
procedures we used to call thinking by then. By 2029 a 1000 $ computer
will have the power of 1000 human brains: It will have processed the whole
body of human literature, fine art and natural science and most probably
have a consciousness [Kurzweil 2001]. I will not
speculate too far into the future but we will have to deal with some deeply
philosophical questions then.
Again, a biological alternative is at hand to complement or substitute
for artificial intelligence. By training and revoking different ancient
and traditional practices to enter other states of mind that transcend
the wave length and frequencies our brain operates in its normal state,
we may learn faster as is assumed in superlearning and understand faster
as reported about states of trance. If we compare the amount of research
money devoted to technical progress and to rediscover old wisdom that is
considered to be esoteric, it becomes quite easy to predict the future.
Kurzweil's spiritual machines will appear with a much higher probability
than humans with a natural gift of Schamanism to put a long story short.
3 Conclusion
In this paper I have argued that scholars and practitioners of knowledge
management have, for some incomprehensible reason, neglected a phenomenon
that threatens to counteract their progress, namely information overload.
They have not only neglected this phenomenon but contribute to it by
following a linear vision of accumulating ever more information to improve
the productivity of knowledge work. I have pointed to empirical evidence
and at the same time argued the case of structural conflict caused by abundance
of information. Both points render the concept of better work based on
ever more information obsolete. Under conditions of information overload
and computer support in thinking.
- Humans will regress in their capacity to think as it is acquired in
evolutionary processes of learning and preserved in constant use. The assumed
high and typically human level of knowledge work, namely dealing with uncertainty,
ambiguity and incompleteness, may never be reached by any human if a future
kid's environment does not offer any training in basic cognitive functions
that will be taken over by computers.
- They will take worse decisions with regard to desired outcome as evidenced
by Dörner's experimental settings and Damasio's neurological cases.
- They will easily fall prey to all kinds of political and commercial
manipulation as their lives will be constantly stressed by the gap between
the amount of information that is relevant and in principle accessible
and the amount they are able to digest. In our contemporary society which
blames this structural gap on individuals and the educational system, low
self-esteem of the non-elites may result and thus threaten democracy.
Remedies to cope with overload while benefiting from the promises of
a knowledge society can be categorised into technical, psychological and
neurological.
I have excluded a scenario of successful neo-Ludditism where the progress
of computer, connective, storage and A.I. capacity becomes stagnant.
I expect a combination of technical and affirmative psychological procedures
to gain market share as they follow the self reproducing dynamic of markets
in general: Each solution breeds a multitude of new problems that require
entrepreneurs who invent new solutions that will again breed problems.
The most probable scenario is not necessarily the most reasonable one.
If we spent as much money on humans as we do spend on their technical
support that may finally result in their substitution or - as Kurzweil
says, in their co-existence with higher intelligence just like animals
co-exist with us today - we may achieve results that are cheaper and superior
even in terms of productivity and most certainly in terms of ethical standards
that are so far bound to biological entities called humans.
References
[Brown/Seely (02)] Brown/Seely, J.: "An Epistemological
Perspective on Organizations Afford Knowing." Keynote Address to The
Third European Conference on Organizational Knowledge Learning and Capabilities,
April 6th, Athens (2002)
[Damasco (97)] Damasio, Antonio, R.: "Descartes
Irrtum. Fühlen, Denken und das menschliche Gehirn." dtv, München,
5. Auflage (2000)
[Despres (01)] Despres, Charles/Chanvel, Daniele:
"The Thinking Behind the Action in Knowledge Management." Proc.
of the Second European Conference on Knowledge Management, Bled (Slovenia),
8 - 9 November 2001 ; pp 133 - 152.
[Dörner (98)] Dörner, Dietrich: "Die
Logik des Misslingens. Strategisches Denken in komplexen Situationen."
Rowohlt, Reinbek bei Hamburg (1998).
[Jackson (98)] Jackson, Maggie: "E-mail just
the latest way workers keep in touch." Associated Press, Compiled
in: Cremers R., strategic-news@olsy.nl,
23.5.1998
[Kurzweil (02)] Kurzweil, Ray: "Homo S@piens.
Leben im 21. Jahrhundert. Was bleibt vom Menschen?" (Original 1999:
The Age of spiritual Machines) Econ München, 3. Auflage (2002).
[Pfeffer (99)] Pfeffer, J./Sutton, R.: "The
Smart-Talk Trap." Harvard Business Review, May-June 1999, pp. 135
- 142.
[Regents of the University of California (00)] Regents
of the University of California (Ed.): "How much information?"
www.sims.berkeley.edu/how-much-info/summary.htm,
downloaded: 28.11.2000
[Reich (01)] Reich, Robert: "The Future of
Success" Knopf
[Schneider (01)] Schneider, Ursula: "Die
7 Todsünden im Wissensmanagement." FAZ Verlag, Frankfurt (2001)
[Students' Group (01)] Students' Group, University
of Linz: Seminar Paper on "Individual Strategies to use e-mail in
the Academic World - an empirical study", Linz 2001
* As all work on knowledge management this paper will be exposed to
the difficulty of talking about concepts that lack clear
definition. In the following I will distinguish between "data" (texts,
images) as observables that turn into information if related to a
context and therefore endowed with meaning. Assuming a common context
of global society I will therefore often refer to information on the
net and thus deviate from Bateson's definition that information is
subjective, namely a difference that makes a difference (which I
highly appreciate). Everything stored on a medium, be it paper or
electronic, will not be conceptualised as knowledge in this paper,
which uses the concept of knowledge as related to a consciousness and
assumes that so far consciousness only exists in humans. Structures on
media, be they texts, images, formulae, therefore are only potential
knowledge. They turn into the latter if a human brain processes them.
|