denotes premium content | Dec 3 2008 Feature
posted 1 Jun 2000 in Volume 3 Issue 9
Organic Knowledge Management - Part
Two
Knowledge Elicitation: Indirect knowledge discovery
In the first in this series of
articles on Organic Knowledge Management, David Snowden identified key issues
relating to the language we use to describe what we know. In this second piece,
he examines some of the dangers of knowledge elicitation and outlines approaches
routed in anthropology that enable the identification of knowledge assets in
organisations.
'There are more things in heaven and earth, Horatio, Than are dreamt of
in your philosophy.'
Shakespeare's Hamlet
The first of this series of articles
described a linguistic framework, designed to enable organisations to identify
what they know in such a way as to move directly to action as a result of the
description: The ASHEN model. The principle of the model was, first, to identify
Knowledge Disclosure Points (KDPs) in the form of decisions, judgements, problem
resolutions and learning events; then, to ask a meaningful question within the
context of those KDPs. For example, when you made that decision, what Artefacts
did you use or have access to? What Skills had you acquired that were necessary?
What Heuristics have you developed that enable you to make that decision quickly
on the basis of incomplete and/or unarticulated inputs? What Experiences have
you had that were essential, or just plain useful, in making that decision? What
Natural Talent is necessary and can you give examples of signs that such talent
exists as potential in others? ASHEN provides for a meaningful question in a
meaningful context.
This second article will explore the need for indirect approaches
to knowledge audits, based on the discovery of KDPs through techniques drawn
from anthropology. In the final article, these techniques will be extended to
cover Story Circles and, more speculatively, the use of anonymity in
virtual communities. The final article will also summarise a series of heuristics
for use in designing 'interventions' in the knowledge and learning ecologies that
are a necessity in the modern organisation.
The need for Knowledge
Elicitation(1)
In the first of these articles I pointed to the self-evident truth
that much valuable knowledge is triggered in context: It is known when it is
needed to be known. This means that we either have to be present at the time of
that need and observe its identification and use, or we need to recreate the use
in retrospect. This requires us to take account of one of the great truths
of knowledge management, derived from a saying of Peter Drucker at Delphi's
IKMS event in San Diego, 1998: 'In the knowledge economy, everyone is a volunteer,
but we have trained our managers to manage conscripts.' Elicitation of knowledge
requires a paranoid attention to the conditions of a volunteer society,
otherwise our investigation will be adversely influenced by camouflage and
conformity within the communities we are studying, and we may fail to understand
the motivation or nature of knowledge retention or hoarding.
Both of the above behaviours are part of survival in a large
organisation and, to a lesser or greater degree, all employees learn to adopt
both strategies at need. In addition, we see knowledge being deliberately
hoarded, sometimes because it presents an opportunity to exercise power, but
more frequently because of fear of abuse of something precious, or simply
through lack of time coupled with excessive and unreasonable demands for
codification(2).
For these reasons and many others, it is vital to avoid influencing the
subject area, as such influence will trigger one of the above problems.
Knowledge elicitation is an indirect process, developed because the traditional
techniques of structured interview, questionnaire and hypothesis-based
consulting have little application in the early stages of a knowledge audit
project, although they are useful once we have worked sufficiently with the
ecology to allow its basic patterns and structures to emerge.
Beyond the Newtonian
metaphor
Underlying all of the above is a fundamental statement about the nature
of knowledge management and associated consultancy. We can see the development
of management science as a linear development, commencing with the adoption of
the principles of Newtonian physics by Fredrick Taylor at the start of the 20th
Century, and culminating in the work of Hamner and others on process
re-engineering. The underlying metaphor was of the organisation as a machine, in
which cause and effect could be separated and understood through the application
of scientific method. The Harry Seldon character of Asimov's Foundation Trilogy
in science fiction probably summarises best the aspiration of this approach: To
develop algorithms that would predict human behaviour in the same way as the
movement of the heavenly spheres could be predicted. The argument of this and
other articles covering Organic Knowledge Management is that the mechanical
metaphor of Newtonian physics fails when we move to the management of
intellectual capital; it is fine for process, quality and any other activity
that is mechanical in nature, but it fails when the principle component or actor
is organic.
In
the realm of physics, scientists realised that an electron can paradoxically be
both a wave and particle simultaneously. They also became aware that the energy
required to see an electron fundamentally changed its position, so you could
never really see the thing itself, but had to look for its traces. In management
science, we are starting to make the same transition. Knowledge management has
been one of the trigger points in this transition. Managing knowledge has a
significant requirement to understand ephemeral natures; the act of knowing is
itself interdependent with a variety of causal factors and there is no clear
distinction between cause and effect. Knowledge is simultaneously and
paradoxically a thing and a capability, an actuality and a potential, tangible
and intangible. Managing knowledge is more akin to managing a complex ecology of
interdependent, unpredictable and fluid entities, than it is to designing and
maintaining a sophisticated machine.
It follows that a knowledge
elicitation exercise needs to avoid influencing the ecology that it is studying.
The formal hypothesis-based techniques of much internal and external consultancy
are inappropriate for knowledge work. Aside from the directive nature of such
enquiry, the very arrival of a consultancy team engenders a reaction in the
ecology. The need to overcome these issues led to the development of the
anthropologically-based technique outlined below; a recognition of the
limitations of that technique to examine the past led to the develop of Story
and other methods which will be outlined in the final article of this
series.
As with
an ecology, the nature of what has happened before profoundly influences what is
now possible. The population may be cynical or just plain weary of new ideas and
initiatives, coupled with new drives for economies and revenue in an ever more
competitive environment.
Anthropological observation
When an anthropologist
studies a community they first become, as far as is possible, a part of
that community. They seek to observe as an unobserved observer before they form a
hypothesis. In contrast, too many consultants draw down a method from their
internal knowledge systems and seek to shoehorn the situation into that model - the economics
of which are those of production line - you can have any solution so long as its
black! Anthropological techniques are gaining increasing credibility in business
and do not require long periods of cultural emersion to be effective. The secret
to this sort of work is not to enter in as a consultant wearing a suit with or
without clipboard, and certainly not with a pre-prepared set of questions and
hypotheses to test. Instead, the consultant enters as a servant of the servants.
Stack vegetables, carry bags, make tea, sweep metal scarf off the factory floor
with the apprentices, dig holes in the road, and shadow office workers. You may
not even use consultants; in one successful project school children were used to
understand the knowledge flows of a head office. The stated reason for their
presence was a half-term work experience, but the day before they started they
were trained by an anthropologist in observation techniques. The school children
exhibited three characteristics of successful operators in the field. They were
naïve, so they asked unexpected questions; they were innocent and as a result
people answered those questions without taking offence and, finally, they had
all the curiosity of youth, and genuine curiosity naturally leads to higher
levels of elicitation.
Observation of this type led to the discovery of the field
engineers café mentioned in the first article and the valuable knowledge artefact of
the journal. In a similar case, the subject of study was a supermarket in which
the management staff was exceptionally good at forecasting fresh products. This is
a real issue for a supermarket: Failure to have a full range on display as
the trading day draws to a close means loss of customers and loss of loyalty,
but nearly all items unsold at the end of trading are a loss. No-one (including
the staff) knew why this particular supermarket was so good. Several days of
patient observation, tea making, trolley pushing and shelf stacking resulted in a
more relaxed environment and then the reason emerged in the context of need:
The individuals knew what they knew when they needed to know it. One
Thursday evening, demand started to soar to exceptional levels. Given that the
normal weekly peak is Friday/Saturday this was anomalous, and pressure grew to
increase the order quantities for the next two days. The management team was not so
sure, and at one point in the discussion, one member suddenly reached into a
drawer and pulled out a much-battered five-year diary. While the others were
talking she thumbed through it apparently without purpose. The next minute she emitted
a quick expletive followed by the statement: 'We've got a Baker Day and there's
a home match on Saturday.' For those not familiar with the phrase it is
a reference to an ex-British Minister for Education, who granted teachers
the right to call occasional holidays for training purposes. Such one-day
events have entered the vocabulary as 'Baker Days' and for working parents they are a
problem. The Baker Day, coupled with a home match for the local soccer team,
meant that weekend shopping was taking place early to allow visits to the beach,
preparation for the absence of the family at football on Saturday, or a like
activity. As a result of this, the team in question lowered their order rather
than increased it, and in consequence had less wastage. It turned out that the
five-year diary was used to record any exceptional or unusual event and the then
best understanding of the reasons why. The human being was able to scan this
written record and identify recurrences and/or similarities with past events and
use them to help interpret the present situation. A very simple mechanism, very
easy to replicate over a large number of stores with little cost, and with the
advantage that the book contains more valuable knowledge than is ever used in
situ, so its capture (by copying) can provide useful information to the
organisation over and above its use for forecast accuracy.
Now consider, in this case there was
no indication of hoarding behaviour, but no amount of questioning, structured
interviews or the like had triggered the knowledge of the knowledge. It was only
the creation of a real need that triggered the use of the knowledge, and thereby
its elicitation. In ASHEN terms, this was an artefact of major use as it was
scalable at low cost. There were also associated skills, heuristics and
experience that could be managed through training programmes and cross-store
mentoring. Such work requires patience and the strength of will not to ask
directive questions that will influence the answer. Opinions can be offered, but
are only appropriate to provoke or stimulate a response and should be indirect,
preferably using a reference to another industry sector all together. Questions
of the form: 'I've noticed that in industry X, Y applies is that the same here?'
are acceptable. The more different that X is, the less intrusive the question
and the more revealing the answer. Not all are capable of doing this type of
work, as it requires a high level of personal tolerance for uncertainty. The
temptation of most consultants is to prove their industry knowledge, to add
value to the process. The observation stage is not the time for this. It also
requires a willingness to sacrifice the status of consultant; in a sense, the
more lowly the work, the better the impact and the more likely you are to find
something. One of the ways in which consultants can be tested for their
suitability is to give them humble tasks such as note taking on an assignment.
Those willing to learn will seize the opportunity; those who are incapable of
learning will demand a role more appropriate to their status.
The instruments of
observation work are tape recorders, video cameras and notebooks. One wants to
record the material as is, rather than filter it through the perception of the
observer. The note pad is for opinions and commentary. One of the main dangers
is that of over-identification with the subject. Working this way involves its
own type of camouflage to the point were the observer can, and almost always
does if they do a good job, go native to a degree.
Some observation will, by necessity,
come from an interview. In these cases, the secret is not to allow the subject
to know the precise nature of the enquiry. For those readers who enjoy spy
novels, you will know that one interrogation technique involves always asking
more questions after you have found what you were looking for. This means that
the subject does not learn from you and you are able to test or validate their
statements. Open questions, discursive questions, disruptive ideas and indirect
enquiry are the tools to apply. One useful technique is to always interview the
individual in the context of their work so that clues such as pictures on walls,
trophies and, the layout of the room, and even the attitude to interruption can
provide clues and opportunities for indirect questions. That way the context
creates material around which the subject can talk without directly tacking the
subject.
The
output of observation or the less effective interview process is threefold in
nature:
- Anecdotal material in the original words of the subjects being studied.
This is the core output, and subsequent listening to tapes allows the
consultant to identify knowledge disclosure points in the form of decisions,
problem resolution, acts of judgement, learning points, solution
identification and the like. A KDP is a focus point that satisfies two
criteria: Firstly, it must be meaningful to the subjects of the study; and
secondly, it must entail the use or acquisition of knowledge. The anecdotal
material can also be used as the base material for the creation of purposeful
stories; this will be discussed when we look at interventions.
- KDPs will also be directly noted during the observation. These should
be captured at the time. If they are clear and distinct then the ASHEN
question may be directly asked: 'What Artefacts, Skills, Heuristics, Experience and
Natural Talent were necessary when you made that decision?' However, this
should only be done when the knowledge is immediate and vivid. The primary
focus of this stage is to identify the various KDPs so that they can
subsequently be clustered and the ASHEN question be focused on individuals and
communities based on the totality of the observations. Again, one is seeking
to avoid biasing the sample at this stage as much as is possible.
- Finally, examples of the various ASHEN components will be encountered. A key rule for the observer is not to comment or to judge but simply to record. A knowledge asset is a knowledge asset; if you look pleased or are disturbed by something you will influence one-way or the other the subject. A fastidious expression of horror when working with engineers in a sewer will ensure more exposure rather than less to some stomach churning tasks! It's the nature of field communities to shock and to initiate. This is a part of being accepted within the community. In one case, the moment of acceptance came when the observer was sent to the local depot to pick up a 'Stand Alone' . Having stood alone in the depot for an hour or so, the individual was accepted. It was an initiation rite. To the observer, this is a sign that they are accepted and should alert them to the immediacy of obtaining a higher degree of confidence and consequently more exposure to what really happens. An over-reaction to such an event would invalidate it or lead to mischievous repetition. In some cases, observers had to be pulled because they lost respect through over-reaction.
Observation has its limitations as a technique. It is highly effective
where the life cycle of knowledge use can be observed over a period of one or
two weeks. Where knowledge use is revealed over longer time periods, there are
two other techniques that are valuable: Story Circles and Anonymity. These
techniques will be discussed in the next article in this series, along with the
process by which the various KPDs are clustered and ASHEN components gathered,
structured and used to create interventions to sustain and build the knowledge
ecology.
One final word of warning: In training and testing consultants in this sort
of work over several years, it would appear that those capable of carrying out
the work pick it up very quickly, whereas as others will struggle forever.
What has become known, as the 'deep end' test is a useful way of sorting sheep from
the goats. The candidate consultants are metaphorically taken to the deep end of
the municipal swimming pool and thrown into the deep end. You then go and have a
cup of coffee and read the paper - the ones who are still swimming when you come
back are worth working with. The reason for this is simple: A more conventional
approach means that it is a long time before you can identify who has 'got it'
and who hasn't. Most consultants are naturals at camouflage behaviour and
conformity; early exposure is less risk to the project and to the method.
References
1. In
the 1998 series of articles, I used the term Knowledge Disclosure. I have now
changed this to Knowledge Elicitation to reflect the need to be more proactive
in identifying KPDs. Although this is necessary, it carries with it an increased
danger of influencing the subject of study. While this is partly reduced through
the use of fiction and faction in Story Circles, it is still present, and
practitioners need to maintain constant vigilance
2. An expansion of the
reasons for Knowledge Retention is due to be published as an article in the
autumn. Advanced copies of material can be obtained on application to the author
David Snowden is European director of the
Institute for Knowledge Management. He runs regular public master classes in
Organic Knowledge Management. He can be contacted at:snowded@uk.ibm.com
