What's Missing From VPEC-T?

[Richard Veryard]

There have been several discussions in the VPEC-T Google Group
identifying possible additions to the VPEC-T lens. Some months ago,
partly in order to find out what kind of lens VPEC-T actually was, but
partly because I actually think it's important, I suggested adding the
letter M for Meaning, and now Roy Grubb has just started a thread
suggesting the letter O for Outcomes. Previously, Sally Bean asked
about Roles/People/Community. (Have I missed any other suggestions?)

One response to this has been to claim that any of these are somehow
already implicit in the original five letters. I don't actually think
this is a very good response. I might as well argue that Trust is
implicit in the Kipling-Zachman lens (in other words, implicit in the
use of the lens by those practitioners I regard as using it
"properly") and nobody could ever prove me wrong. The point of calling
it VPEC-T is to draw attention to these five concepts in particular
(or six if we count the dash as a concept).

The next response is to challenge the notion of completeness.
Obviously there is an indefinite number of concepts we might want to
pay attention to. There are many other lenses (Kipling-Zachman,
CATWOE, SWOT, and so on), and each of these focuses attention on a
different set of concepts. I don't think it makes sense to try and
produce a single all-purpose lens that covers everything, and I don't
see why adding lots of other stuff to VPEC-T would be an improvement.
We can agree with Sally that Roles/People/Community are important
without necessarily concluding that these concepts must somehow be
bundled into VPEC-T. A better solution might be to select another lens
that provides particular insight into Roles/People/Community, and use
this lens in combination with VPEC-T.

However, I think it still makes sense to talk about ways to improve
the VPEC-T lens, provided this doesn't just mean making it more
complicated and more likely to lure people into a false sense of
completeness. One of the reasons I like VPEC-T is that it draws our
attention to the complexity associated with each of the five letters.
Many traditional methodologies have a place for a simple unified
statement of value (goal or objective); but VPEC-T tells us about many
interacting and often conflicting value systems (V), many interacting
and inconsistent policies (P), and so on. One reason I wanted to add M
for Meaning (and the same argument might also apply to Sally's or
Roy's suggestion) was that I see Meaning as the same kind of concept -
we are faced not with a single unified meaning emerging from the
analysis (which I think was Nigel Green's argument) but with a
multiplicity of meanings - each event or policy can mean something
different to each Role/Person/Community.

To take the Dinner Party example from the Lost in Translation book,
each of the things that happens at a dinner party can be interpreted
and experienced in different ways. Policy dictates whether the guest
should offer to help with the washing up, but Meaning tells you
whether people actually feel grateful or threatened when this occurs.

I also think Meaning is a key element of change management. We frame
change proposals so that they can mean something positive for each of
the stakeholders. (I hate all this talk about "overcoming resistance",
but that's another discussion.) And we create Meaning by telling
stories. So one way or another, I am determined to have a lens that
helps me think about Meaning.

As far as I can see, I have three possible options. My first option is
to use a completely different lens for understanding Meaning. My
second option is to create my own idiosyncratic version of VPEC-T,
which I might call VPEC-TM. (Meanwhile, Roy creates his own version,
and the VPEC-T community fragments before it has even really formed.)
My third option is to persuade the VPEC-T community to make a
collective change to VPEC-T.

This is of course raising a question about the VPEC-T community. Now
Sally, what lens do you think we should use to think about this
question of Role/People/Community?

[Sally Bean]

Richard,
Here are some candidate lenses. I'm thinking more here of domain
analysis/general problem solving that may well be broader in scope than
information system analysis.

Firstly, you've already alluded to CATWOE/SSM, which identifies customer
(beneficiary), actor and owner roles. This is enriched in the 'developed
form' of the SSM, described in Checkland and Scholes book of 1990, which
includes something called Analysis 1/2/3:
Analysis 1 is the analysis of an intervention itself - who caused the
intervention to happen, who is conducting the intervention and who is
affected by the outcome of the intervention.
Analysis 2 covers Roles, Norms and Values, which overlaps with VPEC-T.
(Interestingly roles could include things like 'trouble-maker' and 'joker'
as well as functional roles)
Analysis 3 covers Power relationships,

Secondly you could draw Rich Pictures or Influence diagrams
See here for examples.
http://systems.open.ac.uk/materials/t552/index.htm


Thirdly, there is Verna Allee's Value Network Analysis, which I've not
actually used myself, but am interested in learning more about.

When I've discussed roles with Nigel Green, he assures me that VPEC-T
analysis does cover roles adequately, but they are not a primary feature. I
can also see that there is a danger that you may perpetuate current
organisational structures that might be better redesigned, if you take them
as a starting point of your analysis. However, I usually like to start an
investigation by looking at the stakeholders involved, if only to understand
the politics and make sure that the right people are consulted and
communicated with.
Sally

[nigelp...@googlemail.com]

Oh no, not this again and while I'm too busy to craft a pithy
response :-)

The fag-packet response:

* I don't want to add to the core dimensions (columns) unless there's
a massively compelling reason (I'm not yet convinced)
* Adding new columns will unhelpfully constrain use and/or make things
confusing
* The additions suggested can all be dealt with as 'rows' or Cross-
cutting Concerns/Subject of the VPEC-T analysis
* Suggested additions can be described in row/col Use Patterns
(blogged about this already) - these UPs would have the effect of
'extending' the framework for a given context
* I do such extensions each rime I use VPEC-T - typically white board
with Topic being looked at.VPEC-T columns, topic related rows (CcCs)
and a flip chart with findings e.g. outputs, concerns, tension-points,
ideas, principles, actions et al. I've not yet experienced a problem
with 'missing' dimensions using such an approach. However I'm keen to
hear of a real example where a VPEC-T analysis was found wanting
(assuming a similar approach is followed).

My two-pennies,

Nigel

[Richard Veryard]

Nigel finds this discussion unwelcome ("Oh no, not this again"), and
is strongly resistant to adding any columns, but he would be willing
to accept additional rows / cross-cutting concerns. For my part, I
already find the rows more confusing than the columns, so I'm not keen
to see any further complication added there.

But my original question "What's missing from VPEC-T?" can be
understood as a rhetorical one. Maybe the answer is "Nothing".
However, we cannot expect people to take this on trust; we must expect
people to want to work this through for themselves. (I'm seeing future
users of VPEC-T as CUSTOMERS.)

As I see it, Nigel's practical use of VPEC-T is based on a policy of
extension - anything that comes up (EVENT) during a VPEC-T exercise
can be handled as another row (TEMPORARY CONTENT) - and an aspiration
(value system, worldview) of using VPEC-T for everything while keeping
VPEC-T itself (CORE CONTENT) unchanged. Am I right?

For me though, the main reason why I can accept the answer "Nothing"
is because I don't expect VPEC-T to be the only lens I'm using. So I
think I'm more likely to use another lens in combination with VPEC-T
than to add an adhoc extension for a particular situation. And I'd be
happy to drop some of the more confusing aspects of the cross-cutting
concerns, if it makes VPEC-T easier to adopt, learn and use.

Meanwhile, I regard the act of reflecting about what we are doing as
extremely important, an essential characteristic of the kind of
systems thinking I wish to encourage. We must constantly keep half an
eye on the lenses themselves, so this kind of discussion is not going
to go away. Sorry Nigel.

[nigelp...@googlemail.com]

-------- Forwarded message ----------
From: Tom Graves <t...@tetradian.com>
Date: Fri, Sep 25, 2009 at 10:34 AM
Subject: What's Missing from VPEC-T?
To: "nigelpsgr...@googlemail.com" <nigelpsgr...@googlemail.com>

Hi Nigel

(Apologies, seem to have lost [or never properly set up] my membership
to
VPEC-T group, so will have to email you this response instead.
Following is
for posting to the group: )

This sounds much like the discussion about 'missing dimensions' in
Zachman,
with people wanting to add columns such as 'interface' and the like.

I'm with Nigel here: I would strongly recommend leaving the basic
structure
of VPEC-T alone. There's an interesting discussion about the role of
the '-'
that might be worth revisiting, but otherwise that should be it.

'Meaning' and 'Outcome' are definitely relevant, but it's misleading
to try
to kludge them on to the existing structure. If we think recursively
as well
as iteratively, 'Outcome' is close to "that which is valued" - i.e. a
Value
- at a higher level of abstraction; 'Meaning' is in effect close to a
kind
of outcome, though in a perhaps different sense.

Thinking in Zachman terms, the VPEC-T lenses are equivalent to
Zachman's
columns. They're not dimensions as such, but a distinct set of
'primitives'
that can not be resolved into each other, yet may also exist as
members of
'composite' or layered composite (e.g. Nigel's 'Use Patterns'). In
Zachman's
case, the original columns had several fundamental taxonomic flaws,
some of
which could only be resolved by adding what truly _is_ another
dimension of
'asset-type' (see http://tetradianbooks.com/2008/12/silos-frame-ref/
). In
VPEC-T I believe the columns are already valid and 'complete-enough':
it may
be useful to add 'Meaning' or 'Outcome' and the like as backplanes
(i.e.
true intersecting dimensions - e.g. 'meaning of value', meaning of
event',
'outcome of process', outcome of trust' etc), but should not be added
to the
column-set itself.

Might also be useful to compare VPEC-T usage with Zachman's rows. The
latter
imply different sets of stakeholders, in layers of abstraction from
strategy
and above to real-time operations and their performance-outcomes. A
recursive as well as iterative usage of VPEC-T might help in this,
too.

Hope this helps, anyway.

Best etc
- tom g.

--
---------------------
Principal Consultant
Tetradian

[SallyB]

Tom has already drawn a useful parallel with the Zachman Framework
which had also occurred to me. And I agree with the point Richard
makes about using lenses in combination. What I typically do for a
piece of work is to take ideas from different methods and write down a
plan of how I'm going to put them together.

I also see an emergent pattern which I discussed with Nigel earlier
this morning.

I think there is a pattern of tension between the inventor of an idea
or methodology and the emergent community of 'secondary practice' -
those people who are trying to apply the principles without the
originator being directly involved in their practice. As Richard
points out, people have to work through their interpretation of the
idea and see how it fits in with their existing repertoire. (I think
Tom blogged something about a learning process recently?). They are
bound to ask 'difficult' questions, and it can be tough for inventors
to have to keep answering the same questions, or bat away challenges
to their thinking from people with a different perspective and
background, especially when the inventors are busy people who have
achieved a level of 'unconscious competence' in the application of
their idea.

This 'going around in circles' effect is also a problem even when
there isnt a 'guru'. Our architecture team in BA blocked off Thursday
mornings for reflection and knowledge-sharing. We did find that some
topics did keep resurfacing, and there is an art to knowing just how
much time to devote to such debates. Too little, and they will come
back to bite you later. Too much and you will disappear down
theoretical rabbit holes and never get any proper work done!

Sally

> But my original question "What's missing from VPEC-T?" can be
> understood as a rhetorical one. Maybe the answer is "Nothing".
> However, we cannot expect people to take this on trust; we must expect
> people to want to work this through for themselves. (I'm seeing future
> users of VPEC-T as CUSTOMERS.)

>.........

[SallyB]

Following my comment here earlier today about the tension between idea
creators and secondary practitioners, I've just noted this blog on a
similar theme by Dave Snowden on the increasing take-up of his ideas.
http://www.cognitive-edge.com/blogs/dave/2009/09/purity_danger.php