## Narrative and Memory

From: andrew cooke <andrew@...>

Date: Sun, 28 Jul 2013 17:52:46 -0400

[This is about software development, not MS or Literature...]

Recovery from an MS outbreak is very slow - I am still getting better after 8
months.  It's also intermittent and inconsistent - there are times when things
get worse again.

When recovery stops, some symptoms are likely to remain (and be variable).  So
it becomes important (at least for my peace of mind) to know whether things
are improving, or whether you have reached a final, noisy equilibrium.

You might think this is easy to track.  Simply remember (or write down) how
you feel at various points in time, and compare.  But that doesn't work very
well.  There are many small details, and it's unclear in advance which will be
important, and which not.  So you simply don't know what to record, while
memories don't have sufficient resolution.

Despite these problems I recently found a reliable argument for the
improvement of my right hand.  A few months ago I had a good explanation for
how my hand felt - as thought I were wearing a tight leather glove.  And that
explanation was closely connected to a restrictive feeling across the knuckles
as I clenched a fist.  Then, a month or so back, that feeling changed.  At
first I only knew that it felt "different", but after some time I realised
that the restrictive feeling had moved from my knuckles to the second finger
joint.  And in the last few days I have felt that change again; the
restriction is now "inside" my fingers (on the palm side) and felt mainly when
I try to unclench / flatten my hand.

Such a coherent "story" of improvement is very rare.  Normally I have to be
satisfied with a vague impression - "it seems better than it was".  And as far
as I can tell it came from the lucky accident of the glove metaphor, which
focused my attention on the restrictive feeling on the outside (and then
inside) of my finger joints.

Without that "story" - the structure of looking in a certain way - I doubt I
would have remembered exactly which part of my body was restricted when.
Things ache all over the damn place.  It was only chance that made see the
changes fit a nice, linear progression.

What I learned from this was that my memory was very much shaped by the
interpretation I used at the time.  The image of a glove structured my memory
in a way that fitted nicely with my recovery.

The last few years I have worked - on and off - on a rather unpleasant
project.  The development of a complex piece of software that has been plagued
by problems and delays.  To "celebrate", when it finally shipped, I gave a
talk at work, where I described the system and tried to draw some lessons from
the experience.

It wasn't a very good talk.  I struggled to find coherent lessons.  And later,
in the discussion that followed, I realised that other people on the same
project had very different memories, and very different conclusions.  Everyone
was unhappy, but the explanations, and supporting evidence, varied widely.

The connection with my hand seemed clear.  People use narratives to structure
their memories.  Then their memories selectively support those narratives.
It's hard for me to be sympathetic to someone else's explanation of what went
wrong, because I have already structured my memory around my own ideas.

Last week I started a follow-up to the same project.  Immediately I was hit by
a slew of familiar issues.  My stomach sank as I realised that we were going
to experience exactly the same frustration and disappointment as before.  We
had learned no lessons, and improved nothing.

So on Friday I wrote a short report - the first of what is planned as a series
of weekly summaries - that describe the main problems.  I am deliberately
trying not to interpret what is happening.  They are just basic, factual
summaries: X was inconsistent with Y; Z failed, etc etc.

My hope is that at the end of this project we can use them as a more neutral
memory.  And learn, and avoid.

[You might ask why this didn't already exist.  What we have is a daily written
report and weekly oral reports.  The daily written reports were too low-level
and numerous.  The weekly oral reports weren't recorded and have anyway lost
support from management.

Andrew