Pragmatic Thinking and Learning
Page 12
abstract. You are more likely to discover new abstract patterns,
which is what all of us programmers are trying to achieve.
This phenomenon of increased abstract awareness was demon-
strated in a study7 of secondary-school students who were given
the following problem: five meshing gears are arranged in a hori-
zontal line much like a row of quarters on a table. If you turn the
gear on the far left clockwise, what will the gear on the far right do?
Some students were selected to work alone, others in pairs, as the
researchers steadily increased the number of gears. By 131 gears,
it was easy to see who had discovered the abstract pattern (in
this case, the well-known computer-science parity rule) and who
7.
The Emergence of Abstract Representations in Dyad Problem Solving [Sch95].
Thanks to June Kim for this pointer and summary.
Report erratum
Prepared exclusively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
ENGAGE AN R-MODE TO L-MODE FLOW
98
hadn’t. Only 14 percent of the solo workers discovered the rule,
but a whopping 58 percent of the pairs did.
In another experiment, one pair of students came up with an
abstract matrix representation based on a very concrete problem
statement. The researchers reported the following:
...The experimenter asked the members how they came up with the
matrix. One member stated, ‘He wanted to make columns and I
wanted to make rows.’ To negotiate their two perspectives on the
problem, they managed to come up with the matrix formalism that
included both columns and rows.
Schwartz, et al.
Working together is a provably effective way to discover helpful and
interesting abstractions.
Meeting in Metaphor
As we’ve seen, L-mode and R-mode processing are radically dif-
ferent, and yet there may be a common meeting ground between
them in your own mind, a place where creativity gives birth to new
ideas.L-mode and R-mode meet in metaphor—in the act of creating
analogies.
“Metaphor, a common ground for both verbalizations and images,
is a way to voyage back and forth between the subconscious and
conscious, between right and left hemispheres.”8
The use of metaphor is a powerful technique to open up creativity.
TIP 14
Use metaphor as the meeting place betweenand.
Now, when you hear metaphor and analogy, you might have a
flashback to some horrid English class in grade school. But in fact,
we use metaphors all the time. What we call windows on the com-
puter screen aren’t really windows at all. The mouse isn’t actually a
rodent. A folder on your hard disk isn’t a real folder, and the trash
can isn’t a real can.
8.
See Conscious/Subconscious Interaction in a Creative Act [GP81].
Report erratum
Prepared exclusively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
ENGAGE AN R-MODE TO L-MODE FLOW
99
When you write a concurrent program using threads, you’re not
sewing. It’s just a metaphor. Don’t even get me started on zom-
bie processes in Unix or the typographic flotsam of widows and
orphans.
We use metaphors constantly; in fact, cognitive linguist George
Lakoff (Women, Fire, and Dangerous Things: What Categories Reveal
About the Mind [Lak87]) maintains that we can’t even think at
all without the use of metaphor. Most humans aren’t particularly
good at dealing with the abstract; using a metaphor to relate some
abstract notion to something concrete, something found in every-
day life, makes it much easier for people to grasp.
But metaphors have differing potency. Common, everyday ones
seem to be more like the symbolic representation of L-mode.
Grander metaphors, on the other hand, are much more power-
ful. They can change our thinking and generate their own answers.
What makes the difference?
Juxtaposing Frames of Reference
Metaphor comes from the Greek meaning “to transfer,” with the
idea that you are transferring the properties of one object to
another in a way that is not literally possible.
This notion of combining two different, incompatible ideas is the
very definition of creativity according to researcher/philosopher
Arthur Koestler.9 In his model, some particular topic area forms
a particular frame of reference. The sudden switch from one self-
consistent frame of reference to a different, unexpected, incom-
patible frame of reference is the basis of a strong metaphor. That
junction of the two different frames is called a bisociation.
The more unlikely the association—the further apart the frames of
reference—the greater the creative achievement when bisociated.
This idea is the basis of Edward de Bono’s Po technique.10 Po is
an invented word that tries to go beyond the binary notion of “yes”
9.
See his article “Bisociation in Creation” included in The Creativity Question
[RH76]. Thanks to Steph Thompson for this information. Koestler had some other unsettling beliefs and was accused of violent crimes against women. Genius and madness are often close companions it seems.
10. See PO: A Device for Successful Thinking [DB72] for more.
Report erratum
Prepared exclusively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
ENGAGE AN R-MODE TO L-MODE FLOW
100
and “no.” Several techniques involve Po; for now, you might think
of it as a super-powerful version of suppose.
One of the Po techniques is random juxta-
Use random
position. You take a word from your sub-
juxtaposition to create
ject area and combine it with a completely
metaphor.
random, unrelated word. For instance,
consider the words cigarette and traffic
light. The challenge is to form a bisociation from these completely
unrelated ideas. For example, cigarette and traffic light might meld
into the concept of using a red band on the cigarette as a stop-
smoking aid.
The further away the ideas are, the harder it is to join them in a
metaphor that works. When we come across a particularly inven-
tive metaphor, one where the frames of reference are suitably dis-
tant, we praise the author for centuries:
But soft! What light through yonder window breaks? It is the East,
and Juliet is the sun!
Love is a smoke made with the fume of sighs.
Adversity’s sweet milk, philosophy.
William Shakespeare
What is that bright light in the window? It’s not a celestial object;
it’s this girl Romeo just met at a masquerade ball.11 Love is an
emotion; it has nothing to do with smoke, fumes, or sighs in a
literal sense, but what a fantastic image that conjures up. You can
almost see the wisps of smoke from a young lover’s obsessive desire
The characteristics of the “smoky” frame of reference joins with the
frame of reference for emotion (love); it begins to impart many of
its known (but unstated) features onto the emotional frame. That
sort of imprinting from one frame of reference to another is very
powerful and something we can use to our advantage.
There’s metaphor, and then there’s metaphor.
11. Modern greeting cards have probably inured us to this sort of comparison; it likely had far greater impact in Shakespeare’s day.
Report erratum
Prepared exclusively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
ENGAGE AN R-MODE TO L-MODE FLOW
101
System Metaphor
The original published version of Extreme Programming (described
in Extreme Programming Explained: Embrace Change [Bec00]) fea-
tured a fascinating practice: system metaphor. The idea is that
any software system should be able to be guided by an appropri-
ate metaphor. For instance, a payroll system might be considered
analogous to a post office, with discrete mailboxes, delivery sched-
ules, and so on. Or maybe a scientific measuring system could be
thought of as a manufacturing system with conveyor belts, storage
buckets, and so on.
All metaphors break down eventually, but the idea was that a suf-
ficiently rich metaphor could help guide the design of a system and
help answer questions that come up during development (this idea
is somewhat similar to our discussion of system invariants in The
Pragmatic Programmer: From Journeyman to Master [HT00]).
The features of the metaphorical frame of reference can imprint
themselves onto the software system; implicit, well-understood
properties of the real world start to transfer into the software itself.
But coming up with good metaphors—ones that help answer ques-
tions instead of causing more—can be hard. System metaphor, as
a practice, is not as widely accepted as, say, test-first development
or pair programming.
I was talking to Kent Beck, the father of XP, about metaphors in
general, and he said this:
“Metaphorical thinking is fundamental in programming, as it is
in all abstract thought. When we aren’t aware of our metaphors,
we can lead ourselves astray. Mixing metaphors negates much of
their power. Why do we override a method in a subclass? Clear metaphors make it easier to learn, reason about, and extend code.”
Clear metaphors are a powerful tool, but we don’t always get it
right. Kent goes on: “Why is it that we mess up metaphors? Why is
the converse of add( ) not always delete( )? Why do we insert( ) some-
thing into a container instead of add( ) it? Programmers are sloppy
in their use of metaphors—tables that aren’t at all like tables,
threads that aren’t like threads, memory cells that aren’t like mem-
ory or cells.”
Report erratum
Prepared exclusively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
ENGAGE AN R-MODE TO L-MODE FLOW
102
We use metaphors so much that we’re not even aware of many of
them (as with windows, mice, and so on). It’s really easy to latch
onto the first metaphor that presents itself without thinking too
much about it. However, that’s probably not the best metaphor
you could use.
Coming up with a really good metaphor
Generative metaphors
that has generative properties suitable for
are hard.
your context is much harder. There is
no “metaphor compiler” that can tell you
whether it’s right or not; you have to actually try it in practice. Use
the metaphor to guide your design, and be aware of how it helps—
or not. You won’t know immediately; the outcome will be uncertain.
And as we saw in Section 4.3, Engage an R-mode to L-mode Flow,
on page 92, you need to be OK with uncertainty. Don’t force the
issue; just be aware of it.
After some experience, you might suddenly realize that the
metaphor you started out with was wrong, but this other idea really
fits more closely (and that’s of course fine; it will just take some
code refactoring).
If you’re not used to deliberately creating metaphors, you might
find the practice of a system-level metaphor difficult.12 But there is
a way you can improve your ability to create metaphor and analogy,
funnily enough.
So, This Duck Walks into a Bar...
Humor is neither a waste of time nor a harmless diversion; instead,
it reflects an important ability necessary for thinking, learning, and
creativity. It’s all about connections.
Humor arises from making novel connections across disparate
ideas. It may be absurd, but humor is often based on identifying
relationships and distorting them. For instance, “My best friend ran
away with my wife. I’m sure going to miss him.” You assume the
primary relationship is between the man and his wife, but instead
it turns out that his relationship with his best friend is more impor-
tant to him; the skewed connection makes it funny.
12. Personally, I suspect that’s the driving reason why it hasn’t been adopted as widely.
Report erratum
Prepared exclusively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
ENGAGE AN R-MODE TO L-MODE FLOW
103
Or the all-time classic line from Henny
Youngman: “Take my wife. Please.” You Take my wife.
initially think that “take my wife” is merely
an idiom that means “Consider my wife’s behavior, for example,”
only to realize it is a plaintive request instead. That sudden lin-
guistic U-turn is the source of humor. The creativity comes from
realizing that “take my wife” can have multiple meanings and from
exploiting the potential for misunderstanding.
Comedian Steven Wright was famous for drawing interesting par-
allels, as with his friend the radio announcer who would disappear
when driving under a bridge. That is, Wright drew the analogy that
just as a radio signal can sometimes fade out under a bridge, so
might the radio announcer himself. He also describes accidentally
using his car keys on his apartment door and taking the building
for a spin around the block.
Instead of drawing an analogy, you might also extend an existing
idea past what’s reasonable. For example, if an airplane’s black box
can survive a crash, then why can’t they make the whole plane out
of the stuff?
In either case, the talent for humor comes from drawing or extend-
ing relationships beyond the norm, truly seeing “out of the box.”
A quick wit—being able to draw connections between things that
aren’t related or to extend an idea past its breaking point—is a skill
well worth practicing, honing, and encouraging in your team.
TIP 15
Cultivate humor
to build stronger metaphors.
Have you seen my fishbowl? The implicit, habitual frame of refer-
ence would lead you to think I’m looking for the glass bowl in which
my fish resides. But if the answer is, “Yes, he just got a strike!” then
we’re in a completely different, cartoon frame of reference where
bowl was actually a verb.
By practicing making these far-flung connections, you’ll become
better at doing so. In fact, you’ll actually begin to change the very
fabric of your brain to accommodate this new activity.
Report erratum
Prepared exclusively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
HARVEST R-MODE CUES
104
Next Actions
! Make more metaphors. You can do this as part of software
design or something more artistic—your own jokes, fables, or
songs.
! If you’re new to creating metaphors, start with something sim-
ple: a thesaurus (you know, that thick book that’s sold next
to the dictionaries in the bookstore or that “other” window in
your online dictionary program).
! For more in-depth exploration, try playing with WordNet
(available for all platforms from http://wordnet.princeton.edu).
This gives you not only synonyms but also antonyms, hyper-
nyms, hyponyms, and other derivations of various flavors.
4.4 Harvest R-mode Cues
Despite years of being ignored, your R-mode remains hard at work,
toiling away in the background to match up disparate facts, make
far-flung associations, and retrieve long-lost bits of important data
from the morass of otherwise uninteresting memories.
In fact, it’s entirely possible that your R-mode already has exactly
the answer to the most important problem that you’re working on
right now.
But how can you get at it? We’ll spend the rest of this chapter
looking at techniques to help invite, coax, ferment, and jiggle great
ideas out of your head.
You Already Know
You may already have that great idea or know the solution to that
impossibly vexing problem.
Your brain stores every input it receives.
Every input gets stored.
However, even though stored, it does not
necessarily index the memory (or if you
prefer a more die-hard computer analogy, “store a pointer to it”).
Just as you can arrive at work with no memory of how you got