40
Work to Rule
In industries or situations where one is not allowed a full-
blown strike, a work slowdown is often used as a means of
demonstration.
Often this is called work to rule or malicious obedience,
and the idea is that the employees do exactly what their
job description calls for—no more, no less—and follow the
rule book to the letter.
The result is massive delays and confusion—and an effec-
tive labor demonstration. No one with expertise in the real
world follows the rules to the letter; doing so is demonstra-
bly inefficient.
According to Benner (in From Novice to Expert: Excellence
and Power in Clinical Nursing Practice [Ben01]), “Prac-
tices can never be completely objectified or formalized
because they must ever be worked out anew in particu-
lar relationships and in real time.”
all developers the same, regardless of ability. This does a disser-
vice to both novices and experts (and ignores the reality that there
is anywhere from a 20:1 to 40:1 difference in productivity among
developers, depending on whose study you believe).9
TIP 2
Use rules for novices, intuition for experts.
The journey from novice to expert involves more than just rules
and intuition, of course. Many characteristics change as you move
up the skill levels. But the three most important changes along the
way are the following:10
• Moving away from reliance on rules to intuition
9.
In 1968, a difference of 10:1 in productivity among programmers was noted
in Exploratory Experimental Studies Comparing Online and Offline [Sac68]. The gulf seems to have widened since then.
10. Identified in From Novice to Expert: Excellence and Power in Clinical Nursing Practice [Ben01]; more on this landmark book in just a bit.
Report erratum
Prepared exclusively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
DREYFUS AT WORK: HERDING RACEHORSES AND RACING SHEEP
41
Figure 2.3: Dreyfus model of skill acquisition
• A change in perception, where a problem is no longer a collec-
tion of equally relevant bits but a complete and unique whole
where only certain bits are relevant
• Finally, a change from being a detached observer of the prob-
lem to an involved part of the system itself
This is the progression from novice to expert, away from detached
and absolute rules and into intuition and (remember systems
thinking?) eventually part of the system itself (see Figure 2.3).
The Sad Fact of Skill Distribution
Now at this point you’re probably thinking that the great bulk of
people fall smack in the middle—that the Dreyfus model follows a
standard distribution, which is a typical bell curve.
It does not.
Sadly, studies seem to indicate that most
people, for most skills, for most of their Most people are
lives, never get any higher than the second advanced beginners.
stage, advanced beginner, “performing the
tasks they need and learning new tasks as the need arises but
never acquiring a more broad-based, conceptual understanding of
the task environment.”11 A more accurate distribution is shown in
Figure 2.4, on the following page.
Anecdotal evidence for the phenomenon abounds, from the rise of
copy-and-paste coding (now using Google as part of the IDE) to the
widespread misapplication of software design patterns.
11. Described in Standards for Online Communication [HS97].
Report erratum
Prepared exclusively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
DREYFUS AT WORK: HERDING RACEHORSES AND RACING SHEEP
42
Figure 2.4: Skill distribution
Also, metacognitive abilities, or the ability of being self-aware,
tends to be possible only at the higher skill levels. Unfortunately,
this means practitioners at the lower skill levels have a marked
tendency to overestimate their own abilities—by as much as 50
percent, as it turns out. According to a study in Unskilled and
Unaware of It: How Difficulties in Recognizing One’s Own Incom-
petence Lead to Inflated Self-Assessments [KD99], the only path to
a more correct self-assessment is to improve the individual’s skill
level, which in turn increases metacognitive ability.
You may see this referred to as second-order incompetence, not
knowing just how much it is that you don’t know. The beginner
is confident despite the odds; the expert will be far more cau-
tious when the going gets weird. Experts will show much more
self-doubt.
TIP 3
Know what you don’t know.
Unfortunately, we’ll always have more advanced beginners than
experts. But even though it is weighted at the bottom, it’s still a
distribution. If you’re lucky enough to have an expert on your team,
you need to accommodate them. Similarly, you need to accommo-
date the few novices, the many advanced beginners, and the small
but powerful number of competent and proficient practitioners.
Report erratum
Prepared exclusively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
USING THE DREYFUS MODEL EFFECTIVELY
43
Expert != Teacher
Experts aren’t always the best teachers. Teaching is an ex-
pertise in its own right; just because you are expert in some
subject is no guarantee that you can teach it to others.
Also, given the phenomenon that experts are often unable
to articulate why they reached a particular decision, you
may find that someone at a competent level might be in
a better position to teach a novice than an expert would
be. When pairing or mentoring within the team, you might
try using mentors who are closer in skill level to the trainee.
The hallmark of the expert is their use of
intuition and the ability to recognize pat- Intuition and pattern
terns in context. That’s not to say that matching replace
novices have zero intuition or that com- explicit knowledge.
petents can’t recognize patterns at all but
that the expert’s intuition and pattern recognition now take the
place of explicit knowledge.
This transition from the novice’s context-free rules to the expert’s
context-dependent intuition is one of the most interesting parts of
the Dreyfus model; so our goal, for most of the rest of this book,
is to see how we might better harness intuition and get better at
recognizing and applying patterns.12
2.4 Using the Dreyfus Model Effectively
By the late 1970s or so, the nursing profession was in dire straits.
In a nutshell, these were their problems, which I’ve drawn from
several case studies and narratives:13
• Nurses themselves were often disregarded as a mere commod-
ity; they just carried out the highly trained doctor’s orders and
; weren’t expected to have any input on patient care.
12. That’s patterns in the usual English sense, not software design patterns.
13. Described in From Novice to Expert: Excellence and Power in Clinical Nursing Practice [Ben01].
Report erratum
Prepared exclusively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
USING THE DREYFUS MODEL EFFECTIVELY
44
Ten Years to Expertise?
So, you want to be an expert? You need to budget
about ten years of effort, regardless of the subject area.
Researchers∗ have studied chess playing, music composi-
tion, painting, piano playing, swimming, tennis, and other
skills and disciplines. In virtually every case, from Mozart to
the Beatles, you find evidence of a minimum of a decade
of hard work before world-class expertise shows up.
The Beatles, for instance, took the world by storm in 1964
with a landmark appearance on the Ed Sullivan Show. Their
first critically successful album, Sgt. Pepper’s Lonely Hearts
Club Band, was released shortly after, in 1967. But the band
didn’t just magically form for a tour in 1964—they had been
playing in clubs since 1957. Ten years before Sgt. Pepper’s.
And hard work it is—merely working at a subject for ten
years isn’t enough. You need to practice. Deliberate prac-
tice, according to noted cognitive scientist Dr. K. Anderson
Ericsson, requires four conditions:
• You need a well-defined task.
• The task needs to be appropriately
difficult—challenging but doable.
• The environment needs to supply informative
feedback that you can act on.
• It should also provide opportunities for repetition and
correction of errors.
Do that sort of practice, steadily, for ten years, and you’ve
got it made. As we noted in The Pragmatic Programmer:
From Journeyman to Master [HT00], even Chaucer com-
plained that “the lyfe so short, the craft so long to lerne.”
However, there is some good news. Once you become an
expert in one field, it becomes much easier to gain exper-
tise in another. At least you already have the acquisition
skills and model-building abilities in place.
Thanks to June Kim for the reference to Dr. Ericsson.
∗.
See The Complete Problem Solver [Hay81] and Developing Talent in
Young People [BS85].
Report erratum
Prepared exclusively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
USING THE DREYFUS MODEL EFFECTIVELY
45
• Because of pay-scale inequities, expert nurses were leaving
direct patient care in droves. There was more money to be
made in management, teaching, or the lecture circuit.
• Nursing education began to falter; many thought that formal
models of practice were the best way to teach. An overreliance
on formal methods and tools eroded real experience in
practice.
• Finally, they had lost sight of the real goal—patient out-
comes. Despite whatever process and methodology you
followed, despite who worked on this patient, what was the
outcome? Did the patient live and thrive? Or not?
If you read that list carefully, you may have noticed that these
problems sound eerily familiar. Allow me to slightly edit this bullet
list to reflect software development:
• Coders themselves were often disregarded as a mere commod-
ity; they just carried out the highly trained analysts’ orders
and weren’t expected to have any input on the design and
architecture of the project.
• Because of pay-scale inequities, expert programmers were
leaving hands-on coding in droves. There was more money
to be made in management, teaching, or the lecture circuit.
• Software engineering education began to falter; many thought
that formal models of practice were the best way to teach. An
overreliance on formal methods and tools eroded real experi-
ence in practice.
• Finally, they had lost sight of the real goal—project out-
comes. Despite whatever process and methodology you fol-
lowed, despite who worked on this project, what was the out-
come? Did the project succeed and thrive? Or not?
Huh. It sounds a little more familiar that way; indeed, these are
serious problems that our industry now faces.
Back in the early 1980s, nursing professionals began to apply the
lessons of the Dreyfus model to their industry with remarkable
results. Dr. Benner’s landmark book exposed and explained the
Dreyfus model so that all involved parties had a better understand-
ing of their own skills and roles and those of their co-workers. It
laid out specific guidelines to try to improve the profession as a
whole.
Report erratum
Prepared exclusively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
USING THE DREYFUS MODEL EFFECTIVELY
46
Over the course of the next twenty-five years or so, Benner
and subsequent authors and researchers turned their profession
around.
So in the best spirit of R&D (which stands for “Rip off and Dupli-
cate”), we can borrow many lessons from their work and apply them
to software development. Let’s take a closer look at how they did it
and what we can do in our own profession.
Accepting Responsibility
Twenty-five years ago, nurses were expected to follow orders with-
out question, even vehemently—and proudly—maintaining that
they “never veer from doctor’s orders,” despite obvious changes in
patients’ needs or conditions.
This attitude was enculturated in part by the doctors, who weren’t
in a position to see the constant, low-level changes in patients’ con-
ditions, and in part by the nurses themselves, who willingly abdi-
cated responsibility for decision making in the course of practice
to the authority of the doctors. It was professionally safer for them
that way, and indeed there is some psychological basis for their
position.
In one experiment,14 a researcher calls a hospital ward posing as
a doctor and orders the nurse to give a particular medication to a
given patient. The order was rigged to trigger several alarm bells:
• The prescription was given over the phone, not in writing.
• The particular medication was not on the ward’s usual
approved list.
• According to labels on the medication itself, the prescribed
dosage was double the maximum amount.
• The “doctor” on the phone was a stranger, not known to the
nurse or staff.
But despite these clear warning signs, 95 percent of the nurses fell
for it and went straight to the medicine cabinet, en route to the
patient’s room to dose ’em up.
14. Described in Influence: Science and Practice [Cia01].
Report erratum
Prepared exclus
ively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
USING THE DREYFUS MODEL EFFECTIVELY
47
Fortunately, they were stopped by an accomplice who explained
the experiment—and stopped them from carrying out the bogus
order.15
We see very much the same problems with programmers and
their project managers or project architects. Feedback from coders
to those who define architecture, requirements, and even busi-
ness process has traditionally been either lacking entirely, brutally
rejected, or simply lost in the noise of the project. Programmers
often implement something they know is wrong, ignoring the obvi-
ous warning signs much as the nurses did in this example. Agile
methods help promote feedback from all members of the team and
utilize it effectively, but that’s only half the battle.
Individual nurses had to accept respon-
sibility in order to make in-the-field deci- “I was just fol owing
sions according to the unfolding dynamics orders!” doesn’t work.
of a particular situation; individual pro-
grammers must accept the same responsibility. The Nuremberg-
style defense “I was only following orders” did not work in WWII,
it did not work for the nursing profession, and it does not work in
software development.
But in order to accomplish this change in attitude, we do need to
raise the bar. Advanced beginners aren’t capable of making these
sorts of decision by themselves. We must take the advanced begin-
ners we have and help them raise their skill levels to competent.
A major way to help achieve that is to have good exemplars in the
environment; people are natural mimics (see Section 7.4, Learn
About the Inner Game, on page 203). We learn best by example. In
fact, if you have children, you may have noticed that they rarely do
as you say but will always copy what you do.
TIP 4
Learn by watching and imitating.
15. This was an older study; don’t go calling the hospital up now with bogus orders, or the feds may well come a knockin’.
Report erratum
Prepared exclusively for Jose Luis Loya
gggggggggggggggg
this copy is (P2.0 printing, January 2009)
USING THE DREYFUS MODEL EFFECTIVELY
48
No Expertise Without Experience
Jazz is an art form that relies heavily on real-world expe-
rience. You may learn all the chords and techniques
Pragmatic Thinking and Learning Page 5