I hope Steve Kline is remembered, not only for the things he did,
but for the special way he did them. Steve Kline was a person of
vision. So are we all. But Steve's commitments to vision were
specially intense, and specially disciplined. I hope some sense of
Steve Kline's commitment to human vision can be remembered,
and can guide us. Pictures are hard. A young child can often
respond with very clear words. Ask that child "draw me a picture"
- see what you get. Ask an adult the same, and you'll often be
appalled at how little detail the adult can actually construct and
remember. Steve Kline wanted pictures, whenever he could get
them. Images were part of his thinking. Steve liked abstractions
all right, but he wanted abstractions that connected, in step-by-step
graphical and pictorial detail, to things that he could SEE. I
hope Steve is remembered for that special, careful, disciplined way
In 1613, Galileo spoke of two kinds of "vision" - sensory
vision "what the eye of the forehead" registers, and
understanding "what the eye of the mind envisions." Steve
Kline worked hard to connect the world that could be seen with his
physical eyes, to useful systems of ideas in "the eye of the mind."
Steve formed and judged ideas with the same concern for detail,
the same desire for traceable connection between different points
of view, that he applied to pictures.
Steve knew how hard it was to get ideas that correlated, step
by step, with pictures, or graphs, or clean outlines. Difficult or
not, that was his standard. Clear, fully connected patterns of ideas,
explicitly matchable to tangible things could be CHECKED, and
focused, and modified. For Steve, getting this clarity and
consistency was an essential part of the creative process.
Steve Kline checked his own ideas against evidence all the
time. He judged other people's ideas by the same incisive
standards he used to judge his own. Sometimes, if he had a clear,
meticulously checked image in his mind, and your image was less
focused and less checked, he could be overwhelming. Often,
though not always, he was also right.
The fluids profession owes a debt to Steve's forceful
temperament, as well as his intellect. The idea had become
dominant that turbulence WAS statistical. Forceful men said
that flows were so complicated that you could "see anything you
wanted in them." These people said that cause and structure
ceased to be when flows became turbulent. Some statistical
models described some data very well. The system seemed
closed and complete, and sterile.
Over a period of about a decade, Steve and others showed
data and presented ideas supporting a different notion. Turbulence
could be thought of in differential and structural terms. Flow
visualization, carefully done, DID show real patterns. Notions of
causality and structure made sense in turbulence.
It is too stark, but still clarifying, to say that this was a fight
between people who were for pictures, and people who were
against them. This was a controversy between people who
believed in pictorial detail, and the three dimensional complexity it
could show, and people who believed in stark mathematical
descriptions, stripped of pictorial content, where "football shaped
correlations" were "all you ever knew, and all you needed to
know." Steve stood on the side of pictorial detail.
It was a long hard fight for Steve and his side. Steve was
shunned at some technical meetings, and it bothered him. Steve
was laughed at by his main opponent in controversy, a powerful
and funny man, whose jokes could stick and burn. Most of all,
Steve remembered this. Steve cornered people with flow
visualization pictures in meetings and said "see those spots?" "see
those streaks?" People refused to look, or REFUSED TO
ACKNOWLEDGE WHAT WAS BEFORE THEIR EYES.
The problem wasn't boredom. Sometimes hands shook. This was
Steve and his colleagues believed their eyes, and ideas that
fit what they saw, and kept on. I'm glad.
So are many other people. Today, flows around whole
airplanes are being modeled in computers. Today at Stanford,
flow modeling of whole turbojet engines is well begun. Better
airplanes and better flow machines of all kinds have already
resulted, and more will come. This work would be unthinkable
according to the older statistical model of turbulence.
Steve had some admirable but sometimes uncomfortable
virtues. Steve Kline was absolutely sure that he could be wrong,
and unashamedly sure that you could be, too. Steve was
remarkably comfortable with the limits of his own understanding,
and unashamedly certain that you had the same limits yourself.
His tests of his rightness or yours, of his limits or yours, were stark
pictorial standards, or as close to pictorial standards as he could
Steve Kline thought that the highest standards of description
and understanding involved pictures, and words, and math, all fit
together. The pictures and words and math had to be traceably
related, step-by-step, with enough sharpness and enough cross-references for checking of many kinds.
Steve thought the documentation in manufacturing
engineering was very good description. It had enough views that,
with hard looking, one could see what was being described. It had
enough words so that interactions were clear. It had enough
numbers and mathematical relations for a sense of detailed
proportion. Pictures, words, and numbers fit together. Steve
worked to get his own ideas that clear, and that detailed, and that
tightly fit to context.
If, sometimes, Steve seemed clearer than other people, it was
because he was. He'd worked to be. Steve's conceptual images
were crafted, checked, reworked, and worried over, like paintings,
not rough sketches. Steve's conceptual images were vivid to him,
and he could often make them seem real to others.
Steve Kline spent much of his career staring at flow patterns
that he could not understand, and that nobody else understood,
either. Lots of times, the patterns were not only too complicated
to fit in his head, they wouldn't fit into a supercomputer. Steve
worked hard to make connections that worked, and sometimes did.
But often when he was working, he had to say "I don't
understand" or "I can't make this fit" tens or hundreds of times a
day. He achieved something rare - a matter-of-fact, comfortable
sense of his own ignorance, and a matter-of-fact sense of what he
had reason to believe.
Steve had a hard won, vivid sense of the boundary between
where he understood and where he didn't. This gave him
unusually clear judgments about what problems needed to be
attacked, and what problems were ripe for attack. These were
judgments that he could explain and justify to others. Some of
those judgments made a difference.
Steve was not a man of religious faith, but he was a man of
intense intellectual faith. Steve believed that reality was there,
even if it was unseen. He believed the world was understandable
in the end. Steve believed that by looking hard at problems,
representing them more and more sharply, more and more starkly,
and checking and rechecking logic and evidence, breakthroughs
would be possible. Steve wanted to define and represent systems
more and more sharply, more and more carefully, more and more
meticulously, until details stood out that might offer clues to
something new and useful. He taught his students to look harder,
look longer, and look more meticulously than they'd naturally
want to. Sometimes Steve's focus would exasperate and exhaust
everyone around him. And yet, sometimes he'd be able to
straighten something out, and solve difficulties, before the rest of
us were even aware that there was a problem.
The connections between his fluids work and his socio-technical work were closer, Steve felt, than people thought.
Steve's strongest commitment, in fluids, was to the notion of
representation, the ideal of models that made detailed
mathematical and observational sense. Steve felt that the first
half of his book, CONCEPTUAL FOUNDATIONS owed an
enormous amount to fluid mechanics, and addressed problems that
had to be solved if certain kinds of progress in fluids modeling
were to be possible. In the later part of his career, some people
felt that Steve sometimes neglected the fluid sciences. So far as I
could tell, Steve himself never felt that way. He was working on
modeling problems that had to be cracked, to permit progress in
fluids modeling, and to permit progress elsewhere, too.
I'll miss him very much.
If you would remember Steve Kline, look at your problems a
little more sharply. Stare at them a little longer. Do more cross-checking. Make sure that what you see and what you say about it
fit tightly. Worry more. Be a little more ready than you might
otherwise be to discard an idea related to your problem, and try
again. Keep at it. Sometimes you may gain new and
important insights, as he did.
That would be Steve Kline's best memorial.