Earlier, there was a thread on Michael Fortune’s “design for
production”.
An example of his approach - the design of a hand mirror and the jigs to
permit relatively unskilled people, using readily available and
relatively inexpensive hand held power tools, to produce multiples of
them. Morris Dovey noted that a CNC machine could make them without all
the jigs Fortune came up with. My repsonse was that what would
distinguish the hand and jigs pieces from the CNC pieces was soul, one
would have some and the other would not.

That raised an interesting question.
If it’s OK to use power joiners, planers, table saw,
bandsaw and the like to do the “grunt work”,
where does the grunt work stop and the craftsmansip,
with the “soul” that goes with it, begin?

Morris asked if the grunt work stopped - say at 0.0012” from the final
finished piece. He implied that once the concept was developed to the
point that a CNC machine could be used to make the piece - the rest was
grunt work -the “soul” in a piece coming from the mind of the designer,
not his or her hands, or anyone else’s hands for that matter.

I noted that there are some things that don’t lend themselves to CNC
machine work - undercuts, flowing curves and the like. Morris noted
that a 3 1/2 axis CNC machine, with the right tooling, could do that
kind of thing as well.

I thought about that. With enough technology - and money - almost
anything a craftsman could do could probably also done by automated
machines. So what was it that was missing with the automated approaoch?

I found the answer in the March-April 2006 issue of Woodworker West
(www.woodwest.com), there’s an articel by W. Patrick Edwards titled
“Form Follows Process - A Historic Look at Workmanship”, first
published in The Journal of American Period Furniture in January of
2001. Mr. Edwards, according to the article, has 30 years of
experience restoring antique furniture masterpieces and constructing new
pieces utilizing traditional 18th Century techniques - and he therefore
may be biased towards hand work and hand tools.

In the article, Edwards presents David Pye’s distinction between “hand
work” and “machine work” and Pye’s definition of “craftsmanship” as the
“workmanship of risk” as distinct from the “workmanship of certainty”.
In both cases, he’s talking about production work, be it with hand
tools, with machines or a combination of the two. Remember that
pre-power tools furniture makers, like their power tools successors,
made their living producing furniture, with the emphasis on production.
Pre-Industrial furniture makers who used efficient methods and
procedures to produce their furniture while maintaining their quality
standards prospered just as their successors have.

Pye boils “craftsmanship” down to answering the question “Is the result
predetermined and unalterable once production begins?”

If “risk” is involved, and the opportunity to modify either the method
of work (what tool to use in a specific situation and method of using
it) or the original design is modified to adapt to specific conditions
at hand, then it’s “craftsmanship”. If not, it isn’t “craftsmanship”.
Both “workmanship of risk” and “workmanship of certainty” can result in
wonderful pieces of furniture - or crap. But I think that “workmanship
of risk” allows for more opportunities to create a great pieces than
does “workmanship of certaintly”.

And that gets us back to today’s automated production methods and,
specifically, the use of CNC. CNC tooling and the capabilities of CNC
machines make it possible to manufacture just about anything a furniture
designer can come up with. With good quality control of both the the
raw materials (wood) and the machine processing, high quality furniture
can be produced at a price point significantly below that of anything a
small “workmanship of risk” shop can produce.

So why go the “workmanship of risk” route rather than the “workmanship
of certainty” route? Isn’t it the idea for a piece, developed to a
sufficient level of detail to have a CNC machine “do the grunt worka’,
where the “soul” of a piece is instilled in the piece?

I think that the answer is no - and here’s why.

“Risk” implies not only the opportunity for mistakes and perhaps
failure, but also the opportunity for success, to “produce” a piece
better than the original “plan”.

“Producing” nice/good furniture involves far more than coming up with an
idea, drawing it in sufficient detail to describe what the componets
dimensions are, what goes where and how the parts are joined together.
Beyond The Measured Drawing(s), there are a series of critical decisions
that must be made on the way to the finished “production” piece, some of
which MAY result in minor, or even major modifications of the original
“plan”.

The decisions begin with the wood. Go with one for all the visible
parts or go with multiple woods? If multipe, which woods would
compliment each other the best - fo this piece? This one’s a push since
both approaches, risk and certainty share this / these decisions.

When you get to stock selection and parts layout the two approaches
diverge significantly. While both can account for grain orientation,
even the better sheet goods/ply “cuts layout” software has this
capability, the type of grain and the location of the figure in a
particular part can’t be selected by a machine. Even if a machine could
“see” the grain pattern/figure, could it select the best area of the raw
stock which would be both most pleasing to the eye AND be harmonious
with adjacent parts? After all - we’re talking about furntirue made of
wood. THAT is the first major difference - for wood, unlike metal or
“wood products” is not uniform nor homogeneous, either structurally or
in appearance.

So parts layout is a significant difference. With the “risk” approach,
“optimization” of the use of the wood is a consideration, but not the
over riding consideration. One may chose to “waste” some wood to get a
nicer looking part that goes better with adjacent parts. In normal
handcrafted
furniture Ive heard allowing for 10 to 15% for waste. In higher end
shops it could be 25% or more (though “scraps” and “cut offs” seldom go
to waste).

For me, parts layout/selection is a fun process of getting out a bunch
of “candidates”, moving pieces around, flipping one or more over,
rotating 180 degrees, slipping and sliding to get close to some
combination that may “work”. Then I make a “viewing frame” for the
part’s size, appropriately using framing squares, and slide my “window”
around ‘til I see something in it that “fits” and “works” for the piece
I have in mind. I start with the major focal point(s) of the piece -
door panel(s), drawer face(s), table top, then the “support” parts -
literally or figuratively.

Sometimes this process may change the original design, maybe merely
modifying the size of the piece or the proportions -or - become the
beginnings of an idea for a completely different piece. This is just
one of the things “workmanship of risk” has over “workmanship of
certainty”- It’s adaptive - throughout the process of “producing” a
piece - Pye’s differentiating question - “Is the result predetermined
and unalterable once production begins?” answered.

And that adaptation throughout the process of “production” in the
“workmanship of risk” approach may, or may not, fine tune and improve
the original design and plans. If you’re preparing the stock for a
piece at “human speed”, be it with handtools or with handheld power
tools, one may see - or feel something in the wood that suggests either
an eminent problem or a possible improvement and make a choice. As edge
treatments, sometimes very important in a piece, are worked on - the
wood of one piece may indicate that something other than a 45 chamfer
might bring out a pleasing grain, or remove some less than pleasing
grain - or maybe a sap pocket or the like. The closer the method of
working the wood comes to human speed with tools that allow the human to
sense changes both visually and tactily, the greater the potential for
incremental improvements - or screw ups.

Rather than continue, let me repeat - at some point, the method of
processing the parts of a piece in the “production” process begins
losing “soul” as the human “maker” gets farther and farther from the
wood.

What say you?

charlie b

I'm rather inexperienced in woodworking, but my day job is a software
designer.

In a computer, the fundamental building blocks are completely mechanical
and absolutely predictable. Barring hardware problems the computer does
*exactly* what you tell it to. However, I think it's still possible to
have "soul"--it just means that the "soul" comes in at a higher level of
abstraction.

Consider the examples you gave--part layout based on "what looks best",
modification based on how it "feels" when being worked. Given a
sufficient amount of time and effort this could be at least partially
mechanized. In this hypothetical case, the rules by which the machine
operated would have to have been created by someone. Thus, the machine
would be carrying out (at one level removed) the wishes of the original
designer.

You could eventually get this machine to be pretty good (almost
certainly better than I), but it would likely be impossible to get it as
good as the real artists, simply because not all of the artistic
decisions are amenable to reduction to rules of logic. (Although they
do have software to generate artwork and musical compositions.)

What I'm trying to say is that even if you have something completely
machine-made it would still be possible for it to have "soul"--it's just
that the soul doesn't come from the machine, but the machine *designers*.

I don't think we're anywhere near that level of sophistication yet, but
I wouldn't be surprised to see layout software eventually get hooked to
a digital camera and given some basic understanding of desirable grain
directions.

Chris

Been around long enough to remember the early days of SIG GRAPH?
interGraph's wire frame raster displayed "3D" helicopters "flying"?
How about the beach, with waves and clouds moving while the sunlight
changed computer animation- done by a guy at Lawrence Livermore
Labs. He agreed to work for them, in return for having one of their
super computers to play with for a few hours during low demand periods?

After two days of temps around 100 degrees F, and more forecast, I wrote
a random waves sound generating program on an Atari 800 using Atari
Basic to control volume, tone and time for each of the four sound
"channels"
built into the 800. Sounds were "outwash", "curl", "break" and
"inwash".
Varied volume and "attack", "sustain" and "diminish" by size of wave,
duration
between breaks varied randomly. Got the program working, crashed on the
floor and imagined I was at the beach. It's OK to be hot at the beach.

But machines only do EXACTLY what you tell them to do, with very basic
instructions. While some interesting "art", be it 2D images or sounds,
computer generated stuff seldom deals with aesthetic things well,
or consistently. When it comes to "soul" in furniture making I'll still
go with "wetware" driving tools and muscles.

Some things lend themselves to abstraction - no interplay between
theoretical and real world. That's where the "workmanship of risk"
approach CAN more consistently produce good or even great pieces.

charlie b

charlie b wrote:

Rather than continue, let me repeat - at some point, the method of
processing the parts of a piece in the "production" process begins
losing "soul" as the human "maker" gets farther and farther from the
wood.

What say you?



Can the 'machine' go to the lumber yard and sort through the cherry looking
for exactly the right pieces for the current project? And does it know
enough to pick up some great bubinga that just came in to use in some
future project? Or pick up a block of pink ivory at a good price because
'I'll use it for something'?

Chris Friesen wrote:

I'm rather inexperienced in woodworking, but my day job is a software
designer.

In a computer, the fundamental building blocks are completely mechanical
and absolutely predictable. Barring hardware problems the computer does
*exactly* what you tell it to. However, I think it's still possible to
have "soul"--it just means that the "soul" comes in at a higher level of
abstraction.

Consider the examples you gave--part layout based on "what looks best",
modification based on how it "feels" when being worked. Given a
sufficient amount of time and effort this could be at least partially
mechanized. In this hypothetical case, the rules by which the machine
operated would have to have been created by someone. Thus, the machine
would be carrying out (at one level removed) the wishes of the original
designer.

You could eventually get this machine to be pretty good (almost
certainly better than I), but it would likely be impossible to get it as
good as the real artists, simply because not all of the artistic
decisions are amenable to reduction to rules of logic. (Although they
do have software to generate artwork and musical compositions.)

What I'm trying to say is that even if you have something completely
machine-made it would still be possible for it to have "soul"--it's just
that the soul doesn't come from the machine, but the machine *designers*.

I don't think we're anywhere near that level of sophistication yet, but
I wouldn't be surprised to see layout software eventually get hooked to
a digital camera and given some basic understanding of desirable grain
directions.

I'm troubled by the assumption that the craftsman is less involved in
the work if the cuts are programmed into the machine. There's nothing
other than inertia there to stop the craftsman from stepping in to
modify the course of the blade to affect a curve in a chamfer, or to
arrange the workpiece to take advantage of some effect of the wood.

'Course, you could argue that inertia... or momentum, is the point of
CNC, and getting that involved in the work works against the whole point
of automation.

In any event, there's nothing unalterable about a programmed series of
cuts and the soul or liveliness of a piece can be found in the
craftsman's involvement in its making. Whether that be through the
handles of a spokeshave, or the fingerpads of a keyboard, it is still
transmitted to the wood.

er
--
email not valid

"Enoch Root" wrote in message

I'm troubled by the assumption that the craftsman is less involved in
the work if the cuts are programmed into the machine. There's nothing
other than inertia there to stop the craftsman from stepping in to
modify the course of the blade to affect a curve in a chamfer, or to
arrange the workpiece to take advantage of some effect of the wood.

I believe the assumption would only hold true if the bulk of entire process
to the finished product is done by the machine. *That* I'd attribute to many
factories. Since most of us don't own all these commercial machines, our
constructions are prone to errors and variances in construction that make
most of what we make unique.

On Mon, 27 Mar 2006 07:35:25 GMT, Lobby Dosser
wrote:

Can the 'machine' go to the lumber yard and sort through the cherry looking
for exactly the right pieces for the current project? And does it know
enough to pick up some great bubinga that just came in to use in some
future project? Or pick up a block of pink ivory at a good price because
'I'll use it for something'?

Given the number of computer geeks that have woodworking as a hobby, I
wouldn't be surprised if the answer to this in the future is 'yes'.


-Leuf

On Sun, 26 Mar 2006 22:10:49 -0800, charlie b
wrote:

But machines only do EXACTLY what you tell them to do, with very basic
instructions. While some interesting "art", be it 2D images or sounds,
computer generated stuff seldom deals with aesthetic things well,
or consistently. When it comes to "soul" in furniture making I'll still
go with "wetware" driving tools and muscles.

I think we've only really scratched the surface in understanding how
math plays into the underlying fundamentals of what makes a thing
beautiful. You look at a fractal and there's this thing more
complicated than anything we could come up with, and yet it's based on
something simple. Music is math. You need not be aware of the math
to make great music, but the number of people who can actually do it
is limited. For the rest of us, let's take any help we can get.
Being able to hit a baseball is math. Again, you can be oblivious to
the math and be the best hitter in the world, and undertanding the
math doesn't mean you can hit the ball.

Imagine a computer program that could spit out designs based on
formulas and you could sit back and say No, No, No, Interesting but
No, No, Hmmm... what if I switched that around like such and such
and...

They make 2d and 3d images out of fractals, they make music out of
them, who wants to try furniture?

Some things lend themselves to abstraction - no interplay between
theoretical and real world. That's where the "workmanship of risk"
approach CAN more consistently produce good or even great pieces.

How much of our own work is copying what some other person made? In a
lot of ways we're just really inefficient and inaccurate CNC machines
substituting our own preferences and trying to make up for our
mistakes.


-Leuf

It has been and it ever was thus that a product by which I mean a
piece that has always been intended for replication is substantially
different from a piece which is intended to be a one-off.

(someone told me that I use too many commas. i hope that you enjoy the
result of my response.)

Pye's theories are the natural reification of the Cotswald movement
which came to be the Craftsman Movement which came to be the
underpinnings of the design ethic of Greene and Greene which came to
be the given group of elements that describe the Craftsman ethic at
least in America.

As you can see we have a doubled opportunity to define the idea of
"Craftsman".

The concept of "The Workmanship Of Risk" was addressed by Diderot and
is not something specific to Pye which is why Pye's book has gone out
of print and Diderot's book is still available even though it is a bit
older.

What we do as independent craftsmen is all about the workmanship of
risk. It doesn't matter if we jig it up. Our jigs our open to
question.

I currently work in a manufacturing environment. They call it
cabinetmaking but it is not.

The entire goal of a manufacturing process is to eliminate the
intelligence that would drive the production of a single product and
replace it with a logic that obviates the need for intelligence in
pursuit of the goal of making many of the same product as efficiently
as possible.

When we look at an empty wall in our house and say to ourselves that a
shelf should be there we aremarket researchers. When we define the
requirements of the shelf we are designers. When we modify the design
to agree with the available stock in our shop we are value
engineering. When we figure out the best way to cut and join we are
engineers. When we decide on the the finish that will go best in our
home given the other partners involved we are marketers.

When we are done we are heroes.




Regards,

Tom Watson

tjwatson1ATcomcastDOTnet (real email)

http://home.comcast.net/~tjwatson1/

Music is math. You need not be aware of the math
to make great music, but the number of people who can actually do it
is limited.

The best music is not in the simple ratios of harmonics but the subtleties
of being *outside* (a little sharp of flat) of these ratios. The perfect
envelope of a guitar bend is what separates the "artists" from the 6-string
craftsman. A good drummer will play a little ahead or behind the beat to
create push and pull (tension).


How much of our own work is copying what some other person made? In a
lot of ways we're just really inefficient and inaccurate CNC machines
substituting our own preferences and trying to make up for our
mistakes.

IMO plaster walls look better than sheet rock, specifically because of their
imperfection. In may cases perfection *is* unapealing.

Then again, sometimes "patina" is really just dirt. :-)

-Steve




--
NewsGuy.Com 30Gb $9.95 Carry Forward and On Demand Bandwidth

Enoch Root wrote:

In any event, there's nothing unalterable about a programmed series of
cuts and the soul or liveliness of a piece can be found in the
craftsman's involvement in its making. Whether that be through the
handles of a spokeshave, or the fingerpads of a keyboard, it is still
transmitted to the wood.

But the fingerpads give no tactile feedback FROM the wood TO the
hands
of the person on the keyboard. And with a machine whirring and a
cutter cutting, it's hard to hear when the sound of the cuttiing edge
changes that subtle little bit that warns of the beginning of a
problem.
The machine driven wood removing device also can't see when the
nice grain begins appearing and indicates "screw the plans, I'm
stopping here to keep this look of the wood!"

charlie b

charlie b wrote:
Enoch Root wrote:

In any event, there's nothing unalterable about a programmed series of
cuts and the soul or liveliness of a piece can be found in the
craftsman's involvement in its making. Whether that be through the
handles of a spokeshave, or the fingerpads of a keyboard, it is still
transmitted to the wood.


But the fingerpads give no tactile feedback FROM the wood TO the
hands
of the person on the keyboard. And with a machine whirring and a
cutter cutting, it's hard to hear when the sound of the cuttiing edge
changes that subtle little bit that warns of the beginning of a
problem.
The machine driven wood removing device also can't see when the
nice grain begins appearing and indicates "screw the plans, I'm
stopping here to keep this look of the wood!"

Jim Krenov's isn't the only approach to doing lasting things in wood.

Lately I've been collecting images of backsaw handles for a project.
(Bob Brode's pages have been a big help.) Handles of that era embody a
craftman's influence on the final form, whereas more modern ones do not.

They were also made in a factory setting, with an emphasis on
production. But modern saw handles, even the really good ones, don't
match them in character and distinction.

I don't think it is because of the tools being used to make them that
they don't measure up.

I think it is because the maker is less involved in their production,
and because less thought is being given to their form, their fit, and
how they will be used.

Well, maybe that is similar to Krenov's way in some respects. But where
his design goal emerges from an evolving process, the final form of
those handles is not. All the factors that contribute to the handles'
design and execution are known prior to the making. And yet, they feel
excellant.

er
--
email not valid

On Tue, 28 Mar 2006 07:14:36 -0500, "Stephen M"
wrote:

Music is math. You need not be aware of the math
to make great music, but the number of people who can actually do it
is limited.

The best music is not in the simple ratios of harmonics but the subtleties
of being *outside* (a little sharp of flat) of these ratios. The perfect
envelope of a guitar bend is what separates the "artists" from the 6-string
craftsman. A good drummer will play a little ahead or behind the beat to
create push and pull (tension).

Which is all still math Computer generated music is lifeless
compared to the real thing, but that's because the program doesn't
know to do those things not because it can't do those things.

IMO plaster walls look better than sheet rock, specifically because of their
imperfection. In may cases perfection *is* unapealing.

So if a piece of furniture one of us made came out "perfect" should we
hang our heads, take it apart and try again? I make these intarsia
leaves where the edges and grain of the pieces simulate the veins of
the leaf. I get the gaps between the pieces to an acceptable level
and then I stop, and I like to think the finished leaf looks better
with the small gaps in it. But at the same time if pieces fit
together perfectly I don't sand a gap in where there was none. I
can't have it both ways.

I think for most of us we have to worry more about the border between
"having character" and "careless" than "perfection". If you do manage
to get to the latter you did it by first overcoming the former and you
probably don't need any help understanding it.


-Leuf

The best music is not in the simple ratios of harmonics but the
subtleties
of being *outside* (a little sharp of flat) of these ratios. The perfect
envelope of a guitar bend is what separates the "artists" from the
6-string
craftsman. A good drummer will play a little ahead or behind the beat to
create push and pull (tension).

Which is all still math

At the risk of being completely pedantic, music can be simulated with a
mathematical model. Most if not all simulations are close but not exactly
the same as the original. The reproduction is only as good as the
mathematical model. For example, a CD recording is a represented by a
series of discete values, not the smooth anolog curve of the original sound.

IMO plaster walls look better than sheet rock, specifically because of
their
imperfection. In may cases perfection *is* unapealing.

So if a piece of furniture one of us made came out "perfect" should we
hang our heads, take it apart and try again?

No. I'm suggesting that "perfect", in the eyes of those beholding, is
probably not mathematically straight/flat/square/smooth etc. While I agree
that it's possible that machinery could render just about anything possible,
the real challenge would would be creating an accurate definition of what
our eye sees as perfect. Traditional tooling in conjuction with tactile
feedback loop, (including the analog circuitry in our heads) can to a better
job than a machine with using a fully automated digital simulaiton.

The limitation of automated systems is not the machines, but the algorithms
that drive them (spoken like a software guy eh?). If you can't define what
it is that makes the undulations of a plaster wall appealing then it's
pretty tough to automate its recreation.

I make these intarsia
leaves where the edges and grain of the pieces simulate the veins of
the leaf.

And therein lies the elusive quality of art. The ability to capure the
essance of something without shown the whole. This is why an illustration is
usually more effective than a photograph, because the irrelevant information
is subtracted. I think this applies to written works, music and physical
arts.

Cheers,

Steve


--
NewsGuy.Com 30Gb $9.95 Carry Forward and On Demand Bandwidth

Stephen M wrote:

And therein lies the elusive quality of art. The ability to capure the
essance of something without shown the whole. This is why an illustration is
usually more effective than a photograph, because the irrelevant information
is subtracted. I think this applies to written works, music and physical
arts.


Strippers and sumie as well. Most of what you "see" is filled in
by the viewer's imagination,more so with sumie. In a sumie
(oriental brush painting, typically done in continous shades
of gray) painting (actually "inking"), the bare minimum required
to triggerer the idea of the piece is actually on the paper - the
viewer's eye, experience and imagination fills in the rest.

James Krenov's pieces have that quality, and like haiku,
change with the viewers mental state as wel as how the
light plays with the surfaces and grain.

Nicely put sir. Computer algorithms may be able to produce "art",
but soul will always be the domain of living things, be it a
human or a bird.

(just noticed rithms vs rythms)

charlie b

On Thu, 30 Mar 2006 08:54:48 -0800, charlie b
wrote:

Nicely put sir. Computer algorithms may be able to produce "art",
but soul will always be the domain of living things, be it a
human or a bird.

Not trying to remove the human from the equation. But first we used
the computer to help draw the design. Now it's being used to help
produce the design. We've only used it to produce the design in the
respect that it can give us a way to fool around with things without
getting a new sheet of paper or piece of wood. I think we could use
it to help come up with the design as well. The same ignorance of the
computer towards not doing certain things also means it doesn't know
it shouldn't do things that the human has preconceived notions not to
do. It's not afraid to take risks, it has no concept of risk.

The best thing is that exploring things like this, and artificial
intelligence, is that we find out how little we actually understand
things that we take for granted. Trying to teach the computer we end
up learning about ourselves.


-Leuf

Leuf wrote:

The best thing is that exploring things like this, and artificial
intelligence, is that we find out how little we actually understand
things that we take for granted. Trying to teach the computer we end
up learning about ourselves.


Personally, I prefer teaching kids and get the same result in
terms of learning about myself. Kids ask questions and
sometimes display what can only be described as glee.

charlie b