I suspect that we all have similar intuitions, more or less practical, that inescapably bear on our thinking whenever and wherever we see contrivances. In "To Engineer is Human" Henry Petroski wrote: "I believe, and I argue in this essay, that the ideas of engineering are in fact in our bones and part of our human nature and experience."

I also perceive that these intuitions are considered suspect by Darwinian evolutionists in that they potentially interfere with an objective approach to biology. For, as Richard Dawkins so famously put it, (paraphrasing): "biology is the study of complex things which have the appearance of having been purposfully designed."

I think it would be helpful to formulate a comprehensive outline of the fundamental process of human design efforts, and to answer the question: "Is it rational to infer from the "humanness of design" that the contrivances of biology are the workings of an intelligent agent (or agents) bearing some resemblance to our own nature?" This question leads to others: "What is fundamental to the design process?; What does human nature entail? and Is agency necessary for contrivances (irreducibly complex, of course) to exist?" It seems to me that this last question has been the main subject of ID discussions, such as Dembski's explanatory filter, and Behe's irreducible complexity. I simply include it for completeness. The first two questions are of particular interest to me.

In closing this post, I'll take a crude "stab" at outlining the design process as I practice it:

1. Problem definition: With the exception of those instances where we "tinker" with ideas, and stumble upon a concept that we then associate with a need, most engineering problems are driven by the initial perception of a problem in need of a solution. (This is often referred to as a 'Marketing Specification.')

2. Characterization of fundamental principles involved: Environmental conditions relevant to the problem at hand lead to a narrowing of relevant physical laws, materials and boundary conditions likely to proove useful in a solution set. (Feasibility study.)

3. Formulation of Design constraints and tests necessary to evaluate progress toward the desired end. (Engineering Specification.)

4. Conceptual, analytical and iterative exploration of design approaches. I suppose this is what most people have in mind when they picture human creativity in action.

5. Documentation. The process of translating a design into code, checking it for errors, and storing and transmitting it, is necessary for any contrivance to be reproduced.

6. Design for Manufacturing and Assembly (DFMA): This topic really exists in steps 3-5, and is the subject of a course of study at Stanford U. It is an extremely fruitful skill to develop, wherein the designer keeps in mind the cost benefits associated with streamlining the manufacturing process.

I'll leave it there for now. I'd really like to hear from others on the questions noted above, and on this outline of the design process.

Respectfully submitted,

Columbo

[ 27 February 2002: Message edited by: Columbo ]

[ 27 February 2002: Message edited by: Columbo ]

[ 27 February 2002: Message edited by: Columbo ]

Great thread. I personally think that studying human engineering is a very rich source of understanding for intelligent design theory. In fact, I've done a fair bit of study on this topic myself, but from the perspective of the Russian study of inventiveness, known as "TRIZ," an acronym for "Theory of Inventive Problem Solving." This science was started in 1946 by Genrich Altschuller as he examined Soviet patents, looking to understand what, precisely, separated inventive solutions from non-inventive or routine solutions.

The results are very interesting. Soviet TRIZ theorists found that there were two kinds of problems and two kinds of mechanisms at work to solve those two kinds of problems. The first kind is the routine problem, which can be solved by "tinkering" or trial and error. However, they found severe limitations on what trial and error can accomplish, and those limitations are called "contradictions." Basically, a contradiction is a trade off in function that arises when the only way to continue improving some parameter of a system requires a decrease in some other parameter. At this point, there is a break, or "edge" which trial and error simply cannot get beyond. Think of a multi-dimensional hypervolume which contains all the types of systems that can be reached by trial and error from some existing system; each dimension represents one adjustable parameter. Furthermore, the edges of that hypervolume are well-defined and concrete; these edges are what TRIZ theorists call contradictions. The bottom line is that trial and error only moves about within the hypervolume and can never penetrate the absolute barriers of contradictions. That is why a second mechanism, the inventive mechanism, is needed for the second kind of problem.

The second kind of problem is what Altschuller called an inventive problem. These problems require the overcoming of a contradiction before they can be solved. In other words, the solution cannot be reached by merely fiddling with the parameters of an existing system. Usually this is because something genuinely new is required. What this does is introduce a new parameter to the system, thereby expanding the dimensionality of the hypervolume in such a way that a solution can be found. Think of solution space being a square 2-dimensional plane, and the solution we need to find is a point that lies above (and out of) that plane. If we can somehow increase the dimensionality of the solution space to a 3-dimensional cube, the solution point will lie within the solution space. This is what happens when an invention is made, something fundamentally new is introduced to the system that overcomes a contradiction and permits a workable solution to be found.

All these ideas are developed at length in my paper in the archive, titled "Inventions, Algorithms, and Biological Design":

http://www.iscid.org/ubbcgi/ultimatebb.cgi?ubb=get_topic&f=10&t=000003

I apply these ideas to specific examples in biology where inventive solutions were clearly required and were found. The implication is that trial and error (the Darwinian mechanism of mutation and selection) was not the operative mechanism for these innovations.

I'd greatly appreciate any comments you might have, coming from an engineering perspective. Do these ideas make sense to you? Or am I totally off base here?

I appreciate your addition to my "model" of the design process. Though I haven't yet read NFL, it's on order, so I'm sure to benefit from your insights there. Thanks,

to John Bracht:

I'm working my way through your very interesting paper on TRIZ with acute interest. I suppose it would be more appropriate for me to wait until I finish it before responding to you, but I can't wait that long! Too many comments are floating around in my head to postpone this reply, so please forgive any resulting misperceptions.

If I read it correctly, the Russian Altschuller and his colleagues have characterized as an actual dichotomy, the two forms of design for two types of problems: Trial and error for routine problems, and creative insight for inventive problems. To quote your paper, psychological inertia (associated with trial and error)... "is doomed to forever wander in the wastelands of sub-optimality." I'm generally in agreement with this, but I think the line between trial/error and innovation is somewhat fuzzy in actuality.

A few comments: 1) This whole discussion is reminiscent of Thomas Kuhn's descriptions of 'paradigm shifts' in science. "Routine" problems are solved within the paradigms, while innovation shift them. Critics of Kuhn have presented some counterexamples, (I cannot recall where I read them at the moment) which may call for a refinement of his work. 2) Petroski (cited above) has a chapter (5: Success if Failure) that examines trial and error in engineering. It seems that it was the dominant method prior to the 19th century. Speaking in contrast to earlier engineering projects such as cathedrals and pyramids, Petroski writes:

"Sudden and catastrophic bridge collapses were introduced into the daily way of life, and they had to be reckoned with not through the classical trial and error method, but with a newer and more abstract method that employed pencil and paper in lieu of chisel and stone. What the engineers of the nineteenth century developed and passed down to those of the twentieth was the trial and error of mind over matter. They learned how to calculate to obviate the failure of structural materials, but they did not learn how to calculate to obviate the failure of the mind."

Elsewhere in that chapter, Petroski associates the introduction of elegance, rather than true innovation, into human design projects, and attributes it to analytical approaches rather than gestalt inspiration. 3) In "Edison: Inventing the Century (Chicago U. Press) Neil Baldwin writes of the criticism that Thomas Edison endured from others because of his dependence on trial and error over more theoretical approaches. Yet he was one of America's greatest inventors.

All this to say that trial and error, when performed by intelligent agents, is not analogous to Dawkins' "blind watchmaker." We cannot really escape that mysterious human capacity to innovate that you wrote about. In my opening essay, I mused about the Darwinist's suspicion of design intuitions in biology. Perhaps there is a similar weakness involved when Darwinists (or anyone for that matter) attempt to conceive of biological contrivances arising from "unguided" trial and error. I rather doubt if there is really such a thing. The 'error' side of the equation implies observation and learning.

More later John ... and thanks so much for the feedback.

Columbo

[ 27 February 2002: Message edited by: Columbo ]

[ 28 February 2002: Message edited by: Columbo ]

quote:

---

If I read it correctly, the Russian Altschuller and his colleagues have characterized as an actual dichotomy, the two forms of design for two types of problems: Trial and error for routine problems, and creative insight for inventive problems. To quote your paper, psychological inertia (associated with trial and error)... "is doomed to forever wander in the wastelands of sub-optimality." I'm generally in agreement with this, but I think the line between trial/error and innovation is somewhat fuzzy in actuality.

---

I want to clarify just a bit. My paper is not especially focused on the mechanisms by which humans invent things (the murky psychological details of innovation) but rather emphasized the application of TRIZ principles to the Darwinian mechanism.

I think it is clear that humans often utilize a trial and error process when they invent. However, they first use an inventive mechanism in setting up the hypervolume in which to move about via trial and error. In other words, we use trial and error only after we have the basic system and want to tweak it. Thus, Edison knew the basic form and principles of the lightbulb and the only thing missing was the filament. Trial and error worked fine for that problem. But the essential point is that setting up the hypervolume itself cannot be a process of trial and error, precisely because trial and error can only operate upon some preexisting system, within some preexisting set of possibilities. It is the defining of these possibilities, the setting up of the hypervolume, which inevitably requires intelligent input.

Furthermore, if we are operating within the wrong hypervolume we will never stumble across the solution no matter how much trial and error we perform. The solution must reside within the space of possibilities for the system.

There are two things I want to emphasize in this post. (1) Trial and error is extremely effective for solving routine problems. Some problems are simply too complicated for a theoretical solution (like the properties of polypeptide chain folding or the vagaries of which material will make the best lightbulb filament). But trial and error is remarkably well-suited for sifting variants until the solution is found. Notice that in each of the above examples the solution can be found simply by sifting variants from a given initial condition. I freely acknowledge the efficacy and power of trial and error once the initial conditions are properly set up.(2) Trial and error is useless for finding any solution outside the possibility space of the system it operates upon (creates variants of). In other words, trial and error simply cannot move outside the hypervolume of possibilities for a given system. If a solution outside that possibility space is needed, something fundamentally new must be added to the system. It is this addition of something new, making available a solution that was unreachable before, that is the essence of inventive problem solving. Keep in mind that once we increase the hypervolume to include a workable solution, we may still use trial and error to find ("home in on") that solution. This, I think, is why Thomas Edison and others who invent via trial and error are still, at heart, utilizing the inventive mechanism.

I look forward to your further comments.

Thanks for the clarification. That helps.

I think what I am fishing for in this string is a synthesis of the notions held by Petroski, Polanyi, Dembski (sounding kinda Slavic here!) and now Altschuller. Perhaps something along this line:

Inherent in human nature and human activity is a recognition, appreciation and exercise of teleology, innovation, efficiency, beauty and economy in design. All of these are present in biology, which is perhaps why it is so hard to speak of biological contrivances in any other than these terms. (I recognize that some scientists, such as Dawkins and Gould, do call into question the suggestion that biological "designs" are optimal, but I don't think they are convincing.) Is this a coincedence, are we this way because we are at (or near) the "top" of the evolutionary tree, or is there epistemic justification (apart from CSI and IC) to infer that biology is the work of an intelligent agent?