Theories, as construed by positivists, have descriptive, explanatory, and
predictive functions. Depending on what function a particular theory is
design the perform most consistently, we can envisage the corresponding
classes of theories. In the positivist world, prediction is based on
explanation, hence predictive theories hold the highest status. On the
other hand, the descriptive function is least valued. This also applies to
ranking disciplines regarding their development and maturity.
However, there is one more class -- normative theories. They are in the
realm of the sciences of the artificial. Hard-core positivists have
difficulty accepting them as science products. In many ways these
structures are personal manifestoes and action manuals. Positivists can not
understand that or simply can not accept it. The problem arises from
imposing natural science standards and criteria on the sciences of artificial.
Generally, the natural sciences culture has problems with the scientific
profanity of engineers and the intellectual bohemianism of humanitarians.
One should not be surprised if these major parties do not communicate with
each other -- they simply despise each other.
About the science of design -- they are at least two. One is about design
as a natural phenomena, and the other is about design as artification
activity. In the first case we will have all these theories that
positivists love and adore. In the second case we have the books of Le
Corbusier about Modern Architecture. Again, people mix them very often and
talk in different languages. That is why our list is like a Babylonian Tower.
Regards,
Lubomir
At 02:03 PM 3/2/2003 -0500, you wrote:
>>If you have a theory, in the scientific sense, it leads to predictions.
>
>This assertion has been troubling me for a long while, and since you
>did bring it up... :)
>
>Is it possible to have non-predictive scientific theories?
>
>Based on my (very little) reading of Popper, he would definitely say
>"no", since science is distinguished by its focus on testability. How
>would you test something that doesn't predict? However I'll bet that
>Popper isn't the be-all-and-end-all, so there might be other
>definitions of science that do support the possibility.
>
>Where this relates to design, in my sphere anyways, is that I'm being
>asked to develop a scientifically valid measure of engineering design
>ability. As I understand measurement, to design a good measure
>requires an underlying theory of the thing being measured. Thus I need
>a scientific theory of engineering design. This would imply that I can
>make predictions about design, which I don't think is feasible.
>
>Then we throw stochastics, fuzzyness, etc., into the mix and I just get
>more and more confused.
>
>Si to haul this rant (sorry to all) back to the original question: is
>it possible to have non-predictive scientific theories? If it isn't,
>then can we have a "science of design"?
>
>Jason Foster
At 02:03 PM 3/2/2003 -0500, Jason Foster wrote:
>>If you have a theory, in the scientific sense, it leads to predictions.
>
>This assertion has been troubling me for a long while, and since you
>did bring it up... :)
>
>Is it possible to have non-predictive scientific theories?
>
>Based on my (very little) reading of Popper, he would definitely say
>"no", since science is distinguished by its focus on testability. How
>would you test something that doesn't predict? However I'll bet that
>Popper isn't the be-all-and-end-all, so there might be other
>definitions of science that do support the possibility.
>
>Where this relates to design, in my sphere anyways, is that I'm being
>asked to develop a scientifically valid measure of engineering design
>ability. As I understand measurement, to design a good measure
>requires an underlying theory of the thing being measured. Thus I need
>a scientific theory of engineering design. This would imply that I can
>make predictions about design, which I don't think is feasible.
>
>Then we throw stochastics, fuzzyness, etc., into the mix and I just get
>more and more confused.
>
>Si to haul this rant (sorry to all) back to the original question: is
>it possible to have non-predictive scientific theories? If it isn't,
>then can we have a "science of design"?
>
>Jason Foster
|