Error paradigms

As must now be clear, error has many different facets and the subject of error, and how to reduce error, can be approached in different ways. While there are a multitude of different taxonomies and error reduction systems, we can discern some broad general perspectives or error paradigms as they are sometimes called. Following Deborah Lucas (1997) and James Reason (1997)... 

Petroski, Henry, Design Paradigms Case Histories of Error and Judgment in Engineering, Cambridge University Press, 1994. 

Continuous subcutaneous insulin infusion (CSII) has been compared with multiple daily injections of insulin in a randomized study in 32 patients, mean age 13 years, over 16 weeks (195). Of the 16 patients who used CSII one returned the pump twice and one returned the pump once, in both cases for pump software errors. Medtronic MiniMed 508 or Paradigm 511 pumps were used in the study. 

The replication paradigm requires that protein enzymes were not present at the beginning, and RNA replication was therefore performed by ribozymes. Some RNAs can in fact behave as polymerases and replicases, but they are far less efficient than the corresponding protein enzymes, and the accuracy of their replications was necessarily very low. The experimental measures, obtained from interacting coupled nucleotides, have shown that without protein enzymes the replication error e cannot be less than 0.01, which means, from formula 5.1, that primitive RNAs could not have, as an order of magnitude, more than 100 nucleotides (Maynard Smith and Szathmary, 1995). 

We are bound to conclude that the replication paradigm does not offer a plausible model even for postchemical evolution. Of course we cannot exclude that future discoveries might modify such a conclusion, but it would be necessary to discover, among other things, that primitive ribozymes were making replication errors comparable to those of protein enzymes, and this is extremely unlikely. 

Let us now come to the second part of postchemical evolution, the stage that was destined to lead to the origin of the first cells. It is in this stage that we must look for an answer to the problem that the replication paradigm has been unable to solve how did primitive systems manage to increase their complexity without being destroyed by error catastrophes The ribotype answer is based on three points. 

The new paradigm of robust optimal control is well on the way to rendering the linear quadratic Gaussian control obsolete (or at least less important) because it can deal explicitly with model error. 

This was all to change in 1970 when the first chink in the armor of the inherent superiority of experiment over computation appeared. Within a few years, computational results pointed out serious experimental errors and thereby secured a place for computational chemistry as an equal partner with experiment in exploring the chemical sciences. All of these centered on the simple molecule methylene, CHj, establishing what Schaefer has called the paradigm of computational quantum chemistry . 

---