In Search of Certainty

Author preface to second edition

In preparing the second edition of the book, I have made as few changes as possible to the text. My goal is not to update or correct it, as much as to repair some deficiencies in the first attempt at explaining an intricate topic. I prefer that the book remain a cultural document, a child of its time, however naive that might prove to be in the long run.

Some obvious points failed to come through the text at first attempt, especially with regard to material structure of infrastructure in Chapter 12. For the sake of keeping the promise of the introduction, I was embarrassed enough to want to improve on that. The republication by O’Reilly seemed like the perfect opportunity to do that. Minor improvements have been made throughout, and substantial changes made to sharpen Chapter 12. I have also added a summary of points from each chapter at the end of the book, as a comprehension aid to all readers, and perhaps as a study aid for college students.

This book is an ambitious project, and there is a lot to swallow in its pages. However, like the books that were just a little beyond my reach (and hence inspired me to learn) as a teenager, I hope that this will be a book readers can revisit multiple times to discover new insights, over several years. No one can absorb the entirety of such a story in one sitting.

One interesting historical point seems worth noting. Following the release of the first edition, it was pointed out to me by Jeff Sussna how the writings of William Ross Ashby on the nascent subject of cybernetics [Ash52, Ash56], in the 1950s, parallels much of the first part of this book. Having recovered copies of these forgotten works, long since out of print, I was amazed to see how closely his thinking tracked my own. Some terms naturally differ. His mention of ‘over-stability’, for instance, is what is referred to here as fixed point convergence. Ashby (who was contemporary with Asimov’s robot stories) described an analogue kind of cybernetics. Such a viewpoint was also my first instinct, given a similar background in physics. However, it is not the right approach for today’s machinery, as readers will discover in these pages. I believe that the true insights came from realizing the digital nature of information, leading to what we might call a modern cybernetics. There is a tendency to grant historical insights, ahead of their time, almost mystical truth, especially amongst laymen—whether it be Ashby or Hobbes, as examples. However, it would be wrong to think that nothing new has been learnt in the intervening years. Far from it. While words are easily compared and assigned significance by a modern reader, precise scientific understandings are harder won, and always slower to emerge. One could say that the digital version of Ashby’s work was covered in monograph Analytical Network and System Administration [Bur04a]. The later addition of functional semantics through Promise Theory [BB14] then went beyond the scope of what Ashby could address in his analogue form, though he seems to have appreciated that the problem existed. His work was remarkable for its time, and its insights, but it is surpassed by modern knowledge.

Finally, I would like to give humble thanks to all the readers, and reviewers, who expressed glad tidings about the first edition. Their remarks exceeded my wildest expectations, and that makes the efforts expended in writing all the more rewarding.

— Mark Burgess, Oslo 2015.