My pseudovisor at Flinders, Professor Paul Calder, politely read my 2011 Honors thesis - TDE: Design of a bioplausible Turing Machine- and then did a most unkind thing- he asked me to prove it! I had thought about this a bit, and said I couldn't. Rather, my work was plausible, where plausibility was defined by me as bipartite- (i) familiarity (ii) utility. If enough people use a theory (familarity) and some of those are impressed enough to go back and use it again (utility), then that theory is plausible. Behaviourism was plausible. Until Skinner tried to use it to explain language in 1959, just after I was born. A young, brash man called Noam Chomsky politely said 'computer says no' (A 'Little Britain' quote, for those British TV comedy buffs out there.)

 

Like many researchers, I was not sure how I achieved success in my analysis and model building. It was a hit or miss affair. Unlike most researchers, I claimed my model to be the truth. This meant that intuitively, I knew how to do magic.

 

But science, though it may be magical (as in wondrous) for many of us, it is not magic. It was those hardy souls of the enlightenment that guided science though the pillars of Hercules - religion was the left leg, and sorcery sophistry) was the right. The centerpath required having  'Hercules balls' (ie courage above and beyond) . Galileo was kept inside his home, under house arrest till he died, and he confessed!

 

 

When my fellow Aussie, the cognitive philosopher David Chalmers, tries to pass repackaged dualism off as legitimate science, he is exercising his right of sophistry, Hercules left leg. In these modern secular times, It is easy to avoid religious prejudice, but not so easy to avoid being seduced by its opposites- peudoscience, pseudosophy and quackery. Apparently, there is some guy out there saying the heart is not a pump. Good luck when he gets old and exhibits symptoms of ischaemic disease- lets see him convince the medicos at the ER, his wife and his kids.

 

True, some science is magical (I hope mine is) but after you do the trick, you need to show the audience how it is done. The revealing of the mechanism behind the magic, the debunking and demystification- THAT is science.

​

I knew that what I was doing was reverse engineering (RE)- going from a given machine back to the blueprint, the reverse of the usual direction of the engineers job. RE has a positively 'dodgy' reputation. I once saw a 'knock-off' of an iphone 3. It looked crude- the package- but it had an FM radio built in, and could recieve analog and digital TV signals by a non-internet pathway. The eurodesigner in me baulked, but the steampunk in me said 'I want one!'.

 

I knew that I was using Occam's Razor- I was deliberately choosing the minimally complex machinery or mechanism that could satisfactorily do the job.  I also knew that I was using that famous dictum invented by the author Arthur Conan Doyle and spoken by his star character, the Victorian detective, Sherlock Holmes- "Eliminate that which you know not to be possible. Whatever remains, however improbable, MUST be the truth." After some thought, I decided to coin the term 'Holmes' Shroud' to describe is use in a research context. One throws the shroud over the dead ideas (those that are impossible, ie do not 'live'), thus revealing those that, though they are lying as still as stone, must nevertheless still have a beating heart at their core. 'Shroud' also has a positively creepy, miskatonic air, as if we are standing in the cold damp chill of the London Fog, knowing that Jack the Bleedin' Ripper is lurking in the dark, dank shadows.

 

The remaining part of my process was only named recently. For years, I knew that my selection of sub-sections of the cognitive model depended on a mental process I thought akin to 'common-sense'. Once one appraised oneself of all the available information, and used Holmes' Shroud (HS) to eliminate the logically impossible hypotheses, the remaining answer often seemed to be the 'bleeding obvious', to use the appropriate cockney phrase. What exactly was going on in my mind, cognitively speaking.

 

I was re-reading an essay on semiotics by C. S. Pierce, when I stumbled across it. There it was, in black and white, well, perhaps sepia-toned monochrome, after all he wrote it in the 1930's. He said of selecting the correct model- when you have chosen the right model, those data which previously appeared to be outliers, now seemed 'unremarkable'. I first called this 'Pierce's Hook' but later changed it to 'Pierce's Arrow' because it sounded jazzier and pointier.

 

Lets rewrite this trio of principles in their usual order of application-

1. Holmes' Shroud (HS)

2. Pierce's Arrow (PA)

3. Occam's Razor (OR)

 

Yet still I wasn't satisfied. This explained the 'nuts and bolts' of my method, but not my overall approach to other people's concepts and theories. Why did my theorizing always proceed to completion, wheras others seemed more circumspect, and happy to find only half of the answer to a given problem.

 

It was when writing down my theory of language, that it hit me- discovering the truth about a scientific problem and encoding that truth in a theory ( a set of interacting rules, a machine, if you like) was equivalent to learning a new language.

 

Now that I had the right metaphor, could I see more clearly what was going on. My method is heavily interdisciplinary. I will almost never look seriously at a theory that someone else has invented. Rather, I will try to develop a new one from first principles, referring to core sciences directly, and not relying on the intermediate ideas and concepts of other cognitive scientists, unless (like the ideas of James, Russell & Whitehead, Hull, Skinner, Chomsky, Simon & Newell, Treisman, Grossberg, and others) they seemed 'correct' to me.

 

It occurred to me, then, that my approach was like that of a first language learner (FLL)- see diagram at left, while the approach of most other scientists was more like that of a second language learner (SLL). SLL's adopt a conscious approach, and do not learn all the complete language forms, while FLL's seem driven by the involuntary processes of their LAD (Chomsky's famous Language Acquisition Device). For FLL's, I described the mechanism as Dependency driven, while for SLL's I described it as Constituency driven. The former is about learning new semantics (ie discovering new knowledge), while the latter is about mapping new syntax onto existing semantics (ie reinterpreting existing knowledge).