Whether an inventor is employed by a company, or hopes to be, his urgent and itching need is to know if his new idea is a valid one. And the classic, obvious, and apparently simple way to find out is by some form of direct test. ‘Make one and see if it works’, ‘Seeing is believing, it leaves no room for argument‘ and so on. All this seems so obvious as to be hardly worth saying. What is not so obvious is that such statements are often expensive self-deceptions. You will be wise not to believe what you see, or what you think you see. The underlying reason for this is that the direct test method tells you what works but not why it works. It tells you something in particular but nothing in general. It has behaviour but no self-evident principles. It acts but does not talk. If you think it has a voice you are only listening to your own echo.

I was once asked to investigate a situation where a new invention behaved in a thoroughly mysterious manner. A small portable machine (whose exact purpose and design I am tactfully leaving unidentifiable) had been invented on a completely new principle. A full-size working model had been made and extensively tested both for performance and reliability. Based on these results a large company signed a favourable royalty agreement with the inventor and production plans were laid on.