A Beautiful Failure

February 4, 2009

There are times in lab, when even a failure brings something positive with it: a brilliant insight, a new direction perhaps, or simply a beautiful sight. Here is a picture I took this morning in a microscope, of a device I had fabricated after much pain. What you see in the image – the stripes – are lithographically fabricated gold wires, about 20 micron wide, on a silica substrate. To put it a bit obtusely, these wires connect the business end of the device to the outside world. Much to my dismay, one of the wires seemed to be busted (i.e., the connection between the device I made and the outside world was severed)

5_t2_inner_bustedwire1

Ordinarily, this would cause me much angst and frustration. Luckily though, I have alternative wires in the arrangement that could still do the job for me. But a closer look at the failure was what ultimately brought a smile. Isn’t this just beautiful ?

5_t2_inner_bustedwire62

I played quite a bit with the filters on the microscope till I got, what I thought, was a fair enough compensation for a failure !! The gold incense stick flaring out into a myriad subtle hues. Here’s the same image with a different focus of the microscope:

5_t2_inner_bustedwire7

The First Post

November 16, 2007

For a blog such as this, what could be a better “first post” than the famous Feynman lecture on “Cargo Cult Science”. This talk which is featured as the last chapter in, “Surely, you’re joking Mr. Feynman !”, was given as a commencement address at CalTech in 1974. The subtitle is, “Some remarks on science, pseudoscience and learning how not to fool yourself.”

Cargo Cult Science by Richard P. Feynman.

The point about scientific “integrity” that Feynman makes – to avoid fooling oneself – is as simple as it is profound. Most experiments are often conducted with a working hypothesis in mind. Sometimes this hypothesis is clearly articulated to oneself, but often its just a lingering expectation at the back of one’s mind. The trap then, is that its all too easy to “see” the results that concur with your hypothesis, and turn a blind eye to ones that don’t. In other words, to obtain the “desired” result ! This problem is especially severe when the given experiment doesn’t give one a clear, unambiguous result (and where a variety of conclusions can be drawn).

My own way of dealing with this ever-present threat is  – Try to prove yourself wrong every step of the way !  I try and conduct my experiments with the aim of trying to prove my own hypothesis about the problem WRONG ! And after trying this as hard as possible, if I find my hypothesis still holds, then to that extent I am more convinced about it. Of course, how hard one must try this, is also a question of degree, and how careful and relatively confident about the conclusions one wishes to be. All said and done, we are still fallible. Mistakes can be made, and made easily. And that is why I find myself going back to this essay often, especially in times when I think I am “slipping”.

There is much else that Feynman, in his inimitable style, talks about in this essay, but there will be other times to talk about those !