Thursday, November 05, 2009

Tony Hoare on Industrial vs. Pure Research

I have just read a very cogent retrospective piece that Tony Hoare has written for the October 2009 issue of CACM to mark the 40th anniversary of the publication of his seminal paper An axiomatic basis for computer programming. (This is the first article he wrote as an academic. Not a bad way to start, is it?) It is a read that I thoroughly recommend and that I'll make available to the students who are now taking my course on the semantics of programming languages. I am posting an excerpt from that viewpoint article on industrial vs. pure research in computer science since it may be of interest to readers of this blog. I pass the word to Tony.

"Pure academic research and applied industrial research are complementary, and should be pursued concurrently and in collaboration. The goal of industrial research is (and should always be) to pluck the 'low-hanging fruit'; that is, to solve the easiest parts of the most prevalent problems, in the particular circumstances of here and now. But the goal of the pure research scientist is exactly the opposite: it is to construct the most general theories, covering the widest possible range of phenomena, and to seek certainty of knowledge that will endure for future generations. It is to avoid the compromises so essential to engineering, and to seek ideals like accuracy of measurement, purity of materials, and correctness of programs, far beyond the current perceived needs of industry or popularity in the market-place. For this reason, it is only scientific research that can prepare mankind for the unknown unknowns of the forever uncertain future.

So I believe there is now a better scope than ever for pure research in computer science. The research must be motivated by curiosity about the fundamental principles of computer programming, and the desire to answer the basic questions common to all branches of science: what does this program do; how does it work; why does it work; and what is the evidence for believing the answers to all these questions? We know in principle how to answer them. It is the specifications that describe what a program does; it is assertions and other internal interface contracts between component modules that explain how it works; it is programming language semantics that explains why it works; and it is mathematical and logical proof, nowadays constructed and checked by computer, that ensures mutual consistency of specifications, interfaces, programs, and their implementations. There are grounds for hope that progress in basic research will be much faster than in the early days. I have already described the vastly broader theories that have been proposed to understand the concepts of modern programming. I have welcomed the enormous increase in the power of automated tools for proof. The remaining opportunity and obligation for the scientist is to conduct convincing experiments, to check whether the tools, and the theories on which they are based, are adequate to cover the vast range of programs, design patterns, languages, and applications of today's computers. Such experiments will often be the rational reengineering of existing realistic applications. Experience gained in the experiments is expected to lead to revisions and improvements in the tools, and in the theories on which the tools were based. Scientific rivalry between experimenters and between tool builders can thereby lead to an exponential growth in the capabilities of the tools and their fitness to purpose. The knowledge and understanding gained in worldwide long-term research will guide the evolution of sophisticated design automation tools for software, to match the design automation tools routinely available to engineers of other disciplines."

5 comments:

Paul Goldberg said...

This article is a timely reminder about the relationship between pure and applied research, coming as it does, during a dispute in the UK about how research quality should be measured. The Government proposal is to award marks for research "impact", which is narrowly defined in terms of benefit to the UK economy. This web page presents the campaign against the proposal, and has links to many relevant articles.

Term Papers said...

It is glad to see this blog, it is good that and detailed fun to read this, nice informative blog, Thanks for share this article.

Term papers said...

I would like to thank you for the efforts you have made in writing this article.

carry on luggage sizes said...

Great information!!! i think that you need add more images.thanks for sharing. I appreciate this post.

Unknown said...

A good way to experience both of academic studies. Wish this would be the best help and thoughts to everyone.