Tying in with our celebrations of the 250th anniversary of Bayes’ theory, StatisticsViews.com interviewed Dennis Lindley, a British statistician, renowned for being one of the founding fathers of Bayesian statistics who celebrated his 90th birthday last week.

Lindley read mathematics at Trinity College, Cambridge in 1941. During the war the degree course lasted only two years and, on finishing, Lindley had a choice between entering the armed forces and joining the Civil Service as a statistician. He chose the latter and, after taking a short course given by Oscar Irwin, he joined a section of the Ministry of Supply doing statistical work under George Barnard.

After the war Lindley spent some time at the National Physical Laboratory before returning to Cambridge for a further year of study. From 1948 to 1960 he worked at Cambridge, starting as a demonstrator and leaving as director of the Statistical Laboratory. In 1960 Lindley left to take up a new chair at Aberystwyth University. In 1967 he moved to University College London. In 1977 Lindley took early retirement at the age of 54. From then until 1987 he travelled the world as an “itinerant scholar.” He has continued to write and to attend conferences. He was awarded the Royal Statistical Society’s Guy Medal in Gold in 2002

I interviewed Professor Lindley back in May at his home in Somerset. He was very kind and welcoming and taking into consideration his age and retirement in 1977, he has not been out of touch with the latest developments in statistics. Next to his chair was a large pile of books, including Nate Silver’s *The Signal and the Noise* and just before my arrival, he was reading articles on his iPad. When talking about his career, Lindley was often moved by his memories, especially with regards to my final question. He talked to me about his career, Bayesian statistics, working with George Barnard and Leonard Jimmie Savage, how he clashed with R.A. Fisher, the best and worst (!) books on statistics that he’s read and the advice he would offer to the getstats campaign.

**1. Do you think over the years too much research has focused on less important areas of statistics?**

I think that is true and that it’s always going to be true. You can’t tell at a particular point in time what is going on and you have to make mistakes on your way. If you’re a Bayesian, you might do a bit better than a non-Bayesian, haha! No, everyone, including myself, has wasted a lot of time on irrevalent research. In fact, I was clearing out some old journals the other day and reading as I went along, and most of the papers are irrevalent now, obviously not at the time, but that’s life.

Working with George Barnard was wonderful…there was five of us in the room and new methods of quality control were being introduced. It was better than being an ordinary graduate student!

**2. What do you see as the greatest challenges facing the profession of statisticians in the coming years? I was at ‘Margins of Error’ held by the Royal Statistical Society and Ipsos Mori and one of the main points that came out is the misrepresentation in the media and that the Royal Statistical Society is now training journalists on special courses.**

I think the greatest challenge is to make society more numerate than it is at the moment. A couple of days ago, I read an article in *The Guardian* by Tracey Emin and she was talking about art, and if there is no art, then the whole world will collapse. Nonsense! The first thing you need is food for your belly, secondly water and thirdly, shelter. Let’s sort out poverty first. We’ve got to turn society into a numerate society.

Related to your question about important developments, I would add that decision-making is important. There are certain decisions that we have no idea how to solve and those are decisions where there is competition – competition between two countries is the obvious example but it is musical game theory, a game between two people. How do we play a game and we don’t know the answer to that. If I was young again and off to the University of Cambridge, that would be the problem that I would want to solve – what the axioms are, write down what happens and we would then know more. There are several solutions but they don’t work, it’s a very difficult but very important problem.

**3. Are there people or events that have been influential in your career?**

Two people, really. My schoolmaster, Mr Messenberg. I wanted to be an architect and he said to me ‘Nonsense, you’re a mathematician!” So he turned me into a mathematician and he made me go to Cambridge.

The other is Wishart, whom statisticians know because he had a distribution called Wishart distribution. One day, whilst I was at the Scientific Civil Service, I received a letter from Wishart and he asked whether I’d like to apply for a job at Cambridge. I was astonished! I went to my boss and showed him the letter, asking what I should do. He read the letter, turned round to me and said, “Goodbye Dennis.” So I went to Cambridge and I don’t know why Wishart invited me. I didn’t know at the time but a distinguished statistician applied for the job and he wrote about it, unhappy that he was not successful. But why did Wishart ask me? I’ve never understood why.

These two people changed my life, Messenburg for getting me into mathematics, Wishart by getting me to Cambridge. Wishart really encouraged me. He once said to me, “I don’t understand what you’re doing but I think it’s pretty good.” I was very distraught when he died suddenly in Mexico.

On an intellectual level, there was de Finetti and Harold Jeffreys who was Professor of Astronomy at Cambridge. Fisher was Professor of Genetics at the time and I used to say the two best statisticians in the world were professors at Cambridge and neither of them worked in the statistics department! Jimmie Savage, of course, influenced me and he died very young. They were the great people. Adrian Smith, who was the brightest student I’ve ever had. One of the joys of life is teaching a really good graduate.

**Copyright:** Image appear courtesy of Professor Lindley