# Mathematical and computational neuroscience

**Jack Cowan had a profound impact on mathematical neuroscience. A large number of his pupils hold chairs today in the field of mathematical biology and computational neuroscience. What is actually the difference between mathematical and computational neuroscience, if there is any?**

There is a difference between *mathematical* and *computational* neuroscience. Jack Cowan explains it being asked by Roger Bingham:

Do you have a simple explanation, something your mother could understand, to sort of explain what you’ve been doing these past 40 years?Cowan answers:

Well I’ve just been trying to apply the methods of mathematical physics to thinking about how the brain works. By that I mean that there is a way in which physicists approach the world, theoretical physicists, that I think really, really works and is really interesting. They don’t try to put in every detail of what the phenomenon is like. They, if they have good taste, they select only those details that are really important for the questions they want to answer. And they construct what are sometimes called toy models, which aren’t facing reality, to quote the title of a book by a friend of mine, Sir John Eccles, but they abstract from reality just what is needed to understand something. And I think that’s what I’ve been trying to do with respect to brain mechanisms: try to make toy models that contain enough details to answer questions about and give you ways to think about what’s going on in the brain. It’s not, I mean, it’s not something that’s commonly done. A lot of the time people do computational neuroscience where they put in a lot of details and make simulations and study what goes on. I don’t do that. I tend to put in as few details as possible and say things that are interesting with few details rather than put in a lot of details.

Computational neuroscience and mathematical modelling are not necessarily but quite often different approaches. Non is superior to the other. Sometimes, you need computational methods to solve even simple toy models.

I like both types of work. Mathematical modeling is much more satisfying for me because you gain a deeper understanding. Computational methods can be fun, too. Usually, you are closer to the experimental data and therefore it is easier to cooperate with experimental groups. This, of course, is essential, as only the experimental work can guide our mathematical theories.

**Thinking out loud with Jack Cowan**

The whole interview can be seen below and the transcript is here. There are many more comments on science today and in fomer times, e.g. rules how to publish if you want to become an amazingly successful scientist, or Jack Cowan's fascination with visual hallucinations.

I just tweeted this blog post, as it is deserving due attention.

Michael: thanks for your comment and tweet! But I think you will agree that it is the interview of Jack Cowan that deserves due attention.

I must modestly say, in this post, I'm just sharing something with you, and hardly added own thoughts.

The interview made a deep impression on me. In fact, I wonder what it was that you obviously liked?