The LMS is monitoring developments closely and will continue to work with the Council for the Mathematical Sciences (CMS) as well as responding as the LMS on areas of particular concern to the Society.

We welcome here all contributions discussing how EPSRC developments are affecting your own institution or department; how should the LMS respond; and what can individual members do to help?

Advertisements

11 Comments
leave one →

Michael Singer at Edinburgh has written a great letter to David Willetts:

http://burttotaro.wordpress.com/2011/09/28/singer-to-willetts-the-undesirable-position-you-warn-against-has-already-been-reached/

I’d like to point bloggers and readers to a letter (25th August)

LMS Response to EPSRC

from the LMS to Philippa Hemmings (`Lead for Mathematical Sciences Capability Delivery’). I should explain that this is a response to her letter of 18th August requesting input from the LMS “in the form of information, facts and figures” to the Shaping Capability process by the end of September.

The LMS felt unable to respond in those terms (due to ignorance of the process, uncertainty about what information was wanted, whether it would be used in context, the unrealistic timescale,…..). However, the LMS is keen to engage constructively, and the letter is an attempt to do so. It includes a selection of questions one would need to answer, and some suggestions of essential features of a process of capability shaping that might have credibility with the mathematical community.

I expect there will be a range of views about the letter, but with luck this will have the benefit of generating discussion and new ideas here. This may in turn even lead to further ideas of how we can engage with the EPSRC to best support a happy, healthy and productive mathematical community, training new mathematicians and supporting the cultural and economic prosperity of the country……

An excellent collection of pure maths “impacts” can be found in answers to these MathOverflow questions:

Real-world applications of mathematics, by arxiv subject area?

Recent applications of mathematics

Applications of mathematics

Applications of math: theory vs. practice

The first one, in particular, has a truly impressive list of answers, and I’d recommend it to anyone seeking ammunition for their arguments.

Also re Ken’s comment: 4. Yes, putting together algebra, geometry, topology and number theory is rather strange. But I also strongly agree with Richard that a better way for EPSRC to allocate maths funding is to tell us how much we get and let the SAT decide who gets what. I therefore think that the cruder EPSRC’s partition of mathematics is, the better.

There is a possible counterargument, which is that if those four subjects are lumped together, then an EPSRC bureaucrat who doesn’t know much about maths will be tempted to think of that “theme” as more homogeneous than it actually is, and will therefore allocate a smaller proportion of the total mathematics funding to it than is appropriate. But if we can somehow get the message across that AGTNT is a hugely, almost comically central pure mathematics “theme”, then we might be better off than we would be if it were split up further. (Do we want EPSRC deciding on the relative importance of algebra and topology, for instance? I don’t think so.)

Re Ken’s post: 1. We can’t thank people like you enough for the job you do; I can’t imagine how hard and depressing it is to spend so much time fighting fires on our behalf, often with little thanks. (Still, seeing “geometry” lumped into “algebra” must have made you smile.)

And 2. my point about the government and yours about whining scientists are related. I’m not at all sure the government know what EPSRC are doing in their name. I think writing to them complaining about cuts is going to be fruitless. But telling them that EPSRC are micromanaging research they know nothing about, picking winners, and basically not letting us loose to do the best we can with the resources that exist (cut or not) might give them pause for thought, especially if we can suggest an alternative. (Eg in maths: just give the money to the maths SAT and let them spend it.)

Apologies for not signing this, but as an interested part (i.e. current ‘about to finish’ PhD student) I prefer to remain anonymous.

According to EPSRC their reason for giving funding just to applied probability and statistics so far is that they didn’t ‘assess’ yet (or whatever term they use) the rest of the areas and that they would open fellowships once this is done. But why don’t they assess all the other areas at the same time? Is it a lack of time? I find hard to believe all the assessment committee do the assessment for all the areas, at least with the same effort and dedication. And then it stroke to me a much simpler reason. If they have a limited amount of money, considerably smaller to the one last year, and they break the funding calls into areas then they get the following:

1) Not everyone complains at the same time.

2) If the areas that ‘grow’ are assessed and given money out first (and so far that seems to be the pattern), by the time they reach the rest of them (and I suspect Algebra, Geometry, Topology, Number Theory may be one of the last ones) there will be little or no money left, so they will not only be ‘reduced’ by almost suppressed by the ‘fact’ that the box is already empty.

Whether this is their purpose or not, this will be the result, and it should be taken in account.

Regarding bypassing EPSRC, as Richard suggested, it may actually be a good idea, but this should be done with care. Pure mathematicians have dealt with (conservative) politicians before, and we do not have to go as far as Newton. It may not be a bad idea to ask them. Professor Swinnerton-Dyer comes to my mind first and maybe he could be contacted. It may be possible also to contact those who assess the board. I think Burt Totaro published a list in his blog. Finally we are not the first ones in EPSRC to be hit by this senseless approach (again Burt’s blog has a post about Chemistry) and it could be interesting to learn from or team up with other subjects.

Some quick responses:

1. Great idea from Tim to use this blog to gather examples of pure maths impact. It gets very boring to be using the same tired examples all the time.

2. The need to aim above EPSRC: Absolutely, agreed. Frank Kelly has written to ask for a meeting with David Willets, and has also written to David Cameron, sent yesterday, press release today. See

http://www.cms.ac.uk/activities.html

On a similar theme, it is important wherever possible to act in concert with other fields within EPSRC’s remit – otherwise it’s too easy for math. science to be targeted as a collection of whingeing eccentrics.

3. Tim’s point regarding the threat to fellowships only being temporary. There may be some truth in this, and the replies from Philippa Hemmings and David Delpy to earlier correspondence suggest as much. But these responses are a response to pressure from the community. Moreover, EPSRC’s budget shows spending on fellowships set to decline (in cash terms) from £51m in 2010-11 to £44m in 2011-12, remaining constant in cash terms for succeeding 3 years, representing a substantial decline even before shaping is implemented -see

http://www.epsrc.ac.uk/plans/deliveryplan/Pages/next.aspx

4. The need to engage with EPSRC: yes, I agree, and in CMS (and LMS) we have been trying to do so. But it is very difficult to do so in a way which maintains at least some vestigial intellectual respectability. To give one example: the maths programme picture of subthemes and their supposed links at

http://www.epsrc.ac.uk/ourportfolio/themes/mathematics/Pages/default.aspx

shows “algebra/geometry/topology/number theory” as one “circle” (often shortened simply to “algebra” in EPSRC documents). The only way EPSRC permits engagement here is to argue for/against the “growth/maintain/reduce” options for this disc, as for the other discs into which it chooses to partition the subject. (This leaves aside the fact that we do not actually know what “grow”, “maintain” and “reduce” actually mean.)

The LMS needs to set up a register of new and useful mathematical ideas to show the world what mathematicians are achieving. This register should include room for comments. This can be done on the internet.

I am an applied mathematician/physicist who is working on problems of industrial significance. At the instigation of Bob Chapman, an NDT (Non-Destructive Testing) specialist from British Energy – now EDF I have assembled an academic team of world-leading specialists who worked on diffraction by elastic wedges – a problem that vexed diffractionists for well over 50 years. We have refined existing semi-analytical techniques, published in one of the best mathematical journals, SIAM Journal in Applied Mathematics, and produced unique computer codes that are currently used by EDF in ultrasonic inspections of nuclear reactors when looking for the backwall surface-breaking cracks. The only reason this success story was made possible is that the industrial contact has a PhD in Applied Mathematics from Manchester, one of the world centres in theory of wave propagation in solids. I am also working with the French Atomic Commission and Doosan Power Systems – again only because there are engineers there with similarly solid mathematical backgrounds. Most technologists these days, even those who are reasonably well educated are aware only of commercial packages, do not understand when they fail and how easy it is not to notice the failure. They have no inkling where applied mathematics could be used and how. This has been my experience and that of other industrial mathematicians I know – there is nothing you can do unless there is “our man” on the other side (that is a (wo)man with a very solid mathematical background). Yes, a lot of problems can be solved with commercial packages, but one never knows when an important and urgent safety or environmental problem will rear its ugly head, which can’t! Then, who do you call ?!

Larissa Fradkin, Sound Mathematics Ltd.

Professor Emerita, London South Bank University and

Associated Professor, Brunel University

Firstly, I love Tim’s idea of this being a forum for listing pure maths “impacts” — we should try this. Of course it’s hard. Much of maths’ influence is too subtle, and includes the trickle down effect of good physicists, engineers, etc wanting to be in universities with good mathematicians. Also there’s the national pride aspect — just as we spend money on the olympics, so we should have the money to spend on Fields medalists & Abel prizewinners, Hawking & Penrose, etc: maths is one of the subjects that the UK is really superb at, and we don’t have many Nobel prizewinners in many other subjects.

Secondly, I am sure the LMS have considered going over EPSRC’s heads to the government, and I would strongly encourage that.

I think EPSRC are very good at showing different faces to different stakeholders, as they would call us.

The give us the impression that they’re the government, while they tell the government that they are (or at least speak for) the scientists.

They tell us that all these changes are pushed by government; they tell the government that they’re best placed to know what’s the right science to fund and so they drive the changes.

So when the government wants to consult us, it contacts them, who speak on our behalf. It looks to me, from the outside, that they go to great lengths to prevent actual scientists (who might be off-message, or might not have been on a speaking-in-bland-and-empty-acronyms course) from giving their views. For instance, when parliamentary committees ask for submissions, they send these requests to EPSRC, who don’t pass them on to us, oddly.

If an economically-liberal government knew that an unaccountable quango was micromanaging research, deciding on the latest fashions, and picking winners, they would probably be shocked. If the government have any unifying philosophy at all it is surely to leave us alone, give the universities the money and trust them to make the best use of it. They presumably think EPSRC is some benign group of scientists deciding funding by peer review of each other’s work, like NSF. We should set them right.

In fact EPSRC seem to have depressingly little interest in the best science, and more interest in building up their powers and roles. As an example, the maths community have for decades asked for the postdoc decisions to be made earlier in the year, to be competitive with the US. EPSRC have always refused — since they are not accountable, so long as they fill the places on the scheme, they consider they have done their job. The fact that this means they very often lose the top few candidates to US universities is apparently of no concern to them. They could have used their money much more effectively, but their impact is unfortunately not measured, only ours. Now they appear to be simply slashing the scheme altogether, or they will offer us some money later in the year when all the best candidates have already accepted offers overseas.

EPSRC’s slavish obsession with research fashions is also terribly damaging. What they consider fashionable is now old-hat and barely research any more (they are getting into the “digital economy” at least 20 years late, and would not now fund the next Tim Berners-Lee because that next big breakthrough will be in a different field). And as Tim (Gowers, not Berners-Lee) pointed out in a different post, when they change their fashions so often, they end up putting money into a field that is weakly supported in Britain (because they didn’t fund that field before it was fashionable) and so they give the money to weak scientists. This is a terrible use of funds. It seems no long view of science is possible because each new leader in EPSRC wants to come in and make their (new) mark without worrying about the predecessor’s legacy. And of course making long term research predictions should be impossible anyway. So mightn’t we be better off without EPSRC at all ?

I say this very unsurely, as one obvious alternative is to put all the money in the RAE/REF, which may have even worse problems deciding how to to split funding between areas. Deciding how much money to give to a science is obviously an extremely difficult exercise, but EPSRC’s current self-important way of doing it is surely the worst. Almost any mathematician would tell them that funding PhD students and postdocs is the most important use of money in maths (and is surely very cheap); without it the long term health of our subject will collapse. But they have explained that they are not listening to us because that would constitute “lobbying by interest groups”.

Like more or less everybody I know, I think that the EPSRC decision is misguided. However, I would urge people to respond to it constructively. There are two things in particular that I have in mind. First, we need to bear in mind that the restriction of EPSRC fellowships to statistics and applied probability is not likely to last for long: the other areas of mathematics are all still under review, and as far as we know at present they will be judged worthy of fellowships once those reviews are complete. As I understand it, we will know about this within a year, and possibly quite a bit sooner. If we ignore this in our response, then EPSRC can simply say, with justification, that we’ve got our facts wrong, and our broader message risks getting lost.

As for the broader message, while I think there are many arguments one can make against trying to direct mathematical research towards areas likely to lead to short-term economic gain, I also think that economic gain of some kind is a legitimate criterion for EPSRC to use (along with other criteria of course) when deciding how to allocate resources in an era of spending cuts. I therefore think that it is crucial that instead of simply dismissing EPSRC’s arguments (which in my opinion will lead them to dismiss us by saying that we would say that wouldn’t we, which we would), we should make a serious attempt to engage with EPSRC on its own terms. We all know that mathematics has a huge economic impact, and that the way it has that impact is often rather subtle and indirect. We must explain this as clearly as we can. We must also try to counter a tempting argument that would say that pure mathematics is published openly, so if it has an economic impact, there is no particular need for the actual research to have been done in the UK. (In brief, the counterargument would be that the kind of understanding you need to spot opportunities to apply mathematical research comes from belonging to a culture that actively participates in that research. You don’t just read a paper in another area of mathematics, instantly grasp its entire significance and go away and apply it.) What’s more, I think all our arguments need to be backed up with convincing examples: if we mention the RSA algorithm yet again, it will look as though all the pure maths we ever actually need was done hundreds of years ago. This blog could be a very good place to collect such examples, especially if people have personal experiences to relate.