The Good Science Project

Recently I was invited to give the keynote lecture at the annual summer symposium of the London Interdisciplinary Social Science DTP, and I was delighted to be there. There were impressive student presentations during the day, and I was reminded again not only of the importance of relations between the disciplines, but also of the importance  of students from different institutions meeting and working together.

In my lecture I debated the interesting challenges and opportunities a STEM-based research institute faces when it begins to integrate the social sciences. As I explain, the challenges and the opportunities are both institutional and philosophical.

The Strange Case of Imperial College London; and how the Social Sciences Ride to the Rescue

Keynote lecture to the Annual Symposium of the London Interdisciplinary Social Science Doctoral Training Centre

Friday May 30^th 2025

I’d like first to thank Professor Mujamdar, of Imperial College London, and also your very own Doctoral Training Programme[i], for inviting me today. We spend far too much time explaining why this particular institution, or that particular institution is the best for this, or for that, when actually the race to be best probably is not terribly creative, on the whole. So your small organisation, the London Interdisciplinary Social Science Doctoral Training Programme, with its aim of bringing students together from across London, is very much to be praised.

I’m going to start my lecture by thinking about conversation – the conversation between the social sciences, and the natural sciences. Surely the two can make common cause. My comments will be wide-ranging, but my case study is Imperial College London, where I work. At Imperial the social sciences are arriving in a big way. I will explain, as far as I can, some of the reasons. And I will ask about the way an institution can make best use of its social science capacity, and its natural science capacity. I will ask: how easy is it going to be, living together?

It seems to me the social sciences are particularly dependent on collegial conversation and trust. Enlightened people will say the same about the sciences, but surely the discussive tone is very evident in the social sciences. Perhaps them LISS, and days like today, carry such importance because, literally, they encourage, and carry, conversation. If our academic institutions are  somewhat rushed these days, and anxious to promote their own virtues, then for sure the conversation between students of different universities will be a very important way for research institutions to learn from each other.  This talkative aspect of the social sciences, you might remember, was celebrated by the philosopher and historian Thomas Kuhn, who was very interested in what indeed might constitute the boundary bewteen the natural and the social sciences. Kuhn’s tool of comparison, interestingly enough, was the conversation. His argument can be found in what now is an elderly book, a classic rather than a sure guide, his Structure of Scientific Revolutions, published in 1962.[ii]  At the start of that book – which, by the way, is quite readable and is consistently interesting – Kuhn suggested that social scientists can never stop talking and arguing, and in particular, disagreeing.

The debates of the social sciences, Kuhn suggested, are endless and never resolve. The same thing, famously, was said by the eminent ethicist Alasdair MacIntyre, in his classic text After Virtue.[iii] Here, MacIntyre is eloquent on the reasons ethical discussion tends to be interminable: they never stop because the antagonists approach the problem – assisted suicide, for example – from different angles, angles that are quite hard to change.  For both Kuhn and MacIntyre, the point about the natural sciences is that at some point they stop talking. Once the answer has been found, these authors suggest, the conversation can stop, and we all can move on, and write the textbooks. For Kuhn, because the talk can stop – the situation in the natural sciences – so the settled paradigm can emerge. Whereas the social sciences, always able to develop a new and opposing view, the talk can never stop, and so the settled truth can never emerge.

We might protest that such a picture is to misunderstand the natural sciences, to fail to see that science also can never ‘settle’ to the truth. But we must admit that a search for, and a belief in, the ‘settled truth’ has always been a feature of the sciences. Perhaps science’s ‘truth project’ began in the early 17th century, when Sir Francis Bacon wrote his influential Novum Organum, a letter to King James 1 explaining why it was time to give science its due, fund it properly, and agree it had its unique method that set it apart from the mental wanderings of the classical Greeks.

To oversimplify, in some sense the natural sciences are naturally mute, while the social sciences are garrulous. Crude though this characterisation might be, it gets at the idea that the sciences confront nature, which ignores the human language. The social sciences however, even when quantitative, sit in the arena of human action, and thus depend on our ideas about that most dynamic and shifting of materials, humanity. An articulate critic of this view was the philosopher Richard Rorty, celebrated for his text Philosophy and the Mirror of Nature[iv]. Anxious to see the links between the social and the natural sciences, Rorty was at pains to develop a constructivist understanding of science, where acknowledgement is made of the ‘human factor’ in scientific knowledge. His Mirror of Nature metaphor, representing the idea that science is simply a direct reading of nature, , is interesting for the suggestion again of silence: we don’t think of mirrors as noisy. Rorty wants to replace the mirror metaphor with another, where he describes science as ‘an edifying conversation’. He wishes to bring the talk back into science, because it will make us better scientists. We might be realists, and consider ourselves to be describing the material world, but still our talk is central.

I don’t want now to exhaustively and philosophically probe the distinction, or lack of distinction, between the social sciences and the natural sciences, extremely interesting as this may be. More, let me turn to my case study, and discuss certain themes running quite warmly at Imperial right now, concerning the relationship between the social sciences, and STEM innovation. In a nutshell, Imperial, famous for being a ‘STEM institution’, finds itself more and more dependent on the social sciences. The reason is easy to see, but still quite startling for an institution that since its founding in 1907 has tended not to set up research spaces for the humanities and the social sciences. Why, then, this recent change? Let me explain.

All academics and all academic institutions are very conscious of the concept of impact: our enduring impact on our profession, and on society. At Imperial, as at QMUL and KCL, the concept of impact is huge. When Sir Francis Bacon wrote to King James with some neat ideas about the scientific method,[v] his method of persuasion was really that science, if wrested from the Aristotelians and the theologians, would have enormously enhanced utility. And this is, and always has been, the guiding philosophy of Imperial. The challenge is that scientific impact, in the 21st century, seems to go hand-in-glove with major issues about the relation between science and society. Big areas of interest at Imperial, such as machine learning, climate science, new vaccines, robotics, data science, transport studies, bioengineering and medical devices, security science and societal resilience, are each increasingly dependent on the social sciences. The advancement of these STEM topics depend on a sophisticated understanding of society.

The debate at Imperial is not really about the fact of the relevance of the social sciences – we can consider the matter settled – but the nature of Imperial’s dependency on the social sciences. Again to over-simplify, one of the questions we are asking ourselves at Imperial is whether, implicitly, we see the social sciences as an assistant to STEM – for example organising the focus groups that help us make sure our new robots get good uptake in society; or, whether, in addition, we see the social sciences as active in setting STEM research agendas, shaping it from the beginning, applying the cautionary note and being a critical friend.

Consider the important role engineers, and engineering, have always played at Imperial. Engineers are interested in optimising. There may be contrasting and even contradictory physical forces at play in a system but the role of the engineer is to remove those conflicts and arrange them so that things work together. A bridge is a case in point. All the forces must work in harmony, or else the bridge falls down. Is this how we feel about things at Imperial? Do we think that disciplinary conflict is bad, a sign of imminent failure? Perhaps our ‘received view’ is that the way to optimise our interdisciplinary work is to hope that in essence the social sciences and the STEM disciplines tread the same path. And maybe they do. Yet we should remember a very interesting statement from the 20^th century Australian philosopher John Anderson, who urged us ‘not to ask of a social institution “What end or purpose does it serve?” but rather, “Of what conflicts is it the scene”‘ In the same vein, Anderson continues: ‘For […] it is through conflict and sometimes only through conflict that we learn what our ends and purposes are’.[vi]

So we begin to see the outline of a question, quite familiar I admit, for Imperial. Is there a conflict – ‘an issue’ – between STEM and the social sciences, or are we happily building bridges and watching the traffic flow? Do we simply take as a matter of principle that conflicts between the social sciences and the natural sciences arise because of some easily solved misunderstanding? Or do we work rather harder to see what such misunderstandings may be, for fear that if we simply smooth them over, to keep things ‘optimised’, we simply fend off inevitable trouble, or perhaps worse, radically reduce the possibility of truly creative results from good interdisciplinary work.

Let me quickly turn again to the past, and to a story that is indeed suggestive of trouble. As you will know, this Doctoral Training Programme is funded by the ESRC, the Economics and Social Research Council. It sits along a handful of others: the BBSRC, the AHRC, the NERC and so on, all of them part of an umbrella organisation, the UKRI. You may have wondered, when you see the logo Economics and Social Research Council, whether perhaps there is a typo here. No-one says ‘Social Research’. We tend to say ‘Social Science Research’. Could it be that, somewhere along the line, the phrase ‘Social Science Research’ has been ‘disappeared’, to be replaced by ‘social research’. That is exactly what happened and let me tell you the story. In 1979 a new conservative government was elected, under Prime Minister Margaret Thatcher, a known radical, one inclined to doubt the value of the state, or the BBC, or the Greater London Council, or the Inner London Education Authority, trades unions in general, nationalised industry and so forth. Her tenure saw a great amount of conflict as she took on and battled with what we might call ‘the centre ground’, allowed the mining industry and the steel industry to perish, instituted what was called ‘monetarism’, and asked us to see the public finances as akin to an individual’s bank account. In her ideological push Margaret Thatcher had a number of influential advisors, of whom the most important may have been the cabinet minister Sir Keith Joseph (1918-1994). As Secretary of State for Education, one charged with making some of the cuts that the Thatcher government became famous for, he looked at his suite of research councils and decided that the axe should fall on one of them, the so-called Social Science Research Council, or SSRC. He was being asked to save money, it is true, but also he was vexed by this notion of ‘Social Science’. One of his intellectual prejudices was to doubt the scientific credentials, and the rigour, of subjects like sociology and psychology, and felt they must be kept distinct from the noble term ‘science’. He duly asked the CEO of the SSRC ‘Do the social sciences follow the Popperian paradigm’, by which he meant: Does the SSRC restrict its funding to work that adheres to the model of hypothesis and test model. As we all know, hypothesise-and-test, definitely a feature of science, may not be its only feature. In the 1970s and 1980s, in the heyday of Sir Karl Popper, the mark of science was its reliance on testable hypotheses.

On the whole, thought Joseph, the SSRC is a bogus institution that puffs up sociology and anthropology into great fields that needed large investment, but never test their hypotheses. He thought: let’s get rid of this absurd funding council, save a tidy sum, and stand down a clan of annoying charlatans. He admitted there probably was some good work going on in this area ‘the social sciences’, but very little, and these good people, he imagined, could perfectly well be supported by their Colleges, or by the British Association.

To his credit, Sir Keith Joseph, a scholarly person, did like to consult, and rather than simply wield the axe, in time-honoured British fashion he asked Victor Rothschild, Lord Rothschild, an eminent scientist and businessman, and an FRS, to look into the matter. Much discussion then took place and in the end Rothschild reported that state funding of the social sciences was an important matter, could be considered a valuable investment, and should be kept going. At which point Sir Keith Joseph, heeding the wisdom of those who know better, signalled a reprieve for the SSRC. But he had one condition. It had to change its name. The phrase ‘Social Science’ – in his view a contradictory term – must go. From now on the council, though allowed to continue, must be known as ‘the Economics and Social Research Council’.[vii]

The philosophical point Sir Keith Joseph was rehearsing was the status of different sorts of knowledge. He was touching on something that we all know as ‘positivism’: the belief that in the quest for reliable knowledge that can form the basis of enduringly successful decisions, numbers, objectivity and scientific observation rank the highest. It is sometimes called ‘the hierarchy of knowledge’ and sometimes called ‘reductionism’. The role of the positivist mindset was a theme of discussion recently, at Imperial, at a meeting I organised a few months ago to discuss this issue of the social sciences in a STEM institution. The room was filled with very talkative social scientists, all based at Imperial, all embedded in STEM departments. You will know that we have no departments of sociology, anthropology and linguistics; no department of history or geography; not even a department of economics or of psychology. Now, as I’ve suggested, we cannot doubt that the social sciences are significant at Imperial; but here is the question: what is it like, being embedded in a STEM department, doing social science work, but surrounded absolutely by scientists?  This was one of the questions we asked at the meeting.[viii] Another was this:  when it comes to social scientists flourishing at Imperial, can we rely on their STEM departments to provide the congenial and creative environment good academic work relies on; or should Imperial find additional ways to support an identified ‘community of social scientists’, perhaps in the spirit of your Doctoral Training Programme.

Here are some quotes from that meeting:

1. I think Imperial is exactly the place that can play a role in encouraging radical social and behavioural science research. [It’s] because we don’t have a dedicated social/behavioural science space.

2. I think what is special about Imperial is that we chose as social scientists to be here, not in an anthropology department or wherever, but instead came here specifically for interdisciplinary work.

3. Although it is good to be embedded in an Imperial College department, social science funders aren’t oriented towards STEM-based projects. This is really where the College could help us, establishing connections with the social science funders.

4. In my opinion […] there are entrenched power dynamics that position STEM as a “harder” discipline and social sciences as “softer” counterparts. While this position may be shifting, STEM continues to hold greater inherent value in many contexts across the College. Achieving truly equal collaborations demands mutual respect for the distinct and valuable contributions of every field.

5. The criticism I often hear from natural scientists is that social science projects feel like they lack rigor or authority and are “wishy-washy”.

This brings me back to the issue of ‘engineering culture’ at Imperial. On the one hand we are now a place with a significant social science presence, where investment in the social sciences is seen as a good thing. And in terms of mood music, we find in the new Strategy the strapline ‘Science For Humanity’, surely a sign that the social sciences are now in situ. In the meeting I organised, social scientists were united in seeing the value of being embedded in a STEM department. There may be work to be done culturally, but – so I heard – the concept of the embedded social scientist, working within a science department rather than a sociology department, can be just fine.

But – a big but –  it must be admitted that in this meeting there was a lot of discussion of the support communities of social science academics need: their ability to work together on methodological issues; their ability to press ahead, with a degree of autonomy, on their own new ideas; their desire to be involved in STEM research from the beginning, not just at the end.

Again, I quote:

1. A more integrated approach is required, where the social sciences are recognised as necessary in the conception, design, testing and implementation of any STEM innovation that aims to improve people’s lives.

2. We need to pool the methodological resources of science and social science, tear down the methodological and intellectual obstacles between them to move forward.

3. What is worrying, is that sometimes social science methods are used in science research, without including social scientists in the process, which may have implications for how we interpret the results of the research.

And now we get to the heart of the problem, which it seems must face all academics and universities seriously interested in interdisciplinarity. If people from across the campus are to work together, it is important to recognise institutional and philosophical boundaries – we could just as well call them ‘traditions’ – rather than pretend they don’t exist. To ‘optimise’ by failing to recognise the problem is just to jinx the research from the beginning.

Now Imperial, anxious to get this matter right, has set up a working group to look at these issues, and report to the University Management Board with some findings and some numbers on the role of the social sciences.[ix] I’m very happy to be on that group and it has been interesting seeing, at an institutional level, how the debate proceeds.

Quite a lot of the work so far has been – of course – ‘mapping’. Where is the social science work taking place, and what is its significance? I have noticed that while always at Imperial we will be drawn to numbers and to solutions, it must be admitted that at our Task and Finish Group,  discussions have usually veered in due course from matters of numbers to matters of philosophy. To me this axis, from positivism to something more interpretative, now seems inevitable at Imperial and is very much to the credit of the institution. And, when we do lay down our spreadsheets, important and interesting though they are, and turn to philosophy, the issue we come to first is the concept of utility.

Let’s turn to this point of the utility of the social sciences to STEM innovation, and also consider a related point – one Imperial is currently debating  – that the social sciences, working in interdisciplinary relation with STEM, are a source of critique, of innovation and the frame of an objective gaze that can help STEM see the wider picture.

On utility, here is another quote, from an Imperial social scientist.

Currently [at Imperial the social sciences are] more of an afterthought, eg “we created this beautiful solution (in engineering, medicine, AI, etc) now let’s convince people to use it”.

That’s what we mean by utility. And it is not wrong in itself. Of course it is not. Moreover, social scientists are no different from anyone else: They need to earn their living. And even if scientists do have quite a reliable route to the world of utility, academics of all types very much want their work to be valued by others.

Now I hope I’ve suggested that for interdisciplinarity to work well, the STEM/Social Science partnership can never be reduced to some simple transaction. Rather, it  is dynamic, ill-defined, and somewhat conflicted. Yet time-consuming and expensive though these tussles may be – dynamic, ill-defined, conflicted – it is this kind of academic work that will make our planet a safer place than it is right now, in terms of health, climate, politics and justice.

Let me tell you a few more stories, to show why I am quite confident about the rich opportunities the social sciences provide to STEM.

I first became interested in the natural science/social science relation when I was a school science teacher, in London,  some time ago. I’d studied zoology first, and my postgrad was in History and Philosophy of Science.  This was before the days of the National Curriculum, and for reasons I’m not sure I understand, when I started teaching there was much interchange between scientists, social scientists, and science teachers. When I trained, at a department in Chelsea College, which then moved here to Kings, we were expected to ask ourselves the question: why are you teaching science to these children, here in London? What will they gain?

It wasn’t enough to say: so they can go to university; or so they can become scientists. We were trained to ask ourselves, what is the value of science to people who don’t plan to study it a lot longer? And in searching for solutions, we had to look quite hard. Always in front of you were exams. When you are teaching science, there are things to be got through:  Newton’s laws, the digestive system, the electron transfer system in photosynthesis. But when I became a schoolteacher the National Curriculum – an idea of the Thatcher government – was not yet in place, and there was more room for teachers to follow their nose. In the freedom we had, issues of politics tended to loom large. With the Cold War still running strongly in the 1980s there was big defence spending on science; Poland was still part of the Soviet Union; we were very conscious of the way science is not simply a bunch of facts but mixes it strongly with geopolitics. Among many initiatives I remember a course available for Sixth Formers, to be used both by STEM students and humanities students, called Science in a Social Context (SISCON). I remember that in the introductory material the authors wrote that with science-and-society issues, expect disagreement, expect discussion. Those authors said that this disagreement and this discussion would very greatly enhance the classroom, and the quality of the science learning.

And that’s what I’ve always found, with the science classroom – that if you can step back from the grind of learning facts, and adopt the interactive and discussive mode, then the deployment of politics, or ethics, or philosophy, very quickly becomes central to the science learning. The same goes for Higher Education. In the university STEM curriculum the social sciences, or for that matter the humanities, should not be lecture 13, after 12 lectures on molecular genetics. The social sciences should be in there from the beginning.

In my experience,at Imperial, until about 10 years ago, the question of how we should deal with the social sciences and the humanities had elements of the long-running Two Cultures debate. We taught the social sciences and the humanities, to broaden our students’ minds, give them ‘soft skills’, and in some sense complement the specialisation of the STEM curriculum. By ‘Two Cultures’ I refer to an influential row, dating from mid-way through the last century, where the novelist and scientist CP Snow, giving the Rede Lectures in Cambridge in 1959, and comparing scientists to humanities scholars, said that in going from Burlington House in Picadilly  (the home of the Royal Academy) to South Kensington (the home of Imperial College) ‘one might have crossed an ocean’. For a long time, in spite of some rather obvious problems with Snow’s diagnosis, the ‘Two Cultures’ was a very recogniseable phrase for people planning a science training. As Snow said, his lectures had ‘touched a nerve’.[x] At Imperial for many years the social sciences and the humanities have therefore been taught, at undergraduate level. You cannot take a degree in history, or in philosophy. But you can take modules and gain ECTs for your degree

But what about the partnership of science and social science in research? What are the relevance of the social sciences to scientific practice itself, or to the research ambitions of Imperial?? The first time I began to wonder whether something serious was in the air was around ten years ago, following the Ebola epidemic in Guinea, Liberia and Sierra Leone. Wilkinson et al, from the Ethox centre at Oxford, put the matter this way: ‘More than any health emergency in recent times, the West African Ebola outbreak has demonstrated the importance of community engagement and the risks of doing it badly’.

Referring to the importance of engaging communities suffering from the virus, they wrote:

In the face of a deadly new disease, and an array of suspicious outsiders who were often dressed head-to-toe in protective suits and spraying chemicals, some people chose to cut themselves off from help. They threw stones at ambulances, rioted and, in one episode in Guinea, killed eight members of an Ebola prevention delegation. Reasons for resistance are multiple, ranging from contradictory messaging, unsafe and degrading conditions in hospitals, and histories of violence, extraction and corruption which fed fears that Ebola (or the chlorine disinfectant spray) was a means of ethnic cleansing.[xi]

This interesting experience was noticed at Imperial. Debate gained ground. Whatever might be the way to prevent, or halt the Ebola epidemic, biomolecular innovation on its own would not be enough. And three other areas of research, much emphasised at Imperial, soon enough showed precisely the same dynamic, where technical knowledge is just one aspect of successful innovation. Among many examples we could choose Covid-19, and the concept of vaccine hesitancy. You need knowledge of the virus, and you need knowledge of society, of political systems, and the media; climate change, ditto; and most recently, what we call ‘security science’ and ‘community resilience’. Just as we mark 80 years since the defeat of Nazi Germany, so once again we have war in Europe, as well as elsewhere. And with war in Europe comes the idea that wars are won by force of arms,  but also by ‘the resilient society’. And resilience it turns out, is not just a matter of hardened electricity grids and protected communication systems: it is also a matter of societal trust in institutions, in social cohesion, and faith in the motivation of our leaders.

Take one further example. AI research regularly wonders whether health consultation, including of care needs, could ‘be streamlined’ by machine-led interviews. Instantly you see the issues: might this be a welcome increase in efficiency; or an alienating intervention, further distancing the patient, or the elderly person, from the health service? We might want to keep an open mind on this. But its obvious that the success of research like this will depend not only on the subtlety of the algorithm, but also on the expectations, traditions and feelings of the user. And the way to learn about expectations, traditions and feelings is through the social sciences.

Dr Stephen Webster, Office of the Vice-Provost (Research and Enterprise)/Science Communication Unit

[i] The London Interdisciplinary Social Science DTP is a consortium of Queen Mary (University of London), Kings College London and Imperial College London: https://liss-dtp.ac.uk

[ii] Kuhn T.S. The Structure of Scientific Revolutions. Chicago, 1962.

[iii] MacIntyre A. After Virtue: a Study in Moral Theory. Duckworth, 1981.

[iv] Rorty R. Philosophy and the Mirror of Nature. Princeton, 1979.

[v] Jardine L. and Silverthorne M. (Eds.) Francis Bacon: The New Organon. CUP, 2000 (orig. 1620)

[vi] In: MacIntyre A.  ibid. p153

[vii] Posner N. Social Sciences Under Attack in the UK, 1981-1983. https://doi.org/10.4000/histoire-cnrs.547

[viii] https://blogs.imperial.ac.uk/the-good-science-project/2025/01/16/a-memo-on-the-role-of-the-social-sciences-at-imperial/

[ix] https://www.imperial.ac.uk/about/leadership-and-strategy/provost/vice-provost-research/vpre-led-initiatives/social-sciences-at-imperial-task-and-finish-group-sstfg/

[x] Snow C.P. The Two Cultures. CUP, 1998 (orig, 1959) p54.

[xi] Wilkinson A., Parker P., Martineau F, and Leach M. Philos Trans R Soc Lond B Biol Sci

2017 Apr 10;372(1721):20160305. doi: 10.1098/rstb.2016.0305

Read The role of the social sciences at Imperial, part II in full

A blog by Lilia Moreles-Abonce, Georgia Christie, and Katinka Hunter-Morris

A place like Imperial College – a leading global institution – has to be successful. We therefore need to know and understand, as a community, what we mean by success. What do we mean by ‘good science’ or ‘good scientists’ and are there conflicts between institutional and personal successes? What are the different criteria we should use to measure this? Actually, can we even measure these criteria, even if we can decide on them? Indeed, for a high-impact and competitive institute like Imperial, how best can we include the human touch within our metrics?

These questions and more were explored in the Friday Forum, ‘Measuring Science, Seeing Virtue’ held at Imperial College in December 2024. Around 60 people attended, from a range of academic disciplines and faculties. First there was an engaging discussion between three panellists, which was followed by a very lively audience Q&A.

Our panellists were:

1. Mary Ryan (Vice-Provost for Research and Enterprise, and Armourers and Brasiers’ Chair for Materials Science, at Imperial College). Well-experienced because of her current role, and her previous career, Professor Ryan stressed at once the need to have different methodologies for measuring different sorts of success. She also stressed, as a senior academic, her view that the act of measuring success, and its problems, is a community issue that matters deeply.

2. Stephen Curry (Emeritus Professor of Structural Biology). He served as one of seven College Consuls, and as Imperial’s first Assistant Provost for Equality, Diversity, and Inclusion. Professor Curry emphasised that these issues are complex as well as important. Thus, events such as this Friday Forum, where staff can debate together the collective criteria for success, are to be welcomed.

3. André Spicer (Executive Dean of the Bayes Business School). Professor Spicer reminded us that an important aspect of measuring ‘success’ is simply leadership: high performance occurs when there is a high level of trust within the university. 

Once the panellists had been introduced, each gave an opening statement of around eight minutes. Professor Ryan elaborated on different dimensions to measurements of success in science. Of course, as Professor Ryan said, university rankings are a factor here, because they have such a strong influence on how a university is perceived. And there are links also to the necessity of Imperial making sure it delivers what it promises. This suggests a commitment to mission and to objectives: naturally the extent of their achievement must be evaluated, which calls for measurement. But probing further, she noted that there seems to be a lot of measuring of the ‘whats’ but less of the ‘whys’. We must always remind ourselves that we tend to measure to measure the obvious things, and so are at danger of missing more elusive aspects of the academic life.

As a final note, Professor Ryan emphasized that we shouldn’t have a single, or simple sense of the measurement of the ‘life scientific’, as there are so many criteria and so many methodologies for looking at what scientists are doing. She therefore ended her talk by inviting the attendees to ask themselves constantly: “How do we make sure that as a community we focus on ‘what really matters’,  and how do we make sure we seek always the right measurements for supporting  ‘what really matters’?

Now it was Professor Curry’s turn. Imperial, he reminded us, is a global top ten university with constantly growing impact. With this success, he said, goes a responsibility to take a sufficiently complex view of rankings. He suggested that a fixation on output – on results –must always be complemented by an appreciation of the quality of the scientific process. Professor Curry reminded the audience that success, as a concept, is in truth somewhat elusive. There is the risk that the more we make success the target, the more we will miss the real thing. Success, he said, is multidimensional, and he talked of ‘weaving into the system’ good training, good leadership, and good mentoring. Get this wrong, and the human cost is high.

Please keep your values to the forefront, urged Professor Curry. After all most people come into science with a desire to make positive changes. Yet trying to hold onto one’s values throughout a scientific career is difficult. Ours is a competitive environment, he said, though also one that always we hope will be supportive. Professor Curry ended his talk by urging the audience to see the importance of talking to one another about what we might consider worthwhile, and worth cherishing, in the university life.

Professor André Spicer began his talk by describing the case of a seemingly prolific Spanish scientist, who was found to follow questionable practices in order to generate a remarkably high research output. With this example, Professor Spicer touched on the recent research showing that while overall research effort is increasing, the amount of significant ‘discovery’ does not have the same upwards trajectory. Such a proliferation of research may ‘pay off’ for individuals, said Professor Spicer, because of the influence of simple metrics, but overall the problem becomes the issue of people confusing the impressive career metric with the actual goal.

This is the dynamic we must be cautious about. The wrong kind of incentive, perhaps those that produce a simple ranking of scientists, leads to researchers getting misdirected in their focus, who thus succumb to cheating, the gaming of systems, and the ignoring of long-term thinking. More optimistically, Professor Spicer offered us six steps that can be followed to keep us from such destructive habits:

1. Get people involved in the development of measurement strategies.
2. Focus on narratives, as these tend to create long term measures, giving people a sense that variables – criteria of success – are attainable, given perseverance and diligence.
3. Loosen the relationship between incentives and measurement.
4. Use a wider range of measures.
5. Add some strategic uncertainty to the metrics, to help people avoid an over-fixation on simple criteria, and to discourage ‘gaming’.
6. Focus on careful metric design, so that the techniques and objectives of measurement are simple, fair, available, immediate, and reliable.

Nicely set up by these fascinating talks, we moved on to the audience Q and A.

interesting points that were brought up: we heard more about Key Performance Indicators (KPIs), interdisciplinarity, and the importance of relationships. The first question asked how social sciences, and particularly a narrative approach to research, could ever fit into evaluation metrics at Imperial. Professor Curry noted that this question is indeed important, as the amount of social science research at Imperial is increasing. Actually, Professor Curry pointed out, it might be that the social science capability we have at Imperial will become a resource to draw on when we evaluate impact and create metrics. On the same theme, Professor Ryan pointed out that if we only look at research’s impact after it is achieved, we automatically limit our vision, and fail to ‘see’ the quality, or the virtue, in science.

Another member of the audience defended KPIs, claiming that they can drive ambitions and assist strategy. Professor Spicer’s response was that balance is key in this: we need to weave between short term goals and long-term strategy. Professor Ryan agreed, and drew our attention to the difference between KPIs and outcomes.

Next, an audience member brought up the importance of relationships in the world of research. How do we maintain these relationships, and how do we measure them? All three panellists agreed that this is indeed an important question, touching on why we enjoy science in the first place, and all panellists agreed that this question is a difficult one to answer. Professor Ryan pointed out that relationships are not time-limited, and so differ from the way we organise science projects. This clearly is a problem when we try to recognise the importance of relationships in science. For example, a relationship might mature into some creative understanding of a technique or theory, and promise great things, while meantime the project time frame has long since finished, and the funding has dried up too. Professor Curry said that the work of relationship-building within research is often hidden and under-appreciated, and varies very much across Imperial. Professor Spicer also reminded us that in science we tend to look at the impact of relationships transactionally, as matters of exchange, but instead we should look at them in a richer, more human, more dynamic way.

The next question asked how a scientist’s doubts about metrics might affect their strategy for achieving promotion, especially when different departments have very particular expectations. Professor Ryan said that of course it makes sense that academics want to know what to do if they are to be promoted. Criteria, she said, are broadening all the time. She for instance mentioned that the way we think about promotion is shifting from solely a focus on research, to one that also has an eye on teaching. Professor Spicer added that we shouldn’t be making promotions based on some numerical metrics, whether via research or teaching, but must depend on a constantly re-elaborated set of criteria.

This Friday Forum itself was a success. The meeting facilitated a fascinating discussion, exploring the topic of how a leading STEMB institution, like Imperial, should measure itself. We saw for ourselves, as we talked over lunch, and then gathered for the discussion, the importance of creating spaces for in-person gatherings. Within the short hour that is a Friday Forum, diverse voices from academic disciplines and faculties could make their points and in turn draw comment. Imperial clearly has both the incentive and the internal gifts that to allow it to take a nuanced and sensitive approach to defining, seeing, and pursuing ‘success’ in science.

Read Measuring science, seeing virtue in full

I was happy this week to announce a third series of Friday Forums, and indeed a third year of life for the Good Science Project. Our first Friday Forum, on October 25th, is on the future of animal experimentation. Like all our Forums, this one will be a discussion of something close and intimate to the life scientific. And like all our Forums we will be able to step back from our busy lives for a short hour, discussing a matter of great importance to our lives as scientists, and gaining new perspectives.

The subtitle for the ‘animals’ Forum was provided by my colleague in Central Biomedical Services, Dr Anna Napolitano, who has been so helpful in setting up the meeting, finding speakers and setting the agenda. The subtitle is ‘Looking back, looking forward’ and is an excellent pointer to how the discussion might go. For in our work to ensure animal experiments are efficient in their outcomes, controlled by agreed conventions, and tied to progress in biomedicine, no doubt we should have a sense both of the history of vivisection and its projected future.

Ethics in science is always a mixture of technical fact and moral principle. With the animal model, the sense of moving forward in ethical discussion is very reliant on technical and methodological advances. Such advances are characterised by the ‘3Rs’, namely ‘replacement, reduction and refinement’, and have as their champion a campaigning organisation that promotes the very ideas we will be exploring at our Friday Forum. It is a very good thing that Imperial College now has its own, excellent website devoted to the 3Rs.

Stepping back a little, I can see that our ‘animals’ Friday Forum is a good example of how I want the Good Science Project – now in its third year – to work. I want our meetings to be properly sensitive to the daily priorities and pressures of daily science, and the things we call ‘ordinary science’. In the last two years the Project has therefore promoted a great amount of such ‘ordinary’ talk. We’ve argued about: the way technicians are important to science; the way scientists sometimes feel they have to be politically active; the way the design of laboratories matters; the problems of moving country to do your science; how your priorities change, or do not change, as you get more experienced in research. At the same time I have wanted to mix in with these quotidian matters much broader and elusive concepts: what do we mean by progress in science? What are the risks in science moving away from ‘blue-skies’ research to the high-impact variety? Can the arts and the sciences find points of contact?

In a recent blog I started to discuss a theme close to the heart of the Good Science Project. I told how our Vice-Provost, Prof Mary Ryan, had described my project as ‘ethics’. This had prompted me to go away and get clear in my mind how good science, and ethical principles, might mesh. As I described in that blog, the Good Science Project’s interest in daily science didn’t seem to match well the demands of utilitarianism, or rule-based ethics, which are the two strands of ethical argument we are most familiar with, and which typically are seen as the moral grounding of science. Instead, I asserted, it is virtue, or Greek ethics, that seems the better fit.

Can I explain this further, by looking at the subject matter of our first Friday Forum, on animals? Can I be confident that, if the Good Science Project is an exploration of ethics at Imperial College, as Mary Ryan suggested, the main philosophical tradition we should use comes not from Kant (rule-based ethics), or the Victorians (utilitarianism), but instead from classical Athens, and the work of Plato and Aristotle?

At first sight it seems likely that the ethics of the animal model will depend entirely on rule-based ethics, and on utility. For the way we describe the ongoing relevance of the animal model is clear. The results obtained from animal experiments are highly beneficial to human health and understanding. This lines up with the utilitarian idea that an act is good if it increases the sum of human happiness. And Kantian ethics are central to our understanding of the animal model too. For legislation and regulation – ‘Home office rules’ – are the basic grounding of good practice.

No doubt at our Friday Forum discussion of animal models these strands of thought will be mentioned. But it is my strong hunch that, when we have finished our conversation , and leave the Sir Alexander Fleming Building for some well-deserved weekend rest, it will be the classical Greeks, and the moral philosophy of Aristotle, that in some quiet way will be echoing in our minds.

For Aristotle, and for the tradition of virtue ethics, the point of focus must be on daily life and on the steady practice of our skills. This is no mere ‘turning of the handle’. On the contrary, for the idea of virtue to take hold, the importance is in the way we constantly enhance our skills, share our knowledge with others, see how best we can do our job, and challenge ourselves to work well. Aristotle describes this vision of virtue at length in his Ethics but for a more contemporary, and highly celebrated account, you should turn to Alasdair MacIntyre’s magisterial After Virtue, published in 1981. It’s a densely argued book, so you’ll need to take your time. It is said that MacIntyre tore up his first manuscript, and started again, after reading the highly influential text of philosophy of science, The Structure of Scientific Revolutions. Written by Thomas Kuhn in 1962, the book is famous for its elevation of the importance of ‘normal science’ and its suggestion that the heart of science is not its progress, but rather its daily practice. And Kuhn went so far as to suggest that if philosophers want to ‘understand’ science, it is laboratory life that matters, rather than simply its developing knowledge. In a sense then, MacIntryre’s subsequent promotion of virtue ethics owes a lot to new ideas then circulating about the life of science.

Let’s be precise. Where does virtue ethics get traction, when it comes to the animal experiment? Why might utility, and rules-based management of animals, be usefully supplemented by ideas dating back two and a half thousand years, to classical Athens?

When we gather in Room 121 SAFB, on January 25th, it is the attentiveness, and the craft knowledge, of our panellists that no doubt will be striking.  They will be triangulating ideas about physiology, biochemistry, and comparative evolution and anatomy. We will hear about animal husbandry and we will hear about veterinary science. We also will hear from our Imperial experts about the regulation of experiments, and the way the value of such experiments can be discussed in public. It is great that joining us is John Meredith, head of outreach and education at Understanding Animal Research. And no doubt our audience will listen, share experience, and ask important questions. This, quite precisely, is what is meant by ‘virtue ethics’: a community, learning together over time, sharing skills, and putting priority into what is in front of us now, rather than some imagined future. Take care of the present, Aristotle would have said, and the future will take care of itself.

Read New series of Friday Forums to discuss the scientific virtues in full

When Vice-Provost Mary Ryan said: ‘Yours is an ethics project’, I had to start thinking.

Two years ago I started The Good Science Project. My idea, as told to Mary, was simplicity itself: the research culture in which we flourish will be one where conversation and debate, on contextual issues as well as ‘scientific topics’, are nurtured. I would set up brief, in person, lunch-included, discussion meetings – the Friday Forums. Nothing much, but it would help. At this point Mary said to me ‘That sounds good, this is an ethics project’.

I went away and asked myself: is communication-within-an-institution, however thoughtful and illuminating, a matter of ‘ethics’? Is Mary right?

The history of the phrase ‘research culture’ gives us a clue why the topic might indeed be an ethical issue. For the previous word we used was ‘misconduct’. As I write in a previous blog, the ancestor to our interest in research culture is a concern about trust. At the start of the millennium a number of high-profile cases from across the world made journal editors in particular, and research centres in general, worry about cheating. To a large extent the concern at that time seemed to be trust in individuals. Are some scientists ‘bad apples’, and what should we do about them? The year 2000 was also the time when the House of Lords Science Select Committee produced their epochal Third Report ‘Science and Society’. The report began with statements about a ‘crisis of trust’ in science. No particular mention was made there of misconduct or research culture, but in highlighting the concept of dialogue between science and society, the idea that the internal workings of science has civic relevance was bound to gain ground.

Thus it was that a discourse grew about science ethics being as much about institutions as it is about individuals. In 2002 and 2003 influential MRC scientist Peter Lawrence wrote for Nature magazine well-received articles on publication norms and the nature of scientific esteem. A broad conception of the culture of research institutions was the force of Sir David King’s code of conduct Rigour, Respect and Responsibility in 2007. Perhaps of greatest significance was the Nuffield Council of Bioethics 2014 report The Culture of Scientific Research in the UK, chaired by Professor Ottoline Leyser (who subsequently went on to champion these issues when she became CEO of United Kingdom Research Innovation).

All of us know that ethics is about the difference between right and wrong, about how to separate benefit from harm, and about how to advance justice rather than injustice. But all this is very abstract-sounding. How do ideas like these get traction on a laboratory? If discussions about research culture classify as an ethics project, what precisely should we discuss? Why might people gathering to discuss perspectives on their work, at a lunchtime Friday Forum, be considered to be engaging in an ethics project?

Let’s do some homework. When it comes to serious descriptions of European ethical thought, three strands exist. At the risk of being dull, I will list them. Firstly, there is deontology, or rules-based ethics. Here you know right from wrong because of rules: religious rules; rules which seem self-evident (murder is wrong; dishonesty is wrong); and, famously, ‘the golden rule’: do unto others as you would wish others would do unto you’. The philosopher most associated with rules-based ethics is Immanuel Kant (1724-1804). In science we have plenty of rules: animals, health and safety, and any number of management, HR and policy requirements.

Secondly, there is utilitarianism. This is where you judge an action by its consequences. If, all things considered, the action makes the world a better, happier place, then that action is the right one. A corollary, and a slightly disturbing one, is that the action itself does not carry particular moral weight. For example, perhaps in some circumstances killing, or telling lies, is the right thing to do, because the consequences are seen as so important. Another way of putting this is ‘the ends justifies the means’. Utilitarianism is a 19th century movement especially associated with the social reformers Jeremy Bentham (1746-1832) and J.S.Mill (1806-1873).

In science we are well-practised in using rules-based ethics, and utilitarianism, as groundings for our work. As regards utilitarianism, the resources we put into science are justified because of the future benefits that will accrue to people. Consider our College strategy: it is called ‘Science For Humanity’.

Almost automatically we think of the values of scientific research as founded on rules and on utility. The rules we set ourselves, the truth of the scientific results that we produce, and the likely future value of our work, dominate the stories we tell about our work.

But, just to give us pause, accounts of what scientists value about their work seem not quite captured by concepts of future benefit, or adherence to rules. Scientists on the contrary prefer to describe their enjoyment in their craft skills, in their steady accumulation of knowledge, and in their sense of being in a community where trusted sharing of ideas is a norm. And when they discuss their concerns about the scientific life it is the distorting impact of intense competition, too hard a fight for grants, and fears about a secure future in science research, that gain mention. And the strong sense often is that these drivers, even if accepted as inevitable and manageable, are not considered as central to scientific practice. These are not the aspects of science that make scientists happy, and cause them to think they are making progress in their understanding. Instead they are a distracting burden from the main task, a tax. We can suggest then that there is more to the life scientific than rules and future benefit. And this is where the Good Science Project comes in, and where Mary’s words prompted its strategy.

For it turns out that a third and lesser-known branch of ethics is perhaps best placed for elaborating the actual lives of scientists – virtue ethics. This is a very ancient tributary of ethical thought, stemming from the classical Greeks. Here, it is character that forms the focus, especially as regards a person’s daily commitment to their work and to their growing skills. For example, to use an example from classical Greece, a ‘good’ farmer is one who understands seed and soil behaviour and knows what needs doing when. Getting good at all these things – the steady development of skills and knowledge, and with that the steady growth of reputation – is the ethical ground for this farmer. There is no emphasis on broader rules, or on consequences. In today’s language, you might say that it is the person’s ‘practice’, and the respect it gleans within their professional community, that matters.

For the Good Science Project, virtue ethics provides the best way of responding to Mary Ryan’s declaration, and indeed establishes the point that this is ‘an ethics project’. By finding ways to help scientists articulate those aspects of daily science that normally lie hidden from view – I mean the pleasures of the technical and intellectual challenges of daily science – we can claim to be followers of Socrates and Aristotle, the very founders of European ethics.

Read Ethics and The Good Science Project in full