Change of Address:

Hallo all.  As has probably been apparent for a while in this blog, my research interests have moved away from science and towards politics.*  This has been cemented recently by me a) handing in my PhD at the start of this month and b) beginning a new role (though still digital data analysis) \in the Cabinet Office (a branch of the Civil Service) earlier this week.

I’ve updated my online presence accordingly.  I’ll now be blogging as and tweeting as @olivermmarsh.  But I’ll be keeping notifications for this blog and my old @SidewaysScience handle active for old times’ sake.

Hope to hear from many of you at the new address. Send cards.

* yes yes I know there’s not clear distinctions between those two things but you know what I mean, sigh.

Posted in Uncategorized | Leave a comment

Posts from Demos Centre for Analysis of Social Media

I am currently spending three months at the Centre for Analysis of Social Media at the think-tank Demos.  Below are some posts I have written while there, largely around the UK General Election:

Activists, Not Bots: An Analysis of #VoteLabour Tweeters

Positive Online Debate: A Rough Guide

Taking Back Ctrl: Language around Social Media ‘Targetting’

Posted in Uncategorized | 1 Comment

Lecture on ‘How To Write A Good Essay’ for first year undergraduates

Created for and delivered to first year undergraduates on the Bachelor of Arts and Sciences programme at University College, run by Carl Gombrich.  If you wish to use or reproduce please contact me first.

Slides here



Posted in Uncategorized | 1 Comment

My Ongoing PhD Research Presentation: ‘Le Geek, C’est Clique? Descriptive and Emotional Meanings of Science in Online Fora’

The full slides, script, and recording for presentation given at Science In Public and 4S/EASST.  Please do not use without contacting me first.

Posted in Uncategorized | Leave a comment

In Partial Praise of Post-Truth Politics

In an odd way, political events of the last few months reminded me of the Cumbrian floods of 2009.  During this episode, news reportage showed endless footage of various political figures looking serious but sounding upbeat as they promised comprehensive relief and future protection.  The visible background of endless unexpected water, and the political background of the recent financial crisis, belied these claims somewhat.  Watching one such display my dad, with uncharacteristic frustration, suddenly exclaimed ‘I just want a politician who says “I’ll do what I can, but just look at it I really can’t do much right now”.  It’s a moment that’s stuck with me, largely because opportunities to remember it aren’t in short supply.

The EU Referendum, the Labour Party Elections, and the popularity of Donald Trump have all developed in somewhat obviously weather-esque manner – fickle, all-consuming, seemingly beyond human control, and altogether just making me just want to stay in bed with a pillow over my ears.  But what struck me was the similar string of endless, visibly, unrealistically over-optimistic promises from all sides.  But this time something was, apparently, different.  So great were the inaccuracies of these promises and the divergence of ‘facts’ presented by each side that we are now, it is claimed, living in a new ‘post-truth era’ of politics.


But how new is this, really?

For me (along with, I suspect, many sociologists) the idea and wording of ‘post-truth’ brought to mind the concept of ‘postmodernism’.  Postmodernism is notoriously hard to summarise, but roughly it is as an intellectual movement which rejects any notion of a single universal right answer.  It’s not alone in that,[1] but postmodernism is distinctive in taking a pretty Big Picture view.  Where (for example) moral relativists ask specific questions like ‘how can someone decide ritual murder is universally wrong?’, postmodernist writers talk about how whole societies come to believe notions like ‘progress’ or ‘expertise’ are universally true.  There have been some major difficulties with the postmodernist turn of thinking.  As I’ve written about elsewhere (as have other more interesting people[2]) postmodernism arguably fostered a self-indulgent approach to writing, stronger on showing the author’s erudition than welcoming readers into the conversation.  Following a series of spats across the last few decades of the 20th century postmodernism (or ‘pomo’) has become a dirty word amongst many public intellectuals.  This is unfortunate, given the issues involved; I’ve heard postmodernist writing described as ‘playful’, a worrying description for authors dealing with systemic oppression and the harms caused by our fundamental inability to find common reference points for talking to each other.

But I do think they were on to something. Aside from its more extreme examples of rhetorical or moral difficulties, I think postmodernism does argue something really important for our day and age: you can’t just strip away all the ‘well they would say that wouldn’t they’ or the ‘it’s media bias’ or ‘Twitter echochambers’ and get at some core of truth underneath it all.  You can’t say a situation is ‘because capitalism’ or ‘because immigration’ – the answers will change depending on where you look, how wide a view you take, what you focus on.  And they don’t somehow average out to one true answer.  Sure, there’ll be little factual nuggets buried in there – even Twitter can’t change what Jeremy Corbyn had for breakfast this morning.  But things only become valuable or useful when embedded in contexts.  To give a simple example, whether 2+2=4 is ‘true’ would be a pointless argument in a non-numerate society.  And context is the complicating bit.

That’s not to say we shouldn’t condemn outright lies, or deny that the real world has any value – in George Orwell’s 1984, Winston Smith’s most powerful act of defiance is holding up two pairs of fingers against The Party’s assertion that two plus two is five.  But I’m going to argue that public debate shouldn’t be so strongly geared around trying to produce ‘truths’ for things that, if everyone is honest, cannot be given a single truth.


Politicians can’t produce the truths they claim to produce.

Let’s start with the wearily predictable example of that claim made during the referendum campaign that the UK sends £350m a week to the EU.  Pedantically, it’s obviously untrue.  It’s a suspiciously round number for a start.  And when we consider the amount of variation which needs to be averaged out, the number and scale of assumptions, and the uncertainties in all the ‘I suppose this benefit is probably worth £x to the economy, ish…’ calculations, the difficulties of coming up with a single ‘true’ figure of how much we send to the EU are vast.  I expect that all calculators, relayers, and receivers of this claim would probably accept that.  But the problem goes deeper than inaccuracy.  The ‘truth’ would depend on the time window you bounded it by, the people asked to do the analysis, which information which happened to be used for input, and much more besides (including the ‘truths’ claimed by every single prior analysis that this analysis built on).  That doesn’t seem a particularly singular truth to me – more a product of decisions, which could have produced a different answer.  For balance, the same can be said of Remain’s claim that UK families would be £4,300 worse off if we left.  Or non-numerical[3] stuff like Tony Blair’s counterfactual history in which ISIS appears even without an Iraq invasion, or who exactly crashed the economy in 2007/2008.  And any large economic, political, or social question – or anything that actually matters to people’s lives.

I’m not saying these sorts of investigations shouldn’t happen, or that results shouldn’t be reported.  Orders of magnitude and trends are still useful.  But they’re rarely the whole story.  To give a non-political analogy, think of how much sex people are having.  Once you’ve done that, think of surveys thereof.  You’ll probably accept that the results are as much to do with people systemically exaggerating or downplaying how much sex they’re having (depending on gender, generation, social views…) as they are to do with frequency of sexual activities.  We could spend lots of time arguing about what the numbers should be.  Or we could spend some time doing that, and some time discussing the more valuable questions buried within all the above.  Does it seem that inaccuracies are more likely to be exaggeration or downplaying?  Did that shift, and when?  How does it vary between populations?  And what does it say about social values, the things society encourages and discourages?  These are not simply corrections; they are important discussions that reveal a lot about what people want, fear, or value.  Those discussions, I’d argue, haven’t been given as much campaign time as the narrow ‘you’re WRONG’ ‘no YOU’RE WRONG’ interchanges.  What might it have looked like if those wider discussions had been allowed?


‘The Truth’ is a powerful tool for dismissing people

I’d argue that[4] focussing on questions of establishing The Truth quickly narrows down the possible scope of debate – the EU is or isn’t a net economic contributor to the UK, immigration does or doesn’t have net benefit for the UK, and so forth.  This risks being pointless, as the UK doesn’t vote – individuals vote, and they most likely have their own views on whether immigration is good or bad for them already.  Trying to tell them their truth is wrong isn’t likely to do much except polarise; but trying to present alternative truths might open up debates further.  Imagine telling that sex survey respondent ‘there’s no way you’re having that much sex, I’m putting this lower figure’.  Now imagine telling that same respondent ‘thanks for that. That’s quite a lot of sex; just to let you know that people I’ve surveyed before have said trying to sustain that amount of sex made them unhappy in the long run.  So keep an eye out for that’.  It doesn’t deny the respondent’s truth; it accepts it and presents them with another one they might not have had otherwise.

Analogously, Remain’s time could maybe have been spent re-formulating the immigration question so it becomes ‘you feel negatively impacted by immigration, so let’s talk about how we our plan for dealing better with its impacts, a plan which BTW would be helped by us staying in the EU’.[5]  Or perhaps coming up with a less vague expression than ‘we’ll lose our place at the table’ (not convincing to people who don’t like the table anyway) – some way of giving us an insight into another truth about what it’s actually like being at that table, and another truth of what their job will be like if they’re not at it any more.  A truth that isn’t immediately accessible to other people, but is reconcilable with our experiences of being or not-being at various tables.  Something that challenged those claims of ‘we would get all the Sovereignty’, not by saying ‘no we wouldn’t’ but instead by exposing the simplistic vagueness, naïve positivity, and irresponsible over-optimism of such a claim.  Or outside the EU issue, the general election claim that ‘trickle-down economics means tax breaks for the wealthy mean everyone will definitely be better off’.  Or pro-Trump claims that a man with his, erm, forceful personality will obviously be great (just really great) at all Presidential jobs – Clinton’s campaign seems to have moved towards the ‘is that really realistic’ approach recently with some success.  But this, I’d argue, isn’t just an alternative spin tactic.  It’s an acknowledgement that in politics things are ‘wrong’ because they’re harmful somewhere, not just because they’re inaccurate.


My Partial Praise of Post-Truth Politics

My overall point: We should not value a public debate which even tries to give unambiguous answers to basically unanswerable questions.  Instead we should value a less precise but more realistic form of honesty, with more nuanced claims which are open about the lack of ‘answers which will be 100% right for everyone’.  This is optimistic: in our current media landscape there’s a huge risk of a dishonesty arms race in which those prepared to ‘ignore uncertainties’ (to put it euphemistically) win at getting their claims accepted more widely.[6]

But for now, the rise of post-truth politics need not be seen simply as a damning indictment of the dishonesty of politicians and the gullibility of voters.  Instead, it acknowledges that politicians and experts do not hand down ultimate truths to the blank minds of an uncertain public; they speak to people who already have their own truths, brought about by a combination of life experiences and personal values.  This is not simply some annoying barrier of ignorance or prejudice.  This is symptomatic of that really deep problem, of how we live with and for one another when our lives are so incomprehensibly different.  Like postmodernism, post-truth politics could turn this deep problem into an important shift in thinking.  So unlike postmodernism, let’s make it make this shift a wide, welcoming, and positive one.  Value people who understand what it’s like to be hit by a flood; value people who understand what it’s like to be tasked with stopping a flood; don’t value people who simply claim the floods will be stopped.  That’s the very best way to ensure they won’t stop ever.



[1] See also relativism and some branches of social constructivism.

[2] For a view supportive of this kind of writing, see this piece by Judith Butler

[3] For these examples, the decisions become things like ‘does a stronger Al-Qaeda basically count as another ISIS or are the differences significant?’ or ‘how far back should we trace the casual chain of crashing the economy, lest we say it was the fault of the Universe for existing?’.

[4] You’ll notice this, and many other sentences in this and other posts I have written, would probably be improved by removing such acknowledgements as ‘I’d argue’.  That this would be an ‘improvement’ is, again, not a universal truth but part of a problem with our cultural expectations of how people should write, read, and argue.  Or that’s what I’d argue, anyway.

[5] I’m aware that theory is much much easier than doing that in practice – nonetheless, this programme gives an interesting insight into the decisionmaking processes behind the Remain campaign, which does suggest they were convinced that accurate economic reportage would win it

[6] Again, this is not a universal truth; it is entirely possible to envisage a cultural revolution in which caution and self-criticism become greater values than snappiness of soundbites.  But I see little evidence of that occurring. Jeremy Corbyn and Ed Miliband seemed to have tried this approach with little success, though personally I’d have more respect for Jeremy Corbyn if he directly acknowledged the criticisms that he lacks personal leadership skills, rather than default to his ‘facts’ that everything is ok because Labour won x in council election y etc. etc.

Posted in Uncategorized | 3 Comments

Do these lenses suit my face? (Or, methodology as a lifestyle choice).

The language of social sciences has a problem with right levels of drama.  We reserve the driest of terminology – ‘positivism’, ‘constructivism’, ‘relativism’ – for describing some pretty fundamental human processes of making sense of the world.  We then describe differences between these labels as ‘struggles’, ‘competitions’, or even ‘conflicts’, all of which suggests two large well-opposed sides confident in their preparations and even more confident in their beliefs.  It does not suggest an image of me surrounded by dirtied coffee cups and PhD data printouts upon which I have scrawled ‘I want to believe something but I don’t know how’.

The point I want to make today is this.  That which social sciences call ‘methodology’ are, at heart, tied to some deeply personal factors.  If you’ve never heard of methodology, it’s the intellectual process of justifying why you did certain things during research.  If you have heard of methodology, you’ve probably been taught about it as communal process, with a researcher becoming encultured into one (or, though more rarely, more than one) of multiple schools of thought.  All schools agree that trying to understand humans and society is quite a complicated task – I’m sure there are people who think that humans and society can be explained using straightforward rules, but I will suggest that these people probably haven’t done much social science (or, indeed, interacted with many humans).  But one of the key issues underlying differences between schools is the following question: how do researchers stand in relation to ‘the real world’?  To illustrate that, have a picture I did at a Drink & Draw meet:



(Meanwhile the person to my left was doing an ink portrait of – I eventually realised – me, and the person to my right was producing a vivid sketch of the Yorkshire Moor.  I’m not a regular.)

As an example of the above, take the question ‘is country A happier than country B?’.  It’s not a straightforward one, but one way to approach it might be to compare suicide rates.  That way, you can (partly) represent the complicated and personal phenomenon of unhappiness as a series of numbers which everyone can agree on.    Not so, say others.  Even the seemingly objective ‘rate’ is actually the product of various social processes.  Different countries/coroners may record suicide rates differently.Ok, say the positivists, but in that case how do you give any answer to the original question beyond ‘unhappiness is complicated and personal?’.  More importantly, how do you give any answer which has any form of reliability (assurance that different parties will agree on the same answer) or validity (assurance that the answer does in some way match up to the world)? Ultimately, how to ensure different researchers can’t just give entirely different answers to the same question, thus reducing so-called ‘answers’ to personal opinion?

That’s a *heavily* summarised social science 101.  My argument here is that these questions go beyond the intellectual and into the rather personal.  To again raise the point of language, we tend– ‘naïve positivism’, ‘anecdotalism’, and suchlike.  Based on my position on that spectrum, the slur that wounds me is anecdotalism – the claim that if you’re going to allow researchers to just give idiosyncratic, rather than standardised, views then all findings are basically just personal stories.  (This criticism often appears as the phrase ‘the plural of anecdote is not data’).  I feel this is an argument which is strong on dismissive language and weak on clarity.  If I was to tell you the anecdote ‘hey so I was in my lab the other day, and I was doing this stuff, and this thing happened which I thought was interesting so I did the same stuff a few more times and, y’know what, the same darn thing kept happening’ then I’m basically reporting an experiment in a rather informal way.  Less facetiously, all research comes ultimately from people encountering stuff and relating it to other people and the label of ‘anecdote’ does no work in distinguishing how different fields actually do research in practice.  Instead, it’s used to dismiss a stereotyped or idealised view of some research that’s different to yours.

On the flipside of this quite negative view, ‘ideals’ also play a positive role in giving things to aim at; for example, many positivists might accept that 100% validity and reliability is unachievable but nonetheless worth trying for.  But a problem I have as a constructivist is that it’s not easy to reject ideas like ‘validity’ or ‘reliability’ when it’s hard to find words which bolster an opposite position – apart from ‘invalid’ or ‘unreliable’, and those labels don’t do wonders for one’s confidence.  And confidence is important in all this.  That earlier diagrammed question can be more brutally phrased as ‘what’s the point of the researcher?’.  What’s the point of all our training, lengthy and sometimes painful and often publicly funded?[5]  For those towards the positivist end, this teaches a researcher how to recognise and extract relevant information from the world and to submit these in the ‘correct’ way for peer scrutiny – in other words, they learn how to show validity and reliability.  But if you don’t believe in validity or reliability or some ultimate ‘correct’ answer, then what are you learning to do?  Are you learning pragmatic skills of how to find and report interesting information, thus making academia a quite extreme and laborious version of long-form journalism?  Or are you learning how to present your individual take on the world in a way which makes colleagues go ‘that’s interesting, ergo you are interesting’ – making research a form of what is often called ‘modern art’, where a small group stand around exclaiming ‘what a clever work, let me now explain why it’s clever in a way which shows you that I too am clever’ while a larger group claims ‘my five-year old could have done that’ and a much larger group simply neither know nor care about it at all.

I’m still not confident on the answer to that.  But look closely at the above paragraph – I’m talking about researchers, long-form journalists, and modern artists in exactly the same idealising and dismissive way as I’ve argued positivists talk about constructivists.  I’m actually not intending to dismiss long-form journalists; my point is that they do their work without the hefty training seemingly required by academics, so why can’t academics?  I am being dismissive of the modern art community, while being conscious that my description is of a stereotype rather than first-hand experience (and also a stereotype that I am actually worryingly adept at slipping into).  The point is this.  My idealised image of the long-form journalist is of someone who presents their reader with a clear, convincing, and detailed picture of a world, while at no point suggesting this is the only or best picture of that world.  This is a description I would be happy to align myself with.  By contrast my idealised image of the positivist researcher is of someone who wants everyone to see the world (or a portion of it) in the same way as them.  I feel this to be a bit simplistic and (potentially) morally problematic.  And my idealised image of the modern art practitioner is of someone more interested in promoting their own personal brand than in actually welcoming input from the wider world, or at any point asking ‘is this possibly just a bit stupid’.[7]  And all that goes a long way to explaining where I place myself on the above spectrum.  That’s not to say I wish positivists and modern artists didn’t exist, or that I automatically dislike their contributions, or that I don’t want to talk to them.  I just don’t want to be one of them.

None of that will appear in my PhD methodology chapter.  None of it would be accepted intellectually.  I think that’s a shame.  Because, setting aside differences between methodological schools, my ideal of academia is of a community that understands the implications of language, avoids uninformed dismissal, and welcomes full honesty and candour.  And whether or not you agree with that, any academic’s methodology has to proceed from the belief (even hope) that some world, or some portion of it, can be understood somehow.  Otherwise we’re a bit pointless.  And that is deeply personal.


(The title, fyi, is a reference to an old metaphor about methodologies being ‘lenses’ one can take on and off to see differently. I suggest it’s not as easy as that)


[1] This example is based on my (secondhand) knowledge of debates around Emile Durkheim’s book SuicideThere’s also an example about vote counting involving some social decisions about what counts as a spoiled ballot, but I’m struggling to find the ref for that.  Send help.

[2] Somewhat ironically I feel, or at least rather annoyingly – see my previous thoughts on clear language

[3] Or alternatively, in a 17th century way when experimental method was becoming a Big Thing. A lab scientist might respond with the point that experimental results are corroborated by replication from other parties, so they’re not so much personal anecdotes as shared stories.  Fine in theory, but in practice that doesn’t really happen.

[4] I’m sure the same is probably true of ‘naïve positivism’, for the record.

[5] Apart from, y’know, providing universities with a pool of cheap teaching labour and seminar organisers.  (This is, somewhat ironically given the main body of this post, a very dismissive way of describing quite a multifaceted problem).

[6] I.e. in the stereotypical sense of art that comes across as more weird and pretentious and of questionable aesthetic value  Apologies to any real, rather than stereotyped, modern artists. One could also think of niche music groups, fashionistas, or any of the people brilliantly, if often disturbingly, satirised in Charlie Brooker and Chris Morris’ Nathan Barley.  

[7] Again, I refer you to Brooker and Morris (ibid.)

Posted in Uncategorized | 1 Comment

Never Explain Pilot Recording

I had the wonderful experience of being on the panel for the pilot recording of a new science-ish (emphasis on the ish) comedy podcast ‘Never Explain’.  Future recordings and ticket details here .

Was a lot of fun, highly recommended.


Posted in Uncategorized | 1 Comment