Official stats, and how to publish them - a post with Taylor Swift

I’ve previously discussed on here how I found the portrayal of one particular set of government statistics – the Scottish Index of Multiple Deprivation – problematic. To summarise that argument – reporting, based on government press releases, called certain neighbourhoods “the poorest”, even though IMD does not measure poverty, is a relative measure, and most problematically, this added to the stigma of the neighbourhoods. I covered some of the impact of this in one of the papers I’m most proud of. So when I was contacted by the statisticians working on SIMD16 to come and have a chat with them about how they could present it in a better way, I leapt at the opportunity. When their report came out I was extremely impressed. You can download it here – it’s quite a big PDF. The things I particularly like were: 

• The very clear statement of what SIMD measures and how it works, and can be used, on page two 
• The case studies that demonstrate that living in a neighbourhood ranked as deprived by SIMD isn’t awful, and that it can help target resources for great things to happen 
• The very clear “do not use SIMD for” list on page six 
• The fact that nowhere does it say which is the “most” or “least” deprived neighbourhood. 

The team that did SIMD were nominated for a UK Civil Service Award, which they won!

Here’s @PermSecScot congratulating member of Scottish Index of Multiple Deprivations team-winners of #csawards Use of Evidence award. Her pride is plain to see! pic.twitter.com/1uYo8ULhiB

— Suzannah Brecknell (@SuzannahCSW) November 23, 2017

I was incredibly pleased on their behalf, because they really deserve it.

via GIPHY

And, they’ve only gone and done it again – not won an award, but produced a report that is staggeringly good.

via GIPHY

It’s a report on a set of experimental statistics that have used new questions on material deprivation asked in the Scottish Household Survey to create a local poverty measure. Have a look at it here. It is soooooo good. 

Again, they’ve combined qualitative data with the presentation of the statistics to make the reality of lived experience come to life, but in a non-stigmatising way, for example on page six “Mary” describes how: 

“My kettle blew up, so I went and got a kettle off my catalogue. Cause I wouldn’t have been able to afford to just go and buy a kettle. And I didn’t want to say to anybody, ‘I can’t afford a kettle.’ Ken, people are coming in for a cup of tea and that, and ‘oh my kettle’s blew up, and I can’t afford another one’” 

An absolutely brilliant way to describe what material poverty means. 

And it gets better. There are bar charts throughout which show a percentage with confidence intervals. Rather than getting bogged down with complex descriptions of confidence intervals and statistical significance, the charts are simply labelled “The bars show measurement uncertainty. Where two bars overlap, there may not be a real difference between the two groups”. I mean!!!!!!

via GIPHY

AND AND AND it gets even better. Not only do they have CI’s clearly labelled, they then go onto interpret them for you. Each chart has a very clear description of “What the data says” and “What the data doesn’t say” to ensure that people don’t misunderstand the charts.

via GIPHY

And the data is really interesting. I just wish all official statistics documents were published this well, with this much thought put into their presentation.

via GIPHY

With thanks to the Taylor Swift Open Data Institute.

Unsolicited cock pics and some wanted attention

I’m currently doing a research project on LGBT+ experiences of homelessness and living in the most deprived neighbourhoods in Scotland – more details here. It has been funded by the British Academy and is quite small. I was aiming for 30 participants in total. I have found it really difficult to find participants, particularly young homeless people.

I did expect this to be the case – I’m working with a subset of a small(ish) population. Estimates are that a quarter of young people between 16-25 are homeless at some point, by the broadest definition (i.e. falling out with their parents and crashing on a mate’s couch). The Albert Kennedy Trust suggest a quarter of this quarter identified as LGBT+ (although there's quite a few problems with that). People often don’t realise they’ve been homeless, and might not be out, so might not define themselves within the target population. I figured using homelessness organisations would be a good gatekeeper organisations.

I contacted quite a few homelessness organisations and still struggled. So, I thought, where do gay people hang out and I might be able to recruit them? Grindr!

For those of you who don’t know, Grindr was one of the first dating apps. I believe it was actually invented by a straight guy who was impressed with how easily his gay brother managed to hook-up in the places he visited. You set yourself up a profile – picture optional, but it can’t be rude – and fill in other details (including your “tribe” [otter, twink, bear etc. I’ll leave you to Google, but don’t blame me]; your preferred sexual practices [it seems versatile bottom is where it’s at, again Google if you don’t know], as well as age and height and other usual stuff. It’s location based, so when you open the app you get the photos of people in a grid with distance increasing as you go down. There is also a “Fresh Faces” bar at the top which lists recent joiners. You don’t swipe people away like Tinder, so it’s more like a chatting app.

I hadn’t intended to use it for participant recruitment. In fact I’ve been in a long-term relationship for 11 years, long before Grindr had been invented, so I had never actually used it. I had used the precursor website Gaydar (I can almost hear the wistful sigh from gay men of a certain age when you mention that) so I sort of knew what it was like. Anyway, I had to apply for a change to my ethics permission (form here for info) from my University’s General University Ethics Panel to use it. Before I got to this stage, Mark Holton at Plymouth University mentioned I should also check the Grindr terms and conditions. These are quite explicit on not using it for marketing, or to recruit people actively, but I wasn’t going to be doing this – I couldn’t tell if anyone was suitable for my research from their profile, so this was more lurking and seeing who approached me.

When I submitted my ethics application it was the week after Stephen Port had been convicted for murdering four young men he had lured to his flat using Grindr. This meant the main focus of the application was actually the risk assessment, particularly since it was a straight male colleague who would be doing the interviews – making sure that people recruited this way were met in public places and that I’d talked to them briefly over the phone to check everything was ok. This was on top of the usual assurances, such as making sure the researcher checked in-and-out-of interviews by text message. I also had to ensure the participants were consenting properly with full information. I was also aware there was a reputational risk to the University – I could chat people up on the app as a representative of my University. Therefore I also set out that after people messaged me on the app, I would move the discussion to email as soon as possible and provide them with full participant information then. The usual participation and consent procedures then kicked-in.

So I set up my profile, example below:

I changed it from that to asking if people lived in particular neighbourhoods in Edinburgh and Glasgow depending on what I was up to and where I was. 

And, the burning question – how successful was it? Well, quite successful actually. I got 49 contacts by the time I wrote this blog post. Not all of them were interested in my research. More of that later. It took me a week to work out Grindr etiquette. I was expecting people to message me and say “I’d like to participate in your research.” However, you only ever start a Grindr conversation with “Hi”. My initial reply to these was an immediate “I’m only on here to recruit research participants”. After a couple of times it became apparent this was extremely rude, so it became a “Hi, are you interested in participating in my research?”

Quite a few of the contacts did fit the criteria, particularly people who had experienced homelessness, which was exactly the population I was looking for. The main problem was turning these into actual interviews. I found this paper about a sexual health project that had used Grindr to find people for a project about men-who-have-sex-with-men. They noted that they got a higher participation rate if they phoned people, not emailed them. I offered to phone or email in my message to people who said they would participate. They all wanted to be emailed. I would email. And then hear nothing. I have no idea why.

And yes, I did get people who just looked at my photo (I used a professional work photo where I do look quite cute) and messaged me for the very purposes Grindr was invented for. And so far, yes, this has also involved some extremely explicit messages. As a gay man, I was sort of expecting this to happen, but if you were to recreate this method of participant recruitment and went into this naïve and oblivious it would’ve come as a bit of a shock. A handful of other messages I got would’ve raised eyebrows out of the context, but with all of these I just ignored the messages, or if someone was a tad too persistent, I politely ended the conversation explaining I was only on the app to recruit research participants. 

So, is it worth the effort to find research participants? LGBT+ people are roughly three per cent of the population, so they are pretty hard to reach. I have exhausted every single recruitment technique, apart from flyering gay bars on a Friday night, with this project and have still really struggled. Unbelievably even snowball sampling only got me two other participants (and one of them was someone’s partner and another someone’s sibling). Riffing on Michael Rosen’s book Bear Hunt I’ve been joking that “I’m going on a gay hunt, got to catch a big one”* and it really has felt like that. I think, like the public health researchers, if you were looking for people who were dating, or after no-strings-fun with a 9 inch top, for your research population, then it is definitely the place to find them. Otherwise, I think my research was a little too leftfield for your average Grindr user. They’re on it for a quick shag, or to meet the love of their life, not to discuss their housing circumstances. 

I also swiftly realised there is ethnography to be done on how people present themselves on Grindr, and the rituals of introducing oneself on the app. A quick Google Scholar search identifies that this has been done. This raises the issue of how “public” social media is (as did this blog post). An ethics panel would probably not give you permission to recreate Laud Humphries’ famous Tearoom Trade participant observation of cottaging. Yet, we could probably do an ethnography of interactions in a gay bar with no issues, comparing, say, how gay men approach each other with how straight men approach women. People do not put themselves on Grindr to appear in a research project, and if I provided any more detail about some of the profiles I’d seen, people could feasibly identify them. A defence of being overly cautious on these issues is we live in a heterosexist society. As I am finding out in this project, we know next-to-bugger-all (pun intended) about the lives of LGBT+ people in very ordinary ways. As a social scientist, I believe the only way to do this is to do research, and this involves identifying people as LGBT+ and then asking them about stuff. 

Anyway, this is all I have to say on the topic so far. This will be written up as a proper methods paper, don’t fear! Oh, and as I tweeted a while back – an idea for a novel: the protagonist is one of the random people that appear in someone else’s online dating profile, and a North-By-Northwest style mistaken identity adventure then starts.

*knowledgeable readers will get the many-levels on which that joke works. 

How many LGBT+ people are homeless in the UK?

It would not be an understatement to say we are currently experiencing a homelessness crisis in the UK, especially youth homelessness, particularly due to cuts in Housing Benefit and wider support services. We need to be campaigning and working hard to ensure these changes are reversed, or proposed changes are not implemented. There is also a broad debate about what should be done, although some of the proposed solutions are not necessarily based on the best evidence. 

One particular aspect of youth homelessness has quietly grown in prominence - the risk and experience of homelessness among people who identify as gay, lesbian, bisexual or trans- (LGBT+)(1).

Due to this, there’s a statistic that is used in a very commonplace way in the UK – that a quarter of all young people (16-24 year olds) that are homeless identify as LGBT+. It most recently turned up at the start of an episode of Queer Britain on BBC 3 to frame the rest of the programme. In this post I want to question this statistic.

First of all though, whatever the real number is going to be, it will be almost impossible to accurately get because the two categories involved are difficult to define. I’ll start with the relatively easier one: homelessness. I discussed this in a podcast with my colleague Dr Beth Watts, if you want to know a bit more. Basically this is a definitional question. The UK has a legal definition of homelessness (it varies between the devolved jurisdictions) because when you “fit” into this definition a housing authority (usually your local council) then has a duty to house you(2). Outwith Scotland, this is associated with being “priority need”. Housing authorities do returns on how many people they have assessed as homeless using these definitions which are collated nationally. Local authorities also provide housing advice and homelessness prevention services (such as family mediation). This includes a wider group of people who would not necessarily be homeless under the legal definition.

We know that these numbers massively under-report homelessness because we know the majority of people who experience homelessness don’t access local authority housing services, or actually define themselves as homeless. An additional issue is that most researchers would define homeless as not having somewhere safe and secure to stay. Therefore, people who fell out with their parents and sofa-surfed for a few nights would have experienced homelessness under this definition. Asking survey questions about this is a little tricky, and unlikely to get data that is useful for statistics. This data is collected at a population-level (the Scottish Household Survey has a fiendishly tricky question on it), but you get a very small n (number of respondees) so you can’t do much  analysis with.

Now, sexual identity. This is extremely difficult to measure, and probably not for the reasons you think. Epistemologically, many queer theorists would legitimately argue you shouldn’t/couldn’t ask this question, as you’re asking people to put themselves into categories that were invented by a patriarchal, heteronormative society to oppress people. It is wrong to ask people to ascribe themselves labels - such as "homosexual" that they vehemently resent and could cause them harm. The less theoretical issue is a technical issue for survey design: what people do and what people say they are is often a different thing; or as my colleague Dr Kirsten Bessemer put it, it’s when a man says “I’m not gay but my boyfriend is”.

The NATSAL survey in the UK demonstrates how this is an issue among young people. Most population surveys in the UK find that c. 3 per cent of the population are not-heterosexual. NATSAL asks about sexual experiences, romantic attachment and sexual identity and gets wildly different answers. It found that 3 percent of men and 4 per cent of women aged 16-24 would describe themselves as gay or bisexual. However, it found that 7 per cent of men aged 16-24 had had a sexual experience with a man and nearly 3 per cent had had at least one sexual partnership with a man in the past five years. Among women aged 16-24, nearly 19 per cent had had a same-sex sexual experience, and 6.2 per cent had had at least one sexual partnership with a man in the past five years. Based on all this, I’m going to be generous and say 5 per cent of 16-24 year-olds are LGBT+.

I’m using the acronym LGBT+ but I’m only really talking about sexual identity. So, an apology to T+ people – I’m not ignoring you, it’s just there is absolutely no data on gender identity, or other complex sexual and gender identities.

Anyway, we don’t know how many people experience homelessness, and we don’t know how many people are LGBT+. So how can we get anywhere near the 25% figure?

In this section I am going to use statistics from England because a) they’re available b) as the biggest nation in the UK, England is most like the UK statistically and c) because Scotland’s homelessness law complicates things (see note 2).

Mid-year population estimates from the Office of National Statistics tell us there are roughly 6,192,870 16-24 year-olds in England. If we estimate that 5 per cent are LGBT+ then we can say about 309,643 of these are LGBT+.

On homelessness, we can use statistics from the Department of Communities and Local Government. In 2016 local housing authorities declared 13,280 people aged 16-24 homeless – this would miss a lot of people, but would include people who had to leave their parental home because their parents threatened violence, or those with mental health problems. It would miss all those in full-time education. If a quarter of them were LGBT+, that would be 3,320 people, 1 per cent of all LGBT+ people. That’s plausible and would mean LGBT+ people are over-represented in this population.

However, local housing authorities deal with a lot more people through an approach called Housing Options – this is an interview with a housing officer where they work to reduce your risk of homelessness, or prevent your homeless, without going down the statutory route. The number of people given these services are collated by DCLG. This gives a (slightly) more accurate figure of who is homeless. In 2016, 200,610 people aged 16-24 in England used these services. Now, 25 per cent of that would be 50,152. If this were the case, this would mean 16 per cent of young LGBT+ people had experienced homelessness, and again it would mean an over-representation compared to a rate for the non-LGBT people of around three per cent.

As I’ve made clear, this is all very difficult to measure. We even know that the 200,610 people who had accessed Housing Options is an underestimate. Some estimates put youth homelessness (including having to sleep with a friend because you have fallen out with your parents) as high as 25 per cent of all young people. This seems high, but even for me, who lived a pretty middle class life as a teenager, it rings true as a few friends did become homeless during these turbulent years. If we use this figure we could estimate that 1,548,217 young people become homeless in a year. Then, 25 per cent of that number is 387,054 – that’s more LGBT homeless young people than there are LGBT+ young people. However, this homelessness rate would be an over-exaggeration as it is a rate for all 16-24 year-olds i.e. 25% of young people experience homelessness at some point when they are young.

Something still does not ring true with the 25 per cent figure though – there are too many questions around it. It also does not match the figures of service users I’m finding in my own research (5-10 per cent). It would be easier to say: we don’t actually know how many LGBT+ young people become homeless. Because we don’t. 

Does it matter? I would say yes, it does, and I say this as a gay man who is passionately interested in this subject and the housing outcomes for LGBT+ people. I would argue it matters for two reasons. Firstly, I think it distracts us from the actual issues around LGBT+ homelessness. The causes of LGBT+ youth homelessness are, from what we know, the same as they are for all young people – family breakdown, unemployment, poverty and mental health. In terms of family breakdown, their sexual or gender identity is very likely to have an intersection with this, or be the primary cause (they are fleeing homophobic or transphobic parents or guardians). However, the key issue to consider then, in a UK context, is what services they can access and some pretty basic issues like:

• Do services record sexual and gender identity of service users? (we’d then know how many were LGBT+!)
• Do LGBT+ people consider themselves “candidates” for housing and homelessness services, or are they self-excluding?
• Are services sufficiently understanding and tailored?
• How are shelters managed? Are they gendered? How do they respond to homophobia/transphobia?

These are all issues raised by LGBT Youth Scotland in their excellent evidence submission to the current inquiry into homelessness by the Scottish Parliament Local Government and Communities Committee. I shall also be following them up with further blog posts coming out of my current research project.

I think there is also an issues that focusing on this 25 per cent has a danger of foregrounding risks of homelessness associated with sexual and gender identity to the detriment of the widespread risks we know are there, are increasing, and we need to do something about: the massive cuts in housing benefit for young people; poor quality work, that is low paid and insecure; problems with tenure security; lack of suitable affordable housing for rent.

Finally, as a social scientist, I think we have a duty to produce rigorous evidence, particularly when it might be used to inform public policy. Yes, we do not know nearly enough about the experience of LGBT+ people at all, let alone youth homelessness. But we should present statistics such as the quarter with suitable caveats – that we simply do not know, in this case.

(1) I am, very badly, using "trans-" here to cover a multitude of sins - the sins being mine. I'm using this term to cover people who identify as transgender or non-binary, and also people who identify as queer. I know this is wrong, and I know this is a lazy shorthand, but I talk more about definitional issues later on, so please let me off. 

(2) So, the key thing here is “priority need”. Everywhere but Scotland, a local housing authority only owes you a duty of housing under legislation if you have “priority need” – basically four reasons: you are pregnant or a parent or guardian with children; you are disabled, ill or have mental health problems; you are a victim of domestic violence; you have been made homeless by fire, flood or disaster. The first one of these obviously commonly excludes LGBT+ people. Something else to add here is that you also need to a “local connection” to the area. This might indirectly discriminate against LGBT+ people who leave small towns to be with LGBT+ communities in larger cities. 

Social media and social research: My first foray

At the moment I have some time to spend on the fascinating AHRC Connected Communities project Ladders to the Cloud, helping to better understand the Facebook page From There to Here… and help the community organisations involved in it turn it into something, well, more. For such a social media addict (my Twitter follower count is now up to giddy heights of four figures) this is my first foray into social media research.

Back in July I was given the login details for the Facebook page and so, for the first time in my life, I got to see the back-end of a Facebook page (rather than a user profile). Because these are essentially the money-making end of Facebook they provide you with “Facebook Insights” which give you age, gender and location information for people who look at your page, comment on your page and the “viral reach” – that is the number of friend’s feeds your posts appear on after someone has commented on them. It’s great for a project like this; with the handy knowledge that the research suggests very few people lie about biographic details on Facebook profiles, the site is essentially doing half of the social science work for you. Which reminds me, this book, which I'm told makes a similar point about the Tesco clubcard data, is on my to-read list for the nearly ending summer.

So, what did I learn from all this data? Firstly, it’s a bit difficult to download what you want. I ended up having to do lots of screen grabs, print out the images, and the then input the data back into an Excel spreadsheet. This done I learnt something very interesting. In March this year a big shift occurred, which is continuing, which has meant the biggest group of people heavily engaging with the site (essentially commenting on the images) is women between the ages of 25 and 44. Before then, the gender split was even and quite a few younger and older people were also commenting.

The next question is, well, there might just be a lot more Facebook users of that demographic in Edinburgh and the UK. Here my next trick came in handy. If you have a Facebook page you can pay to advertise it. If you click on the buy an advert section you have a form where you can target who you want to see your advert. As you close in on your ideal target market a counter on the right hand side changes to tell you how many people that is. So, there are around 38,000,000 Facebook subscribers in the UK and about 8,000,000 of them (21 per cent) are aged 25-34. Of this 359,000 of them live in Edinburgh (about two thirds of the city’s population) and 100,700 of them are aged 25-34 (around 28 per cent).  From all this, I know that Edinburgh has an oddly young profile of Facebook users, with these 25-34 year olds being the biggest group, whereas across the UK it’s 35-44 year olds who are the biggest group.

As a result I would expect more users at this age, but we’re getting a lot more – we’re getting 10 per cent more of these age groups then we’d expect if they were representative of the population of UK Facebook users. And 60 per cent of them are women. Why? I suspect this is just being driven by the content – these are photos of Wester Hailes in the late 1980s and early to mid-1990s and the memories being discussed, such as this fantastic exchange of people working out who lived where, are of people of my age remembering growing up in the neighbourhood.

The trouble with all this, is that when I put out a (very successful) call on Twitter for info on social media use, the journals I was pointed too, like the Journal of Computer Mediated Communication, predominantly report on studies done on college attending young people. This is partly because Facebook was at first limited to this group; partly because they are early adopters to new social media; partly due to a social-psychology research interest in this period as one of emerging adulthood; and finally just due to the publication lag - most of these studies were done in 2006-7 (and thus I've learnt about a lot of long lost social media sites! Bebo anyone?). So it’s difficult to find studies that might help explain these trends and put them into some sort of broader context. Although, I imagine if I can get a paper off swiftly I can hit a zeitgeist wave of research on other users of Facebook.

Finally, one last little thing, which I need to be very careful about how I write because I might end up getting sued by Facebook. When I was first getting the advertising data on the number of people in each age group I did not check a little box saying “exact age matches only” (or similar). So, I got the numbers for each of the pre-defined age groups and then as a double check got Excel to work out the total population to compare with Facebook’s total number of UK users. Miraculously the number of Facebook users in the UK had almost doubled, and a vast number of these people turned up in the 25-44 age group that marketers most want to target.

And in actual fact, to be precise, at the moment I am stuffing my face with sea salt and balsamic vinegar Kettle Chips dipped in humous. You should try it, it’s very tasty.

History and social science

So, I've been thinking a lot about the relationship between history and social science, particularly in terms of methodology as applied epistemology and ontology. This was occasioned by reading Christopher Pollitt's wonderful book Time, Policy, Management.

It's also been occasioned by me reading about Habermas' distinction between the Lifeworld of everyday experience and language and the System of zweckrationalitat. And I've recently started reading Dorothy Smith's 2001 article 'Texts and the ontology of organizations and institutions'.

Now I have to see if I can put these thoughts in order...

One thing that really angers me in social science is badly written social science. By badly written, I mean the sort of clever post-structuralist stuff that is actually hard work to read. I was reminded of this recently when I said to a colleague that I was very impressed they'd actually managed to read Chouliaraki and Fairclough. Their reply was: "parts of it I read and parts of it I understood... or at least think I understood".

I always find this quite ironic, since a lot of this writing is supposed to be empowering and a critique of language and society. Yet it's unreadable. It links to an idea I'll return to below - literacies. As an academic and a social science I have a different literacy to most people. I read academic texts very regularly and can write in this way as well. My academic literacy has changed recently well - being a busy lecturer I now have to be a lot better at skimming and I'm getting there. But the literacy expected of people like Fairclough and Howarth is a literacy I don't have. And I don't think I wish to have.

As far as I've got in Dorothy Smith's article she's made two points I find interesting that link to this. Firstly, that institutions are recreated by text and secondly that social science uses language to abstract institutions to understand them.

On the second point I have to agree. This is what Habermas was getting at with his notion of The System - reifying things to comprehend them when they actually exist in the Lifeworld. On the first point, this is something I keep coming back to as a historian turned social scientist. Whenever social scientists emphasise the importance of language/discourse/text in creating society I just think "no shit, Sherlock". As a historian the past doesn't exist. It only exists as a few shreds of predominantly written evidence you gather together to tell a story of the past. You critically analyse the text - who wrote it and why? And it's implicit that society and culture is discourse. And this isn't a problem. The standard of evidence is that which Habermas suggests is that which exists in the Lifeworld - the power of the persuasive argument and therefore there are very few certainties in history and as Habermas suggests of communicative action in the Lifeworld: 'the typical states are in the gray areas in between'.

And now I'll return to the first point. In developing something as simple as narrative history doesn't reify to the same extent as social science. It specifically tries not too abstract too much from the texts. You're expected to reference your evidence so other historians can see where you have developed it from. History, I think, as an intellectual project is just about focusing on the contrast from the now. I was reminded of this watching the wonderfully game Lucy Worsley on If Walls Could Talk as she explained that until the advent of modern plumbing in the late nineteenth century the very idea of a "bathroom" was entirely alien. This "problematised" the bathroom and hygiene as an idea immediately. It showed it was partial, social constructed and therefore of interest. When I learnt about the history of literacy in early-modern England this did the same for me. The vast majority of people before the eighteenth century were illiterate. Yet because the written word was things like penny ballads, literacy was actually just different. A penny ballad would be bought by a village and the one literate person would sing it to other villagers. If it was especially good it would be kept and become folklore. The reformation and its emphasis on reading the bible to reflect on one's own spirituality made reading the private act it is today. Therefore we shouldn't talk about literacy, but literacies. When it comes to Chouliaraki and Fairclough I'm as illiterate as a seventeenth century farmer.

Social science, being obsessed with the now can only ever achieve this through reifying objects. If it's positivism it abstracts them into theory and statistics. If it's ANT you turn them into "things". If you're into critical discourse analysis then you use a load of post-modern wank badly translated from Foucault and Lauclau to leave your reader baffled.

This has purposefully been a bit of a rant and I know there's lot of holes in my argument and also I'm generalising about history. But it is a rant because I think there's political implications for social science here. As a planner and social scientist I want to make the world better. If what I write is unintelligible to people then how can I do this? This was something I criticised in my research - the reification of social "things" in strategic policy divorced them from the Lifeworld of community activists, therefore there could not be radical democratic engagement. If I'm doing the same, am I not as guilty as the policy-makers I criticise.

The public popularity of good history highlights this. Pretty decent history books from good historians are often in the best-seller lists. Christ, there's even a few TV channels devoted to the subject. Is there a "social-science TV"? No, there's 24 hour news mis-reporting the "now". I think social science can learn a lot from history to reflect on methodology to make it more readable and change the literacy of social science.

Does this make sense?