Lorraine Daston ed. Science in the Archives – new from University of Chicago Press

Progressive Geographies

Lorraine Daston ed. Science in the Archives: Pasts, Presents, Futures – new from University of Chicago Press.

9780226432366.jpgArchives bring to mind rooms filled with old papers and dusty artifacts. But for scientists, the detritus of the past can be a treasure trove of material vital to present and future research: fossils collected by geologists; data banks assembled by geneticists; weather diaries trawled by climate scientists; libraries visited by historians. These are the vital collections, assembled and maintained over decades, centuries, and even millennia, which define the sciences of the archives.

With Science in the Archives, Lorraine Daston and her co-authors offer the first study of the important role that these archives play in the natural and human sciences. Reaching across disciplines and centuries, contributors cover episodes in the history of astronomy, geology, genetics, philology, climatology, medicine, and more—as well as fundamental practices such as collecting, retrieval, and data…

View original post 60 more words

Why I prioritise writing books over articles, even in an era of research assessment

Good pushback by Stuart on the value of books in a REF world. Obviously this is not for everyone, nor is everyone able to make this choice. But books are undervalued (until it comes time to read one). Maybe for every x books you read, you could write one? Unfortunately, edited books are not favored by publishers. One thing to add are good open source outlets (eg ACME books), like the move toward open source articles (though there is still bias against self-published books).

Progressive Geographies

There is a widespread perception that the UK higher education system emphasises quantity over quality in terms of publications, and that there is a constant need to write and submit journal articles. Yet in a six or seven-year research cycle, academics have – until now – needed to submit just their best four pieces. Only four pieces, which for most people is a fraction of what they have actually produced. It is worth noting that there is a possibility, following the Stern review, that the number will change, possibly downwards and perhaps to an average number, which may require some people to submit less, and some more. While there is, and will continue to be, a need to get that right number of pieces, the question of the perceived quality of those pieces is much more significant. Pressure to publish more often comes because of a perception that what has…

View original post 1,384 more words

Schools switch to Peters projection

Several readers have sent me the following story:

Natacha Scott, director of history and social studies at Boston public schools, said it was “interesting to watch the students saying ‘Wow’ and ‘No, really? Look at Africa, it’s bigger’”.

“Some of their reactions were quite funny,” she added, “but it was also amazingly interesting to see them questioning what they thought they knew.

Individual schools in the US have used the Peters maps, Scott said, adding: “We believe we are the first public school district in the US to do this.”

The district has 125 schools and 57,000 students, 86% of whom are non-white, with the largest groups being Latino and black. After changing the maps, Rose said, educators plan to look at other subjects and shift away from presenting white history as the dominant perspective.

Thanks to Pat McHaffie and Sue Roberts.

“Prisons have been substituted for steel”

Interesting piece on why Trump still has support in rural areas. Sample quote:

For years now, leftists have been arguing identity politics versus economics—as if somehow the two could be neatly separated. As if racism and sexism weren’t intrinsic to how economic oppression works in America. As if the fairy tale of equal opportunity only short-changed people of color. Or women. As if class were simply a matter of income.

Nobody around here needed Bernie Sanders, or Elizabeth Warren—or Donald Trump—to tell them the system is rigged. What Trump did give them was the sense that they mattered. Not just their votes, but their culture, their sense of themselves as people who worked with their hands and played by the rules. People who felt they’d been written off by the Democratic Party—and had given up on politics. People for whom opioids, not religion, were the opiate of the masses. Trump found a way to reach these people on what felt like common ground: the ground of culture.

The library of missing datasets–now a github repo and art piece


Enter a caption

The Library of Missing Data Sets by Mimi Onuoha

Brilliant. As Mimi Onuoha explains:

What is a Missing Data Set?

“Missing data sets” are my term for the blank spots that exist in spaces that are otherwise data-saturated. My interest in them stems from the observation that within many spaces where large amounts of data are collected, there are often empty spaces where no data live. Unsurprisingly, this lack of data typically correlates with issues affecting those who are most vulnerable in that context.

The word “missing” is inherently normative, it implies both a lack and an ought: something does not exist, but it should. That which should be somewhere is not in its expected place; an established system is disrupted by distinct absence. Just because some type of data doesn’t exist doesn’t mean it’s missing, and the idea of missing data sets is inextricably tied to a more expansive climate of inevitable and routine data collection.

Here are a few missing datasets:

When the US Foreign Intelligence Court meets
All extinct languages
Number of nuclear weapons Israel has
Number of EU migrants currently in the United States
People excluded from housing due to criminal records
Undocumented immigrants for whom prosecutorial discretion has been used to justify release or general punishment

For more, see here.

Onuoha describes herself as an artist and researcher. Her previous pieces include writing for National Geographic voices, and she has taught a class on the art of digital mapping at NYU.

As she points out, because a dataset is missing, it doesn’t have to stay that way. We now have data on the number of civilians killed in interactions with the police, thanks to the Guardian The Counted project. What’s brilliant about Unuoha’s contribution, is that it’s not about missing data, but about missing datasets. Missing data is where you have a dataset but it’s incomplete. A missing dataset however is a hole in our thinking; a missing concept. So as much as anything her project is about a form of “epistemic closure” or belief systems that are unable to recognise their gaps.

In this day and age of “post-truth,” “fake news” and a certain kind of lack of purchase of the empirical, her project is very timely.

(h/t @baratunde here.)

New Census data reveals rural America

Little noticed by most people except data-hounds, yesterday the Census Bureau released another slug of data. This time it was the American Community Survey (ACS) releasing the five-year data for 2011-2015. In their blogs about this, the Bureau chose to concentrate on rural America.


The Census has a fairly long explanation of what it defines as urban and rural, here, but basically after following a process of identifying what is urban, the rural is what is left over (about 97% of the land area of the USA).


Kentucky is a fairly rural state by national standards. Using the same categorization scheme as the Census map above, here is the map for KY:


What else can the ACS data tell us? I am interested in not so much unemployment, but the degree to which places are not in the labor force (NILF). The following map shows the percent of 16 year olds and older who are not in the labor force in 2015:


This shows better I think that when 50, 60 or even 70% of adults are not working in a community, even though they may not be on the unemployed listing (because they’re not looking for work) that leaves a fairly significant psychological impact.

For example, Martin County has the state’s highest proportion of adults not in the workforce (69.9%). It is 99% white, with a total population in 2010 of 12,929. The unemployment rate is at 10.3%, and its other characteristics, according to a now infamous article in the NYT make it one of the “hardest” places to live in the US:


A similar map from the NYT a few years ago shows that Martin County gets 42.6% of its county income via government benefits:


Trump beat Clinton 88.6 – 9.2% (3,503 – 363) in Martin County, about a wide a margin as anywhere in the county. The state’s fifth Congressional District has a Cook partisan voting index (PVI) of R+25 (the most extreme is a district in Texas, R+32). In the Civil War it was Unionist, and even during the second half of the 19th century it was Democratic, but it has been reliably Republican since. This is what people are referring to when they say that poor white working class “vote against their own interests.”

Interestingly, next door in Pike County, 4,848 people voted for Bernie Sanders in the Democratic primary and 2,335 for Clinton (source). That’s 7,183 Democratic voters. However, Trump took it 80.1 – 17.4% or 19,740 – 4,277 in the election. What happened to the nearly 3,000 Democratic primary voters (40% of the total)? Did Sanders supporters stay home, switch to Trump, or did Democratic supporters per se stay home? Either way, you can’t have nearly 40% of your primary voters drop out of the national election. In fact, you’re supposed to increase your totals for the actual election if you have a good GOTV ground game!

This supports the “depressed turnout” line of thinking in explaining Clinton’s loss. However, the relative weight we place on this, along with the “left behind” explanation, the racism explanation and others, will no doubt be argued over for a while yet.

Did the opioid epidemic help Trump win?

New article in The Nation:


Enter a caption

New research shows that Trump made huge gains in counties with the highest rates of death from drugs, alcohol, and suicide.
By Zoë Carpenter

The Nation article draws on empirical work by Penn State prof Shannon Monnat, who argues that counties “heavily burdened by opioid overdoses and other ‘deaths of despair'” [eg suicide] were where Trump “overperformed” in terms of his voting outcomes compared to Mitt Romney.

There’s been a lot of debate recently about whether Clinton “lost” Democratic votes (ie suffered lowered turnout) or whether Trump found some appeal, beyond explanations deriving from a politics of fear or racism. Logically, both could be true at the same time of course, but for my money I am most persuaded by Monnat’s analysis. I see it fitting a tradition of writing about rural/rustbelt anxieties about being left behind in terms of the economic transition. In Kentucky for example, there’s plenty of angst around the coal industry (Appalachian Kentucky voted overwhelmingly for Trump…even as they massively preferred Bernie Sanders over Clinton in the primaries.)

Here’s a picture of a typical car license plate I took on the UKY campus, outside the computer science building, part of the “Friends of Coal” movement:


You could respond with the observation that this is a rear-guard defense of the 6,254 (Q3 2016) people still in mining in Kentucky, and/or are long-standing trends (and were apparent during Romney’s and Obama’s election campaigns). Indeed, the shift from industrial to post-industrial “service” economies was taught in my undergraduate curriculum in the early 1980s by my former Liverpool university lecturer Peter Daniels, among others. [Sidenote: he’s emeritus now, wow].

The counter to that would then be something like, yes, but Trump mobilized sentiment in effective ways [how?]; just like Brexit, there’s something affectively going on at this particular time. Opioid use has certainly shot upward in the last few years alone. The recent Surgeon General report claimed that more people have a substance use disorder than who have cancer (20.8 million vs. 14 million). More precisely, 47,000 died from a drug overdose (30,000 from prescription drugs, largely opioids). This has led to headlines like the “Oxy electorate” and “opioid alley” referring to Kentucky, where 1,248 people died from drug overdoes in 2015. This has knock-on effects of high rates of hepatitis C. These do seem new developments at this scale.

Some of these eastern Kentucky counties are also the most heavily dependent on government benefits (well over 45% of county income in some cases). This means a full explanation will also have to account for stagnant/declining wages since the 1980s. And for that we need an account of neoliberalism…