My new book, The Dangerous Art of Text Mining, comes out from
@CambridgeUP
in 2 weeks!
To learn why text mining (the technology behind many
#AI
platforms, including
#GPT
) is a dangerous art, follow the thread below.
Available to Preorder:
In Finland, we take clean trains with WiFi all day at reasonable fee.
People get mild ailments checked out at minimal cost thanks to national healthcare. There is free universal daycare from eight months. Plus compartments on trains for children to play.
@ebruenig
I'm thrilled to be joining Emory's Department of Quantitative Methods as a professor. With amazing colleagues like
@laurenfklein
and
@soni_sandeep
, I'll be joining one of the most robust departments in North America for pursuing the data-driven analysis of culture and politics.
Reading this with disappointment, as I pack up my own office in a History Department to move to a department of data science. History departments and history public life are shrinking when we need them most.
Information wants to be free, and bad news travels faster than good. It's my sad duty to report that History News Network no longer has sufficient funding to maintain its operations as an active project of the History Department at George Washington University.
My article -- "The Revolution in Text Mining is Here" -- is now live at the AHR.
Read about the pathbreaking work of
@roopikarisam
@jmjafrx
@richardjeanso
@LukeBlaxill
@_akpiper
& many others & how they created the tools we need for turning data into trustworthy, historical
The American Historical Review’s March 2024 issue is now available. This issue features articles and forums that rethink approaches to intellectual, migration, imperial, Indigenous, and digital histories.
Friends, I’m beside myself: The Long Land War is officially out today from
@yalepress
Amazon:
Seminary Co-op:
So many thanks to everybody who helped make this possible!!!
For those who haven’t heard... we have a date!!
The Long Land War will be coming out on April 19th from Yale University Press!
Book is ready to preorder:
Amazon:
Seminary Co-op:
At Princeton, I gave a lecture on the institutional requirements of digital history research under the title, "Launching a Digital History Lab for the Twenty-First Century." I touched on the accomplishments of European history labs and the many reasons that U.S. departments have
@Lamoille85
@ebruenig
There’s thing thing you can invest in when you have shared infrastructure, and it’s called “maintenance.” Things that get dirty can be made clean again. Success depends on having feedback and a budget to respond. There’s a word for that, too, in socialist countries! “Planning”
Hey everybody! I’m on The Dig, talking about land dispossession, climate change, and the capacity and incapacity of modern people to solve really big problems.
I had so much fun talking to
@DanielDenvir
, and I like to think that our conversation could be fun to listen to for
New
@thedigradio
! My interview w
@joguldi
on The Long Land War: global fight for agrarian reform and tenant rights from late-19th century Ireland thru mid-20th century China and India, late 20th-century UK—into a future of climate-induced mass displacement
I am so excited to report that I am now affiliated faculty at the Department of Computer Science here at Emory.
So many exciting conversations to be a part of as Emory launches its AI initiative!
🤓
My new book, The Dangerous Art of Text Mining, comes out from
@CambridgeUP
in 2 weeks!
To learn why text mining (the technology behind many
#AI
platforms, including
#GPT
) is a dangerous art, follow the thread below.
Available to Preorder:
Friends! My text mining article for the American Historical Review is out in the world. It's open-access for 30 days. "The Algorithm: Mapping Long-Term Trends and Short-Term Change at Multiple Scales of Time"
Text mining -- the technology behind GPT -- means counting words to find patterns. Text mining is DANGEROUS because bias is everywhere -- in incomplete data sets, in the racial and gender prejudice of the past. But there's an art to doing it right.
Friends, exciting news… The Long Land War will be released on May 3, but the advance copies have arrived!
Twitter challenge, round one: I will send a copy to the lucky reader whose retweet (or quote tweet) gets the most retweets this week!
🎶 Five hundred pages of copyediting on the wall... 500 pages of copyediting... take one down, pass it around, 499 pages of copyediting 🎶 -- and it's done! The Dangerous Art of Text Mining is coming out soon from
@cambridgeup
! 🍾
My new book, The Dangerous Art of Text Mining, comes out from
@CambridgeUP
in 2 weeks!
To learn why text mining (the technology behind many
#AI
platforms, including
#GPT
) is a dangerous art, follow the thread below.
Available to Preorder:
Honestly one of the hardest things about having given up a life of privilege to write the books I thought important to write is that one becomes just invisible. The hegemonic visibility of elite institutions is no joke.
I am thrilled that The Long Land War was just named a "book of the year" in the New Statesman. Profound thanks to
@zeithistoriker
who appreciates how my book "flips categories," showing how "ownership rights" were displaced by a new regime of "occupancy
Just finished revising the syllabus for a new lab-style course at my new home (
@EmoryUniversity
).
Title: "Let's Survive Climate Change."
Content: we text mine building codes, world bank reports, corporate reports, & Congress, figure out where regulation is happening & failing.
@ebruenig
Historians here tell me it's not about ethnicity or scale. It's about the women's movement in these places, a century ago, making demands on behalf of a better future for kids. Some of the rhetoric was conservative and about family.
Antoinette Burton identifies a digital lag in
#history
, calls for
#historians
to get on board with digital tools: "As a newcomer to the field of digital humanities/digital history, I am struck by how dominated it is, even now, by the concerns and interests of literary scholars."
The great
@RobGMacfarlane
on being "bad ancestors." Landscape studies, Macfarlane style, is one of the most important longue-durée approaches to the environment to watch.
New this week! _Digital_Histories_ is a must-read examination of digital techniques applied to every period in history from the medieval to the contemporary. Open-access and in print from the wonderful folks at Helsinki University Press!
This is a very important thread. The funding behind all the research in all the fields in which I trained is being dismantled.
We are de-skilling our society, just like when we need more trained interpreters of democracy, the root cause behind LLM hallucination, and the culture
The big move in the humanities funding world that people are finally talking about is Mellon abandoning humanities research in the academy. But the other big move is that the SSRC is also leaving behind support for the humanistic social sciences in favor of policy.
Hi everyone!
In July, I summed up my last decade of research into what history tells us about how to survive climate change.
This is, hands down, the most important thing I have ever done.
Please watch it and share.
Chicago! I’m giving the first ever Shapiro lecture on environmental history Thursday evening at 5. Come join us! “The Administration of Land, Air, and Water in the Longue Durée and how to study it.”
@CambridgeUP
Text mining is a dangerous art because analysts trained in quantification often lack a sense of what could go wrong when archives are biased, incomplete, and imprinted with the suppressions of the past.
Proud to have been part of the team that computationally analyzed the New York Times' biased coverage of Israel and Palestine. This is part of a broader effort by Writers Against the War On Gaza
@wawog_now
, who released today a tour-de-force exposé of NYT:
.
@adam_tooze
is preaching the grim economics of truth this morning in his newsletter.
I can't help but play out the scenario in my head: we aren't on track to make our climate deadlines; we don't make them, because we believed market incentives were all we needed and regulation
At my last trip to Stanford, virtually every historian I spoke to had some kind of digital project. Here's the MIT review talking to some smart people who use algorithms to research the past.
Last night's talk, the Inaugural Shapiro Lecture in Environmental History at the University of Chicago, is up. This talk begins with a history of the administration of infrastructure and landscape over seven hundred years before turning to the challenging question of how to write
My new book, The Long Land War, is a history of climate that you haven't read before. I'm using history to think about the grounds for binding international action to treat atmosphere, land & water as common property that has to be regulated.
@CambridgeUP
Text mining is a dangerous art because analysts trained in quantification often lack a sense of what could go wrong when archives are biased, incomplete, and imprinted with the suppressions of the past.
Exactly why history degrees are needed: ‘Ask an Economics concentrator to think critically about the assumptions underpinning a supply and demand graph and, in our experience, they’ll have no idea where to start.’ What Economics Lets You Forget
'It can be said that Jo Guldi is a traditional historian and that she is not a traditional historian at all. She is traditional because her subjects are, when she studies topics such as the history of British ideas about property rights or the history of the landscape, the land
At Chicago, which is usually ahead of the curve, history PhDs are working on land reform and others are anxious to begin chapters on text mining, having imbibed from faculty that the revolution is here and text mining will soon be changing everything in history. The
@ebruenig
Of course it's the tweet that gets RT'd by a NYT reporter where I'm half-awake on the train and missing an article in the sentence.
Not trying to be Finnish; y'all. I'm a Texan who's visiting and wondering why we can't do better.
Columbia's
@adam_tooze
is proving that international analysis politics -- not economics alone -- is crucial to understanding the climate emergency.
"The US...must use its unique geopolitical leverage not to sustain fossil fuel production, but to curb it."
We are delighted to host Jo Guldi for the inaugural Shapiro lecture on Thursday May 2 at 5pm. Jo will be speaking about how new historical methods might help us better understand the unfolding planetary emergency. Please join us if you can.
@UChicagoHistory
Readers' reports for our code textbook on digital history just arrived last week. Oh my gosh. Thank you, thank you, beloved digital historians, wherever you may be.
This is the single most thoughtful, constructive review process I have ever been through. Careful remarks on
The Dangerous Art of Text Mining: A Methodology for Digital History is in PROOFS!!! 🎂 🍾 And what a beautiful job the designers at
@cambridgeup
are doing!
"The Earth for Man: Redistributing land was once central to global development efforts -- and it should be today." The fine folks at
@BostonReview
worked on a beautiful excerpt from my latest book.
@yalepress
Sitting down at last with my old dear friend, the one who go taught me to drink whiskey and introduced me to Catan when we were both children and this astonishingly articulate argument was merely a twinkle in his ever-restless eye…
@add_hawk
@Lamoille85
@ebruenig
Perhaps you believe that poverty is the absence of grace. But you don’t believe you can be the presence of grace. Jesus would laugh out loud. What kind of grace is the one that can’t help its neighbor.
"Towards a Computer Science of History" (a thread)... When tens of millions of dollars endow expansions of CS and data science+, the hires might go to the economics of information, scientific cyberinfrastructure, urban computing, even data science for the social good.
I've been so rapt with
@Ted_Underwood
's computer-generated DALL-E art where the Lady of Shalott uses a Victorian computer, that I couldn't resist asking DALL-E to tell me about the future of digital history. A thread.
My contribution to
@ExtinctionR
: A meditation on what it would mean to use humanities and social sciences classrooms in the university as a place to audit pollutors, to research the history of exploitation, & to plan a more livable future.
We have an 18-page document proposing how AI, stats, and text mining for historical analysis can make the debates of democratic institutions more transparent to journalists and citizens. Just turned down for funding… again. Anybody know a funder who likes the idea of democracy?
I just shared the best panel of my life talking space, tech, data, farms, and climate change with
@xrw
@shannonmattern
my blissed out smile in this picture says it all
This is your regular reminder that even while History lines disappear because History is dismissed by deans as a non research field, the History profession in the U.S. has chosen to remain technologically and methodologically underdeveloped.
Leading U.S. institutions have all
We are delighted to host Jo Guldi for the inaugural Shapiro lecture on Thursday May 2 at 5pm. Jo will be speaking about how new historical methods might help us better understand the unfolding planetary emergency. Please join us if you can.
@UChicagoHistory
My lab has a beta version of a web-based app called "Democracy Viewer," which allows citizens to text mine the debates of Congress & Parliament.
You can ask questions like 'how have speeches about the environment changed since in the 1970s.'
Data science needs to be smarter. GPT-tech can't analyze change over time. It can't really eliminate bias either -- because the kind of knowledge we need lies in a different part of the university. Historical thinking can help.
Just out! My essay, "The Birth of Rent Control" -- early fruits of a 10-yr project on digital history + archives together over the longue durée as a tool for examining economic inequality & racism. . See also
@ebruenig
Whatever the reasons, the Swedish and Finnish women's movements promoted a version of "family values" or a "culture of life" that included *actual children. These are talking points worth adopting.
Here is the most thorough reckoning w property law, climate, & history I can give. "Can Humanity Change its Future? On Climate Change, Global Politics, and the Long History of Land Rights" Recording from my talk
@turinginst
/
@britishlibrary
.
So, the field of Law is having a debate about how computers can help us to specify what words mean. The debate boils down to whether we use statistical "prediction" to channel popular opinion, or consult corpus linguistics, i.e. counting words in context.
This is an important
From time to time, a mighty colleague in their glory issues a dismissive review that simply gives us anxiety without casting light. I've been struggling with one such review for nearly ten years. The reviewer's unnecessarily dismissive tone filled me with such dread and self
I'm doing a lot of preaching right now to colleagues at History departments about what's happening w/ text mining research in Europe (which is booming) vs the US (which has invested in data science but excluded humanists). If you're a History chair, you could invite me to talk.
Today, as the glorious monarch butterflies are winging their way across the blue skies of Texas, is a great day for Digital History. A great day. Three big points in a short thread. You should pay attention to what just happened.
#twitterstorians
Love getting mail from other historians about how important my take on land reform has been for them — especially for scholars of India. It’s a huge topic, and I can hardly claim to have exhausted it, but it was very important to me to set up the Indian struggle for occupancy
I’m honored to be on this panel! If you’re headed to
#aha24
, make sure to stay through Sunday and hear me talk about the Long Land War — alongside these brilliant historians who are asking other questions about the c20 international governance and its failures.
The
@ToynbeePrize
is proud to host a panel at
#AH24
on January 7th at 9am, titled "What Is Global History Now?" and chaired by
@adenknaap
. This panel engages the potential of global history at its intersections with the new history of Capitalism and the new international history.
1) Yesterday was the first time a Macarthur or any equivalent prize for intellectual achievement was granted to a digital humanist. Congratulations,
@gabrielleforman
, whose
@CCP_org
should be on your syllabus if you're teaching DH, American History, or community organizing.
Once upon a time, we wrote about 'enclosure' as something specific to eighteenth-century Britain. We've learned that enclosure is a process that took place in many other places, and reappears wherever democracy is at stake. Important work here.
GPT gets historical information wrong all the time.
The "hallucinations" -- also known as "factual errors" -- mean trouble for anyone whose industry depends on a factual relationship with the actual truth.
.
The universities of Oxford and Cambridge are ancient landlords and experiments in collective housing. We need long-term food and housing strategies for the climate change era. ***
A thread.
Curious about what's been happening in text mining,
#NLP
, and historical research into texts from the past?
Here's an overview I published a few months ago:
Just sent
@CUPAcademic
500 pp on text mining & why
#datascience
needs the historical method + 53 ill. (mostly data-driven) + 7 tables + 80 pp of footnotes.
Champagne all round!🍾Esp you,
#dh
&
#history
friends; this is about how smart you are. Archival/micro-history included.
New books network interviewed me about my new book on
#nlp
, what can go wrong, and how historical thinking can make
#llm
and
#datascience
smarter. Jo Guldi, "The Dangerous Art of Text Mining: A Methodology for Digital History" (Cambridge UP, 2023)
Ok, there’s humble bragging, 🙄, and then there’s the day when the dentist’s secretary tells you unprovoked that your last book is soo vital and soo important and sooo well written 😳😳😳
A few weeks ago, I issued a warning about the dangers that
#gpt3
poses to democracy today through an attack of fact-based historical knowledge. Watch "Pseudo-History and Digital History: The Dangerous Art of Text Mining" Recording from my talk
@turinginst
“The first annual lecture, "The Administration of Land, Water, and Atmosphere in the Longue Duree -- and How to Study it," will be delivered by Jo Guldi on Thursday, May 2nd at 5pm in the John Hope Franklin Room.”