Popular in Economics

Central Monetary Authorities

Macroeconomic modelling of CBDC: a critical review

European Central Bank Ulrich Bindseil

Central Monetary Authorities

Financing Private Credit

Federal Reserve Bank of New York Nina Boyarchenko

Economics

Failing Banks

Massachusetts Institute of Technology Emil Verner

Economic History

Golden Fetters

UC Berkeley Barry Eichengreen

Economic Inequalities

Visions of Inequality

City University of New York Branko Milanovic

Economics

School Finance Equalization Increases Intergenerational Mobility

Yale University Barbara Biasi

Economics

Olivier Blanchard on U.S. Pandemic-Era Inflation

Peterson Institute for International Economics Olivier Blanchard

Morts Pour la France: A database of French fatalities of the Great War

View more insights in Social Sciences - Economics - Economic History

View more figures from UNSW Sydney

View the page of Pauline Grosjean

Read the Study

Next in Economic History

The Price of Housing in the United States, 1890-2006

Rowena Gray | University of California, Merced

Planting the Seeds of Research

Louis A. Ferleger | Boston University

Pauline Grosjean describes the construction and content of the Morts pour la France database.

Image courtesy of interviewee. September 18, 2024

Log-in or Sign-up to Faculti
Currently viewing this subject insight as a guest. You have insight(s) remaining for this month. Login to view 8000+ figures on the platform.

Generating Workspace...

Export as PDF

This video discusses the process of analyzing and understanding the fatalities of French soldiers during World War I through a comprehensive database. The author reflects on how previous historical studies lacked a unified perspective on casualty data, often focusing on localized events without comprehensive national overviews. A pivotal discovery was that the deadliest day for French soldiers during the war was September 25, 1915, during a failed offensive, rather than in the early battles typically emphasized in history classes.

This video outlines the efforts to create a thorough database that included a significant amount of soldier information, such as reasons for death and individual characteristics. This video involved improving the accuracy of historical records by cleaning up duplicates and integrating various data sources. The researchers explored factors affecting soldiers' survival on the frontlines, noting that officers had a significant advantage in terms of longevity compared to enlisted soldiers.

Additionally, the research delves into cultural indicators reflected in soldiers' first names as a means to gauge socio-religious characteristics across France. This video suggests that names correlate with Catholicism and regional identity, taking into account the historical context in which these names were given.

The author also discusses how understanding the military structure, including changes in officer demographics during the war, could be valuable for studying organizational dynamics and the transmission of values in military cohorts. A notable point made is the observation that soldiers' experiences did not necessarily lead to a cohesive national identity but instead highlighted internal diversities influenced by the political leanings of their commanding officers.

Overall, the text emphasizes the potential for this dataset to enhance historical understanding and answer questions about culture, military organization, and social movements in post-war France, making it a significant contribution to historical and social research.

Language:

Export as PDF

Export as PDF
Here are five key points from the video:

1. **Lack of Comprehensive Historical Data**: Historically, there has been limited understanding of overall casualty data in World War I, particularly for France. The speaker points out that while localized studies exist, there was no comprehensive overview of casualties, leading to misconceptions, such as the belief that the deadliest day was during the early battles rather than later in the war.

2. **Findings on Fatalities**: Analysis of daily fatality data revealed that the deadliest day for France during World War I was actually September 25, 1915, during a failed offensive at Champagne, rather than during the earlier, more glorified Battle of the Frontiers. This finding raised questions about why certain historical narratives and dates are emphasized over others.

3. **Cultural Insights through Naming Patterns**: The video discusses how the analysis of soldiers' names in relation to their places of birth can yield interesting insights into cultural and religious demographics in France, emphasizing how names can indicate Catholic influence or reflect opposition to historical events like the revolution.

4. **The Role of Soldier Characteristics**: The video mentions a study of soldier characteristics and experiences, highlighting disparities in service duration based on rank and the demographic composition of regiments. It notes that officers typically had longer service durations than enlisted soldiers and discusses the challenges of explaining variations in casualty rates amidst troop rotations.

5. **Future Research Potential**: There is enthusiasm for further research opportunities using the data compiled, with suggestions about exploring organizational economics, culture, and the impact of military leadership on soldiers’ experiences. The speaker notes that the database is rich and can be used to analyze factors like the effects of military leadership on soldier outcomes and how these experiences influenced broader societal trends post-war.

Export as PDF
Here are some key quotes and ideas from the video you've , organized by theme:

### Historical Understanding and Data Analysis
1. **Lack of Comprehensive Overview**:
- "But there was no comprehensive, um, overview of what had... what had actually happened."

2. **Revelation of Fatalities**:
- "I still remember that moment when I looked at the data and discovered this... we historians and the public at large didn't even know what was the day with the highest number of fatalities in the Great War for France."

3. **Misconception about Fatalities**:
- "It wasn't by quite a large margin. The deadliest day for France was September 25th, 1915."

### Battle Convideo and Research Findings
4. **Significant Dates**:
- "You have to wonder, you know, why does history select specific dates over others, right?"

5. **Insights on Specific Battles**:
- "The second day of the Battle of Champagne is a bit of a shameful day for France."

### Database and Methodology
6. **Citizen Science Initiative**:
- "There was a great initiative of citizen scientists called... people from all over France... went in and read the manuscript card, the handwritten cards and indexed the database."

7. **Data Cleaning Challenges**:
- "There was about 200,000 duplicates... we matched I think 99.9% of the soldiers to their municipality of birth."

### Sociocultural Observations
8. **Cultural Insights Through First Names**:
- "We can talk about first names and how this database... can actually be really interesting to have an idea of... what were the most Catholic areas of France."

9. **Correlation Between Names and Values**:
- "If a man is called Marie, it's a strong indication that you probably come from a Catholic family."

### Military Structure and Societal Impact
10. **Demographics of Service**:
- "Every man had to do it... pretty much every French man between the ages of 20 and 48."

11. **Impacts of Leadership on Soldiers**:
- “What happened is that after... the first few months of the war, a lot of officers are either fired or dead... a lot of them are teachers... leading to a big change in the leadership of the army.”

12. **Transmission of Values from Officers to Communities**:
- "So, it's really specifically the guys who started under him who adopt his values... and convince other people around them to follow him into collaboration."

### Research Directions and Conclusions
13. **Future Research Potential**:
- "There is a lot of interesting question of organizational economics or management that could be asked with this database."

14. **Reevaluation of Historical Narratives**:
- "So this paper is kind of explaining why... some went left."

These quotes capture pivotal points regarding the study of historical events, data collection methodologies, societal impacts during the Great War, and the suggested avenues for future research. They reflect both an analytical perspective and a deeper inquiry into the societal dynamics of early 20th-century France.

X-TIMESTAMP-MAP:LOCAL=00:00:00.000,MPEGTS=900000

00:00.105 --> 00:02.605
We really had very little idea about

00:02.755 --> 00:04.325
what actually went on, right?

00:04.325 --> 00:07.125
So you had a lot of studies by historians,

00:07.125 --> 00:09.245
but this is mostly localized studies.

00:09.305 --> 00:11.445
So, you know, some people had studied really well

00:12.155 --> 00:14.645
what happened to the people who died

00:14.645 --> 00:17.845
during the Great War in the Vien department, right?

00:17.905 --> 00:19.325
And you knew everything about this

00:19.585 --> 00:21.885
or in that other department, or in that village.

00:22.385 --> 00:26.325
But there was no comprehensive, overview of what had,

00:26.435 --> 00:27.685
what had actually happened.

00:28.225 --> 00:30.535
And I think this is the first, you know,

00:30.535 --> 00:33.335
and I still remember that moment when I looked at the data

00:33.395 --> 00:37.455
and discovered this, is that we historians

00:37.835 --> 00:41.815
and the public at large didn't even know what was the day

00:42.445 --> 00:44.615
with the highest number of fatalities

00:45.195 --> 00:46.575
in the great war for France.

00:47.275 --> 00:48.455
So, you know, we went to school

00:48.995 --> 00:51.975
and we were told that the day with the highest level

00:51.975 --> 00:54.815
of fatalities was during the Battle of the Frontier.

00:54.875 --> 00:57.015
So August, 1914, right?

00:57.035 --> 00:59.775
So this was a particular,

01:00.525 --> 01:02.535
high fatality moment of the war.

01:02.535 --> 01:04.375
And that's, that's entirely true, right?

01:04.435 --> 01:06.655
So you see that, you know, when you look at the data,

01:06.795 --> 01:10.215
you see that consistently this, the, the highest level

01:10.215 --> 01:11.255
of fatalities day

01:11.255 --> 01:12.775
after day is really during that period

01:13.385 --> 01:16.495
where the German army advance is very fast towards Paris,

01:16.595 --> 01:18.895
and then eventually our turnaround at the Battle of the Mar.

01:19.715 --> 01:22.535
Um, however, when we looked at the data,

01:23.075 --> 01:25.375
and I'll, I'll caveat this by saying

01:25.405 --> 01:27.335
that the data is specific

01:27.355 --> 01:29.135
to metropolitan French shoulders, right?

01:29.155 --> 01:31.855
So we don't have soldiers from the colony.

01:31.855 --> 01:33.095
And we'll come back to this because I think

01:33.825 --> 01:35.455
there is some really interesting things

01:35.455 --> 01:37.815
that we've discovered, even with that contrast

01:37.835 --> 01:39.015
for particular battles.

01:39.715 --> 01:40.775
But what we discovered,

01:40.775 --> 01:44.095
and I remember just plotting the time series of daily death,

01:44.955 --> 01:49.615
seeing that battle in, you know, in August, 1914,

01:49.615 --> 01:51.775
that was supposed to be the deadliest battle for France.

01:52.695 --> 01:56.695
And just discovering that by quite a large margin, it wasn't

01:57.945 --> 01:59.165
by quite a large margin.

01:59.305 --> 02:02.205
The deadliest battle for the day, deadliest day

02:02.265 --> 02:05.125
for France was September 25th, 1915.

02:06.195 --> 02:08.335
And what's actually interesting about that day is

02:08.335 --> 02:10.975
that is the first day of the second offensive of champagne,

02:11.625 --> 02:14.375
which ultimately was a failed offensive for France.

02:14.515 --> 02:15.695
And you have to wonder, you know,

02:15.755 --> 02:18.815
why does history select specific dates over others, right?

02:18.815 --> 02:20.295
So the Battle of the Frontiers

02:20.755 --> 02:23.135
and that specific dates close the Battle of the Mar,

02:23.135 --> 02:25.255
which is a very heroic moment for France.

02:25.915 --> 02:28.375
And the second day of the Battle of Champagne is a bit

02:28.375 --> 02:30.295
of a shameful day for France, especially

02:30.295 --> 02:32.935
because that offensive was actually led by peta who then,

02:33.355 --> 02:35.895
you know, had a particular role in, in history.

02:36.115 --> 02:38.535
We didn't have a good overview of fatalities per day.

02:38.755 --> 02:40.495
And then we can reconstruct by battle,

02:40.555 --> 02:41.695
and there's a lot of things we can do.

02:41.695 --> 02:42.935
So the first thing is to document

02:42.935 --> 02:44.095
this in a comprehensive way.

02:44.635 --> 02:47.015
And the second thing is that I think we can address a lot

02:47.035 --> 02:50.415
of questions, in, in, in research.

02:50.995 --> 02:52.855
And also maybe we get to this as well.

02:52.995 --> 02:57.695
Um, it enables us to, to peek into,

02:58.155 --> 03:00.525
the, some of the cultural characteristics

03:00.525 --> 03:01.565
of the population, right?

03:01.565 --> 03:03.285
So we can talk maybe about first names

03:03.305 --> 03:04.405
and how this database

03:04.405 --> 03:06.605
of first name can actually be really interesting to

03:07.115 --> 03:10.205
have an idea of, you know, what were the most Catholic areas

03:10.345 --> 03:13.725
of France, or what was their areas with the, you know, well,

03:13.725 --> 03:15.885
well with the, you know, highest proportion of

03:16.605 --> 03:19.165
Catholic names or highest proportions of rural names,

03:19.225 --> 03:22.045
you know, people who are named after Kings.

03:22.185 --> 03:23.565
We were lucky that citizen

03:23.585 --> 03:25.725
and scientists had done an excellent job for us.

03:25.725 --> 03:29.285
So basically the hist, the, the story of our database is

03:29.285 --> 03:31.045
that it's a database that had been collected

03:31.045 --> 03:32.285
by the French authorities right

03:32.285 --> 03:35.245
after the Great War, kept in South Archive,

03:35.825 --> 03:38.845
and then these archives themselves,

03:39.065 --> 03:40.845
put the database online.

03:41.385 --> 03:43.005
And then there was a great initiative of,

03:43.065 --> 03:47.165
of citizen scientists called, so like one day, one soldier

03:47.825 --> 03:49.405
and people from all over France

03:49.425 --> 03:51.605
and, you know, maybe in other parts of the world, went in

03:51.605 --> 03:54.325
and read the manuscript card, the handwritten cards

03:55.025 --> 03:56.445
and index the database.

03:57.425 --> 04:00.725
Um, and so, so we basically took this

04:01.345 --> 04:02.445
and cleaned it up.

04:02.465 --> 04:04.325
So there was a lot of mistakes, as you can imagine,

04:04.325 --> 04:06.165
because sometimes it's really hard to read.

04:06.985 --> 04:08.565
Um, sometimes, you know,

04:08.795 --> 04:11.165
several people had entered the same person.

04:11.345 --> 04:12.685
So there was a lot of duplicates.

04:12.695 --> 04:14.965
There was about 200,000 duplicates.

04:15.665 --> 04:18.165
Um, and I think the biggest job, and I know

04:18.165 --> 04:19.765
because I did quite a lot of it myself,

04:20.075 --> 04:21.845
because, you know, we hired a lot

04:21.845 --> 04:23.085
of recession system to do it.

04:23.105 --> 04:26.885
But really what was difficult is, so, so Victor,

04:27.025 --> 04:29.565
my co-author had basically done all the work

04:29.565 --> 04:31.685
before getting all the database,

04:31.785 --> 04:35.125
and cleaning it up at the, at the level of 87 departments.

04:35.665 --> 04:38.885
And then what we did together is we cleaned more fields.

04:39.105 --> 04:41.925
So the reasons for death, the rank of the soldier,

04:42.545 --> 04:43.805
the circumstances of death,

04:43.865 --> 04:45.965
but importantly the municipality of birth.

04:46.825 --> 04:48.845
And as you know, France has an insane number

04:48.845 --> 04:50.005
of municipalities, right?

04:50.005 --> 04:52.805
So I think the whole EU has like 80,000 municipalities.

04:53.045 --> 04:55.685
I mean, you know, more or less take, take away, 10

04:55.685 --> 04:58.605
or 20,000, 35,000 of those are in France.

04:59.315 --> 05:02.455
So it was an insane, job to just, and,

05:02.475 --> 05:04.775
and you, you can imagine the amount of spelling mistakes.

05:05.755 --> 05:08.135
So we had to clean that all up so that we can merge it

05:08.135 --> 05:10.055
to other database that would give us other

05:10.055 --> 05:11.575
information about these municipalities.

05:12.475 --> 05:15.055
We could have had a process that was a bit more efficient.

05:15.075 --> 05:16.335
But we started this

05:16.555 --> 05:19.615
before AI was what, I mean, you know,

05:19.615 --> 05:20.815
we're really talking ai,

05:20.815 --> 05:22.655
like it's really the last year, right?

05:22.755 --> 05:26.575
So Chad g PT is amazing now at reading handwritten

05:26.935 --> 05:29.765
documents, but that wasn't the case when we did this thing.

05:29.785 --> 05:32.205
So it's really development in the last year or two years.

05:32.665 --> 05:33.925
And so we did it just before.

05:34.505 --> 05:38.045
Um, so, you know, it was a process of just looking at

05:38.435 --> 05:41.245
what matched and then go over the ones that didn't match

05:41.265 --> 05:43.005
and manually, clean them.

05:43.805 --> 05:46.485
I mean, you know, you could automatize part of it, right?

05:46.545 --> 05:49.485
You could, you know, force it too much to some database.

05:49.485 --> 05:52.645
But yeah, we had, very good RAs from, the University

05:52.645 --> 05:54.885
of Toulu and a few, yeah,

05:54.945 --> 05:58.635
and I, I dunno, it, it is not a very, you know,

05:58.705 --> 05:59.755
it's time consuming,

06:00.055 --> 06:02.395
but it, yeah, it's, it's, it's feasible.

06:02.395 --> 06:05.835
And, and we matched I think 99.9% of the soldiers

06:05.975 --> 06:07.155
to their municipality part.

06:07.185 --> 06:09.675
There's a lot of things that we could do at scale

06:09.785 --> 06:11.195
that hadn't been done before.

06:11.575 --> 06:15.035
So in particular, what we do in, in the paper is

06:15.035 --> 06:17.195
that we do a duration analysis, right?

06:17.215 --> 06:19.795
So trying to see what were the characteristics of soldiers

06:19.985 --> 06:22.875
that made them last more, you know, longer

06:22.895 --> 06:24.795
or shorter duration on the front.

06:25.375 --> 06:27.235
And so here you find clearly, you know,

06:27.235 --> 06:29.275
that officers had a big advantage, you know,

06:29.275 --> 06:31.995
so on average an officer would, leave for

06:32.755 --> 06:35.715
134 days more than an enlisted soldier.

06:36.495 --> 06:39.355
Um, and then you also see the, the big hit

06:39.355 --> 06:40.675
that the line infantry took.

06:41.015 --> 06:43.115
But, and you can even see a discontinuity.

06:43.135 --> 06:46.075
So the, the, the men, so the line infantry,

06:46.575 --> 06:49.955
are the active army was the youngest guys.

06:50.155 --> 06:52.795
I think it's important to say, because, you know, modern

06:53.355 --> 06:54.795
audiences cannot understand,

06:54.795 --> 06:56.595
even though it was a hundred years ago,

06:56.755 --> 06:58.395
I think it's such a different universe, right?

06:58.455 --> 07:01.955
So prince, in 1914, everybody had

07:01.955 --> 07:03.395
to do three years compulsory service.

07:03.605 --> 07:05.275
Every man had to do it.

07:05.775 --> 07:07.915
And they would do their three years compulsory service.

07:08.615 --> 07:10.955
Uh, they would start the year, they would turn 20,

07:12.315 --> 07:14.575
and then they were in the army for 28 years.

07:16.165 --> 07:18.545
So you would do three years of military service,

07:18.715 --> 07:20.705
which would mean you would be in the active army

07:20.805 --> 07:23.545
for three years, then you would be in the reserves

07:23.565 --> 07:25.585
of the active army for 11 years.

07:25.585 --> 07:27.945
Then you would go to the territorial army for seven years.

07:27.945 --> 07:29.945
And then you had seven more years in the reserve

07:30.145 --> 07:31.225
of the territorial army.

07:31.965 --> 07:35.705
And all these men were mobilized in August, 1914, right?

07:35.705 --> 07:37.865
So it's about, it's more than 8 million men,

07:38.085 --> 07:40.705
and it's pretty much every French man between the ages

07:40.705 --> 07:42.025
of 20 and 48.

07:42.865 --> 07:44.125
You know, I don't know how old you are,

07:44.385 --> 07:48.485
but you know, we have this vision that war is a thing

07:48.485 --> 07:50.205
for young men, but it's not the case, right?

07:50.205 --> 07:51.325
So you had, and

07:51.345 --> 07:54.485
and you have very few people in France who actually escape,

07:55.065 --> 07:56.325
the drafts, right?

07:56.325 --> 07:59.045
So we're not in a situation like, you know,

07:59.065 --> 08:01.645
in other countries where a lot

08:01.645 --> 08:03.125
of people didn't do their military service

08:03.185 --> 08:04.285
and didn't go to service war.

08:04.285 --> 08:06.645
And there's a very simple reason to this, is that

08:07.545 --> 08:10.525
France has a big demographic deficit compared to Germany

08:10.525 --> 08:13.685
because France had the demographic transition much earlier

08:13.755 --> 08:14.885
than other countries in Europe.

08:15.345 --> 08:16.845
So we don't have enough men, right?

08:16.845 --> 08:20.245
So we're, we're fighting a war against an opponent

08:20.315 --> 08:23.045
that has a lot more men, an innocent infantry war, right?

08:23.045 --> 08:24.925
So it's very hard to escape,

08:25.185 --> 08:27.165
to escape military service at the time.

08:27.585 --> 08:29.965
You know, we talk a little bit more about in the paper

08:30.115 --> 08:32.685
that some men were withdrawn from the front

08:32.705 --> 08:33.925
to work in factories.

08:34.665 --> 08:38.245
Uh, but we're only talking about, you know, 500, five,

08:38.545 --> 08:40.725
you know, 550,000 men overall.

08:40.725 --> 08:44.645
So it's not a huge amount, you know, out of 8 million,

08:44.645 --> 08:46.685
people, we did this little fun exercise

08:46.705 --> 08:47.765
and we want to write it up

08:47.765 --> 08:51.365
because there is a, a very active literature on culture

08:51.505 --> 08:54.725
and economics, and people have issues measuring

08:54.725 --> 08:56.085
Culture in the past, right?

08:56.105 --> 08:58.245
So now we have a whole bunch of measures of culture.

08:58.385 --> 09:02.605
We have surveys, we have what people post on Twitter,

09:02.605 --> 09:05.765
while we used to be a able to access that, now that it's X,

09:05.825 --> 09:08.285
we can't access, that data so easily.

09:08.345 --> 09:09.605
But, you know, there is reddi posts,

09:09.605 --> 09:11.045
like people can scrape social media

09:11.145 --> 09:13.445
and have a sense of what people are talking about and,

09:13.565 --> 09:15.405
and what's going on in the past.

09:15.505 --> 09:18.045
We don't have very good measures of culture, right?

09:18.105 --> 09:20.685
We can have, you know, what's written of tombstones,

09:20.705 --> 09:22.445
you know, how many churches there are,

09:22.665 --> 09:24.365
but we also have people's first names, right?

09:24.365 --> 09:26.485
Because people, when you name your child,

09:26.545 --> 09:28.005
you convey some value, right?

09:28.105 --> 09:29.885
So there is a whole literature,

09:29.885 --> 09:34.125
there is a literature showing that whether a kid has,

09:34.905 --> 09:37.885
a Christian first name is actually quite well correlated

09:37.995 --> 09:39.525
with even occupational church.

09:39.525 --> 09:41.605
Like, these people are more likely to become priests.

09:41.705 --> 09:44.205
And, you know, so, so

09:44.305 --> 09:45.645
that's the first thing that we can measure.

09:45.705 --> 09:47.285
We can measure. And you know,

09:47.285 --> 09:49.245
it is particularly interesting in the case of France,

09:49.445 --> 09:51.725
'cause the census doesn't ask about religion,

09:51.725 --> 09:53.205
because it is kind of forbidden.

09:53.205 --> 09:54.605
So since the revolution, the census,

09:54.795 --> 09:55.885
there's a few exceptions,

09:56.105 --> 09:58.285
but in general, we don't ask about the,

09:58.285 --> 09:59.965
the franchises doesn't ask about religion,

10:00.385 --> 10:01.605
but then we have this first name,

10:01.625 --> 10:03.325
and so we can, we get a measure, right?

10:03.345 --> 10:05.485
So if, if a,

10:05.545 --> 10:07.045
and actually, so the most common

10:07.615 --> 10:09.725
first name in France at the time is Jean.

10:10.545 --> 10:13.405
And actually a lot of men have Marie as a,

10:13.545 --> 10:14.605
as a first name, right?

10:14.705 --> 10:16.485
As a second name or third name, right?

10:16.485 --> 10:19.005
So this could be a strong indicator of how

10:19.805 --> 10:21.205
Catholic your parents are, right?

10:21.205 --> 10:24.925
So if you're called Marie as a man, it's a strong indication

10:24.925 --> 10:26.765
that you probably come from a Catholic family.

10:27.395 --> 10:29.645
Then you also have people whose first names are

10:29.645 --> 10:30.885
after former kings, right?

10:30.905 --> 10:34.245
So you have Louis Francois,

10:34.625 --> 10:37.925
and actually it's, it's pretty, it's actually interesting.

10:37.945 --> 10:41.845
So if you plot at the municipal level, the proportion

10:41.845 --> 10:42.925
of people who have a name

10:42.925 --> 10:47.645
after a king, you see concentration in

10:48.265 --> 10:52.805
VDI and north of Belfor, which were very stronghold

10:52.805 --> 10:56.045
of anti-revolutionary revolts, right?

10:56.145 --> 11:00.765
So, so even in 1914, you see some reminiscence of opposition

11:00.765 --> 11:02.445
to the revolution, and people, you know,

11:02.765 --> 11:05.605
manifest their opposition by calling their kid Louis,

11:06.025 --> 11:08.525
as opposed to completely Republican names, right?

11:09.305 --> 11:12.285
Um, and then you also have more or less common names.

11:13.305 --> 11:16.245
So there's also a literature that that correlates,

11:16.905 --> 11:19.325
the fraction of rare names, rare first names

11:19.355 --> 11:20.485
with individualism.

11:20.705 --> 11:24.125
So people who are more individualistic tend to

11:24.795 --> 11:27.285
have a more rare first name for their child.

11:27.345 --> 11:28.925
So you can also have a measure of this.

11:28.945 --> 11:31.205
And then the map of France is pretty interesting, right?

11:31.205 --> 11:35.005
So from the point of view of religion, we have a map

11:35.275 --> 11:39.405
that corresponds pretty well to a map that an,

11:40.245 --> 11:43.365
a re a priest had done in the forties to, he went

11:43.365 --> 11:45.285
around every church in France to measure attendance.

11:45.825 --> 11:48.005
And you really have this empty middle of France.

11:48.985 --> 11:51.925
So, you know, you have very religion regions in Britain

11:52.065 --> 11:54.925
and the south of France, and then this, this middle region

11:54.945 --> 11:56.045
of France, like the diagonal,

11:56.045 --> 11:58.565
like this is actually much less strongly

11:58.885 --> 11:59.925
religious than the rest of France.

11:59.985 --> 12:02.645
So, you know, at the moment it's just measurement things.

12:02.645 --> 12:04.925
And we have this data and we haven't really done much

12:04.925 --> 12:07.605
with it, but, it's, it could be useful

12:07.625 --> 12:08.725
for, for future research.

12:08.825 --> 12:10.885
And yeah, so another thing that we did as well is

12:11.165 --> 12:13.845
that we correlated it, we correlated the fraction of

12:14.685 --> 12:17.405
Catholic, of, Christian first names with the share

12:17.505 --> 12:20.205
of refractory clergy during the revolution.

12:20.205 --> 12:21.565
And the correlation is quite strong.

12:21.765 --> 12:25.285
I have another paper with another group of co-authors,

12:25.825 --> 12:28.325
and we look at the relationship

12:28.325 --> 12:30.645
between serving at the Battle

12:30.665 --> 12:34.765
of Verda under the general ship of, Pitton

12:35.025 --> 12:36.405
during the First World War,

12:36.665 --> 12:40.885
and the propensity of people in those municipalities to,

12:41.025 --> 12:43.965
become collaborators in World War ii, right?

12:43.985 --> 12:47.605
So maybe the, the, you know, it's a,

12:48.025 --> 12:50.485
so Pitton was a general in the First World War,

12:50.485 --> 12:52.045
he becomes a marshal at the end of the war

12:52.045 --> 12:54.005
because he comes out as this hero of world.

12:54.245 --> 12:56.485
I, and when, France is invaded

12:56.485 --> 12:59.925
by Germany in World War ii, they call him,

13:00.305 --> 13:04.205
to save France, you know, again, except what he does is

13:04.205 --> 13:05.965
that he puts in place an authoritarian regime

13:05.965 --> 13:09.405
that is collaborating with Germany until 1944.

13:09.985 --> 13:13.805
Um, and so we can establish this link, between, and,

13:13.905 --> 13:18.445
and so one argument why Pean was so popular in World War II

13:18.585 --> 13:20.165
and the whole of the interwar is that

13:20.165 --> 13:23.325
because he was such a hero of the first of the Great War,

13:24.105 --> 13:28.405
and so we actually show in this paper that the,

13:28.515 --> 13:30.765
there's this transmission of values that the people

13:30.765 --> 13:32.685
who serve under him adopted his values.

13:32.785 --> 13:34.165
And that's true of every veteran,

13:34.165 --> 13:35.685
because we can trace

13:36.035 --> 13:38.605
what municipality served under him at different times.

13:39.305 --> 13:41.605
And you can see that any municipality that belonged

13:41.605 --> 13:43.085
to a regimen that served under him

13:43.345 --> 13:47.255
during the First World War, is so the,

13:47.255 --> 13:49.295
the veterans are more likely to become collaborators,

13:49.355 --> 13:50.655
the people who served under him,

13:51.115 --> 13:52.775
but specifically in the municipality

13:52.805 --> 13:56.215
that served under him at Verda, you see that these people,

13:56.215 --> 13:57.895
because they're heroes themselves,

13:57.895 --> 13:59.575
they transmit these values to the rest

13:59.575 --> 14:01.055
of the people in their municipality.

14:01.115 --> 14:03.775
So it's not only they're just the veterans, it's really

14:03.775 --> 14:05.295
because themselves, they're heroes.

14:05.645 --> 14:07.175
Everybody follows them, right?

14:07.195 --> 14:08.215
So the, the old,

14:08.435 --> 14:12.095
the grandpa who's this amazing veteran Froman and every,

14:12.155 --> 14:13.775
and you know, who's still been telling his son,

14:13.875 --> 14:15.215
the only reason you're alive is

14:15.215 --> 14:18.415
because of, bean, like they have the fact of going

14:18.415 --> 14:23.215
to is such a heroic, trait for the, for the,

14:23.315 --> 14:25.175
for the veterans, is that it's,

14:25.175 --> 14:26.335
it's both a combination, right?

14:26.335 --> 14:28.095
Because we've basically exploit

14:28.195 --> 14:30.735
who was at Verda under Marshall peta,

14:31.115 --> 14:32.415
and then who served at Verda

14:32.415 --> 14:33.815
after he was removed from command.

14:35.145 --> 14:36.165
So the guy who served

14:36.165 --> 14:37.605
after he was removed from command,

14:37.605 --> 14:38.845
they don't adopt these values, right?

14:38.865 --> 14:40.645
So their heroes, they can transmit values

14:40.765 --> 14:41.845
to their local communities,

14:41.845 --> 14:43.005
but they don't have these values.

14:43.065 --> 14:45.605
So it's really specifically the guys who start under him

14:45.605 --> 14:48.125
who adopt his values, follow him into collaboration

14:48.125 --> 14:49.445
and convince other people around them

14:49.445 --> 14:50.765
to follow him into collaboration.

14:51.385 --> 14:55.485
And so we're using the data of the France in that paper,

14:55.835 --> 14:57.965
because you imagine that the main concern

14:57.965 --> 15:00.245
that people had was like, hang on a second.

15:00.265 --> 15:03.045
You're using the people who served at Verda under Pitton

15:03.265 --> 15:06.565
versus the people who, who served at Verda under the guys

15:06.565 --> 15:07.725
who succeeded Pitton,

15:07.725 --> 15:09.285
but maybe the battle was very different.

15:09.785 --> 15:12.125
So we're able to use the d the daily

15:12.655 --> 15:15.405
fatalities from the multiple La France to show

15:15.405 --> 15:16.845
that it doesn't make a difference, right?

15:16.845 --> 15:19.685
So his removal of command happens on a date

15:20.035 --> 15:21.085
that doesn't correspond

15:21.085 --> 15:23.205
to act any actual change in the battlefield.

15:23.205 --> 15:26.445
There is this whole future research, using some

15:26.445 --> 15:27.445
of the characteristics of the

15:27.645 --> 15:28.765
database for proxies of culture.

15:28.945 --> 15:31.805
But I think there is a lot of interesting question

15:32.225 --> 15:33.885
of organizational economics

15:33.885 --> 15:35.885
or management that could be asked with this database,

15:35.885 --> 15:38.125
because we have such a rich database, we know

15:38.395 --> 15:41.845
what units they were in, we know, their grade,

15:41.915 --> 15:43.845
like whether they're in this soldier or an officer.

15:44.745 --> 15:46.965
Um, and so I think there is a lot of,

15:47.345 --> 15:48.885
and you know, we, we know about,

15:49.105 --> 15:50.485
so you can also reconstruct,

15:50.505 --> 15:52.165
you can aggregate at different levels, right?

15:52.165 --> 15:54.365
So we aggregate at the level of town or department,

15:54.625 --> 15:57.565
but we, you can also aggregate at the level of,

15:57.665 --> 15:59.285
of a regiment.

15:59.385 --> 16:01.765
And then using, you know, who's in charge of these regimens

16:01.785 --> 16:06.325
to see, whether differences in who your leader is,

16:06.465 --> 16:09.045
is leading to different, amount of death threats.

16:09.065 --> 16:11.365
You can also use the characters as an officer, right?

16:11.365 --> 16:13.205
So a super interesting question,

16:13.705 --> 16:16.525
and again, this is something I use in another paper, is

16:16.525 --> 16:19.845
that we see a big change in who the officers of the,

16:20.225 --> 16:21.485
I'm just having this idea now, like,

16:21.485 --> 16:22.565
and maybe I should write this paper,

16:22.665 --> 16:24.325
but you know, maybe other people can do it.

16:24.985 --> 16:26.485
Um, so what happened is that

16:26.495 --> 16:28.845
after this, this battle of the frontier

16:28.845 --> 16:30.765
after the first few months of the war,

16:30.795 --> 16:32.245
that is like the first month,

16:32.245 --> 16:33.765
that's such a disaster for a friend.

16:34.565 --> 16:37.885
A lot of officers are either fired or dead, right?

16:37.885 --> 16:39.805
Because the, an officer in the French Army,

16:40.005 --> 16:43.005
a professional officer has to stand, straighten his boot.

16:43.425 --> 16:45.685
And so a lot of them di died of a bullet in the head

16:45.685 --> 16:47.805
because it's like shameful to duck down.

16:48.785 --> 16:51.125
So there is incredible casualties of these officers

16:51.125 --> 16:53.525
during the first, the first month,

16:53.865 --> 16:56.245
and then they're replaced by reserve officers

16:56.245 --> 16:57.725
who are not professional leaders,

16:58.425 --> 17:00.405
and a lot of them are teachers.

17:00.405 --> 17:02.485
They're just basically educated people from,

17:02.665 --> 17:05.085
so you have this big change in the leadership of the army,

17:05.825 --> 17:09.005
that corresponds to whether your officer was fired

17:09.065 --> 17:10.085
or whether he was killed.

17:10.145 --> 17:11.525
And, you know, whether he is killed

17:11.545 --> 17:12.725
is pretty exogenous, right?

17:12.725 --> 17:14.205
Especially given their mentality

17:14.205 --> 17:15.765
of standing up in their boots.

17:15.765 --> 17:18.005
And so you could see whether the characteristics

17:18.005 --> 17:20.805
of the officer and how congruent those characteristics are

17:20.805 --> 17:23.845
with the troops matter for how they lead the troops

17:23.845 --> 17:25.925
and matter for the death rates of the troops, for example.

17:26.505 --> 17:30.365
Um, so that's, you know, one example of organizational kind

17:30.365 --> 17:31.365
of management questions

17:31.365 --> 17:32.525
that could be answered with this data.

17:32.745 --> 17:34.685
What's really interesting and what was really discovery

17:34.685 --> 17:37.325
of this pa, of this paper,

17:37.385 --> 17:40.725
and the other one is that it's actually very hard

17:40.725 --> 17:45.005
to explain the variation in death rates, specifically

17:45.005 --> 17:47.045
because in in the line, in the line in,

17:47.045 --> 17:48.605
in the line infantry in particular

17:48.755 --> 17:51.165
because troops were rotated to sort

17:51.165 --> 17:53.525
of equalize, death rates.

17:53.865 --> 17:55.205
So French high command wanted

17:55.225 --> 17:57.845
to avoid some revolt in some area of France.

17:58.585 --> 18:01.005
And this was the fear that if you decimate the,

18:01.005 --> 18:03.285
the guys from this area, then people would, might,

18:03.585 --> 18:05.405
might rebel.

18:05.785 --> 18:07.605
Um, so it's actually hard to explain.

18:07.665 --> 18:11.165
So in that other paper, what we, the, the question I wanted

18:11.165 --> 18:14.325
to answer is, is it true that the French nation was made?

18:14.325 --> 18:17.285
Because you had people from Brittany who went

18:17.285 --> 18:20.885
to the same battle as people from Ika, and they met

18:20.905 --> 18:23.085
and they, you know, this is another myth I learned at

18:23.085 --> 18:26.405
school, that the French nation was built in

18:26.405 --> 18:27.605
the trenches of World War I.

18:28.385 --> 18:32.285
So I took our data and a bunch of other data that we have,

18:32.285 --> 18:33.845
and actually, if you look at the data,

18:33.915 --> 18:35.005
it's not at all what happened.

18:35.865 --> 18:39.645
But what happened is that people within,

18:40.445 --> 18:43.845
a military recruitment bureau became much more homogeneous

18:44.155 --> 18:45.605
from the point of view of vote, right?

18:45.605 --> 18:47.725
So you have a vote in 1914 in the parliament

18:47.745 --> 18:50.525
for the ity election one in 1919,

18:50.865 --> 18:54.645
and you see this great homogenization within a recruitment

18:54.645 --> 18:56.445
bureau, but not necessarily across, right?

18:56.445 --> 18:58.005
Some people go left, some people go right?

18:58.265 --> 19:01.165
And so this paper is kind of explaining why, you know,

19:01.275 --> 19:02.965
some went, went left.

19:03.385 --> 19:05.845
Um, and so kind of explaining the, the emergence

19:05.845 --> 19:08.005
of the socialist movements in France

19:08.065 --> 19:11.245
and the role that military recruitment, military rotation,

19:11.985 --> 19:15.005
and in particular the, the characteristics of your officers.

19:15.005 --> 19:17.325
You know, whether they were these reserve officers

19:17.325 --> 19:18.965
who were teachers who themselves or socialists,

19:18.965 --> 19:22.245
and then like diffuse these values within those

19:22.245 --> 19:23.285
regimens or not.

19:23.305 --> 19:25.365
So we, so this is what the other paper is doing.

"Uncovering the Untold Stories of the Great War: A Comprehensive Analysis of French Military Fatalities"

00:00:00 - We really had very little idea about what actually went on, right? So you had a lot of studies by historians, but this is mostly localized studies. So, you know, some people had studied really well what happened to the people who died during the Great War in the Vien department, right? And you knew everything about this or in that other department, or in that village. But there was no comprehensive, overview of what had, what had actually happened. And I think this is the first, you know, and I still remember that moment when I looked at the data and discovered this, is that we historians and the public at large didn't even know what was the day
00:00:45 - in the great war for France. So, you know, we went to school and we were told that the day with the highest level of fatalities was during the Battle of the Frontier. So August, 1914, right? So this was a particular, high fatality moment of the war. And that's, that's entirely true, right? So you see that, you know, when you look at the data, you see that consistently this, the, the highest level of fatalities day after day is really during that period where the German army advance is very fast towards Paris, and then eventually our turnaround at the Battle of the Mar. however, when we looked at the data,
00:01:25 - that the data is specific to metropolitan French shoulders, right? So we don't have soldiers from the colony. And we'll come back to this because I think there is some really interesting things that we've discovered, even with that contrast for particular battles. But what we discovered, and I remember just plotting the time series of daily death, seeing that battle in, you know, in August, 1914, that was supposed to be the deadliest battle for France. And just discovering that by quite a large margin, it wasn't by quite a large margin. The deadliest battle for the day, deadliest day for France was September 25th, 1915.
00:02:08 - that is the first day of the second offensive of champagne, which ultimately was a failed offensive for France. And you have to wonder, you know, why does history select specific dates over others, right? So the Battle of the Frontiers and that specific dates close the Battle of the Mar, which is a very heroic moment for France. And the second day of the Battle of Champagne is a bit of a shameful day for France, especially because that offensive was actually led by peta who then, you know, had a particular role in, in history. We didn't have a good overview of fatalities per day. And then we can reconstruct by battle, and there's a lot of things we can do. So the first thing is to document this in a comprehensive way. And the second thing is that I think we can address a lot
00:02:50 - And also maybe we get to this as well. it enables us to, to peek into, the, some of the cultural characteristics of the population, right? So we can talk maybe about first names and how this database of first name can actually be really interesting to have an idea of, you know, what were the most Catholic areas of France, or what was their areas with the, you know, well, well with the, you know, highest proportion of Catholic names or highest proportions of rural names, you know, people who are named after Kings. We were lucky that citizen and scientists had done an excellent job for us. So basically the hist, the, the story of our database is
00:03:31 - by the French authorities right after the Great War, kept in South Archive, and then these archives themselves, put the database online. And then there was a great initiative of, of citizen scientists called, so like one day, one soldier and people from all over France and, you know, maybe in other parts of the world, went in and read the manuscript card, the handwritten cards and index the database. and so, so we basically took this and cleaned it up. So there was a lot of mistakes, as you can imagine, because sometimes it's really hard to read. sometimes, you know, several people had entered the same person.

"Data Cleanup and Challenges in Analyzing French Soldiers' Demographics During the Great War"

00:04:12 - There was about 200,000 duplicates. and I think the biggest job, and I know because I did quite a lot of it myself, because, you know, we hired a lot of recession system to do it. But really what was difficult is, so, so Victor, my co-author had basically done all the work before getting all the database, and cleaning it up at the, at the level of 87 departments. And then what we did together is we cleaned more fields. So the reasons for death, the rank of the soldier, the circumstances of death, but importantly the municipality of birth. And as you know, France has an insane number of municipalities, right? So I think the whole EU has like 80,000 municipalities.
00:04:55 - or 20,000, 35,000 of those are in France. So it was an insane, job to just, and, and you, you can imagine the amount of spelling mistakes. So we had to clean that all up so that we can merge it to other database that would give us other information about these municipalities. We could have had a process that was a bit more efficient. But we started this before AI was what, I mean, you know, we're really talking ai, like it's really the last year, right? So Chad g PT is amazing now at reading handwritten documents, but that wasn't the case when we did this thing. So it's really development in the last year or two years. And so we did it just before.
00:05:38 - what matched and then go over the ones that didn't match and manually, clean them. I mean, you know, you could automatize part of it, right? You could, you know, force it too much to some database. But yeah, we had, very good RAs from, the University of Toulu and a few, yeah, and I, I dunno, it, it is not a very, you know, it's time consuming, but it, yeah, it's, it's, it's feasible. And, and we matched I think 99.9% of the soldiers to their municipality part. There's a lot of things that we could do at scale that hadn't been done before. So in particular, what we do in, in the paper is
00:06:17 - So trying to see what were the characteristics of soldiers that made them last more, you know, longer or shorter duration on the front. And so here you find clearly, you know, that officers had a big advantage, you know, so on average an officer would, leave for 134 days more than an enlisted soldier. and then you also see the, the big hit that the line infantry took. But, and you can even see a discontinuity. So the, the, the men, so the line infantry, are the active army was the youngest guys. I think it's important to say, because, you know, modern audiences cannot understand, even though it was a hundred years ago,
00:06:58 - So prince, in 1914, everybody had to do three years compulsory service. Every man had to do it. And they would do their three years compulsory service. they would start the year, they would turn 20, and then they were in the army for 28 years. So you would do three years of military service, which would mean you would be in the active army for three years, then you would be in the reserves of the active army for 11 years. Then you would go to the territorial army for seven years. And then you had seven more years in the reserve of the territorial army. And all these men were mobilized in August, 1914, right? So it's about, it's more than 8 million men,
00:07:40 - of 20 and 48. You know, I don't know how old you are, but you know, we have this vision that war is a thing for young men, but it's not the case, right? So you had, and and you have very few people in France who actually escape, the drafts, right? So we're not in a situation like, you know, in other countries where a lot of people didn't do their military service and didn't go to service war. And there's a very simple reason to this, is that France has a big demographic deficit compared to Germany because France had the demographic transition much earlier than other countries in Europe. So we don't have enough men, right? So we're, we're fighting a war against an opponent

"Exploring Cultural Indicators Through First Names: Insights from Historical Data"

00:08:23 - So it's very hard to escape, to escape military service at the time. You know, we talk a little bit more about in the paper that some men were withdrawn from the front to work in factories. but we're only talking about, you know, 500, five, you know, 550,000 men overall. So it's not a huge amount, you know, out of 8 million, people, we did this little fun exercise and we want to write it up because there is a, a very active literature on culture and economics, and people have issues measuring Culture in the past, right? So now we have a whole bunch of measures of culture. We have surveys, we have what people post on Twitter,
00:09:05 - we can't access, that data so easily. But, you know, there is reddi posts, like people can scrape social media and have a sense of what people are talking about and, and what's going on in the past. We don't have very good measures of culture, right? We can have, you know, what's written of tombstones, you know, how many churches there are, but we also have people's first names, right? Because people, when you name your child, you convey some value, right? So there is a whole literature, there is a literature showing that whether a kid has, a Christian first name is actually quite well correlated with even occupational church. Like, these people are more likely to become priests. And, you know, so, so
00:09:45 - We can measure. And you know, it is particularly interesting in the case of France, 'cause the census doesn't ask about religion, because it is kind of forbidden. So since the revolution, the census, there's a few exceptions, but in general, we don't ask about the, the franchises doesn't ask about religion, but then we have this first name, and so we can, we get a measure, right? So if, if a, and actually, so the most common first name in France at the time is Jean. And actually a lot of men have Marie as a, as a first name, right? As a second name or third name, right? So this could be a strong indicator of how Catholic your parents are, right? So if you're called Marie as a man, it's a strong indication
00:10:27 - Then you also have people whose first names are after former kings, right? So you have Louis Francois, and actually it's, it's pretty, it's actually interesting. So if you plot at the municipal level, the proportion of people who have a name after a king, you see concentration in VDI and north of Belfor, which were very stronghold of anti-revolutionary revolts, right? So, so even in 1914, you see some reminiscence of opposition to the revolution, and people, you know, manifest their opposition by calling their kid Louis,
00:11:09 - and then you also have more or less common names. So there's also a literature that that correlates, the fraction of rare names, rare first names with individualism. So people who are more individualistic tend to have a more rare first name for their child. So you can also have a measure of this. And then the map of France is pretty interesting, right? So from the point of view of religion, we have a map that corresponds pretty well to a map that an, a re a priest had done in the forties to, he went around every church in France to measure attendance. And you really have this empty middle of France.
00:11:52 - and the south of France, and then this, this middle region of France, like the diagonal, like this is actually much less strongly religious than the rest of France. So, you know, at the moment it's just measurement things. And we have this data and we haven't really done much with it, but, it's, it could be useful for, for future research. And yeah, so another thing that we did as well is that we correlated it, we correlated the fraction of Catholic, of, Christian first names with the share of refractory clergy during the revolution. And the correlation is quite strong. I have another paper with another group of co-authors, and we look at the relationship between serving at the Battle

"Legacy of Leadership: The Impact of World War I Commanders on Collaboration in World War II"**

00:12:35 - during the First World War, and the propensity of people in those municipalities to, become collaborators in World War ii, right? So maybe the, the, you know, it's a, so Pitton was a general in the First World War, he becomes a marshal at the end of the war because he comes out as this hero of world. I, and when, France is invaded by Germany in World War ii, they call him, to save France, you know, again, except what he does is that he puts in place an authoritarian regime that is collaborating with Germany until 1944. and so we can establish this link, between, and,
00:13:18 - and the whole of the interwar is that because he was such a hero of the first of the Great War, and so we actually show in this paper that the, there's this transmission of values that the people who serve under him adopted his values. And that's true of every veteran, because we can trace what municipality served under him at different times. And you can see that any municipality that belonged to a regimen that served under him during the First World War, is so the, the veterans are more likely to become collaborators, the people who served under him, but specifically in the municipality that served under him at Verda, you see that these people,
00:13:57 - they transmit these values to the rest of the people in their municipality. So it's not only they're just the veterans, it's really because themselves, they're heroes. Everybody follows them, right? So the, the old, the grandpa who's this amazing veteran Froman and every, and you know, who's still been telling his son, the only reason you're alive is because of, bean, like they have the fact of going to is such a heroic, trait for the, for the, for the veterans, is that it's, it's both a combination, right? Because we've basically exploit who was at Verda under Marshall peta, and then who served at Verda after he was removed from command. So the guy who served after he was removed from command,
00:14:38 - So their heroes, they can transmit values to their local communities, but they don't have these values. So it's really specifically the guys who start under him who adopt his values, follow him into collaboration and convince other people around them to follow him into collaboration. And so we're using the data of the France in that paper, because you imagine that the main concern that people had was like, hang on a second. You're using the people who served at Verda under Pitton versus the people who, who served at Verda under the guys who succeeded Pitton, but maybe the battle was very different. So we're able to use the d the daily fatalities from the multiple La France to show that it doesn't make a difference, right? So his removal of command happens on a date
00:15:21 - to act any actual change in the battlefield. There is this whole future research, using some of the characteristics of the database for proxies of culture. But I think there is a lot of interesting question of organizational economics or management that could be asked with this database, because we have such a rich database, we know what units they were in, we know, their grade, like whether they're in this soldier or an officer. and so I think there is a lot of, and you know, we, we know about, so you can also reconstruct, you can aggregate at different levels, right? So we aggregate at the level of town or department, but we, you can also aggregate at the level of, of a regiment. And then using, you know, who's in charge of these regimens
00:16:06 - is leading to different, amount of death threats. You can also use the characters as an officer, right? So a super interesting question, and again, this is something I use in another paper, is that we see a big change in who the officers of the, I'm just having this idea now, like, and maybe I should write this paper, but you know, maybe other people can do it. so what happened is that after this, this battle of the frontier after the first few months of the war, that is like the first month, that's such a disaster for a friend. A lot of officers are either fired or dead, right? Because the, an officer in the French Army, a professional officer has to stand, straighten his boot.

"Transformations in Military Leadership and Their Impact on Social Homogeneity During World War I"

00:16:45 - because it's like shameful to duck down. So there is incredible casualties of these officers during the first, the first month, and then they're replaced by reserve officers who are not professional leaders, and a lot of them are teachers. They're just basically educated people from, so you have this big change in the leadership of the army, that corresponds to whether your officer was fired or whether he was killed. And, you know, whether he is killed is pretty exogenous, right? Especially given their mentality of standing up in their boots. And so you could see whether the characteristics of the officer and how congruent those characteristics are with the troops matter for how they lead the troops
00:17:26 - so that's, you know, one example of organizational kind of management questions that could be answered with this data. What's really interesting and what was really discovery of this pa, of this paper, and the other one is that it's actually very hard to explain the variation in death rates, specifically because in in the line, in the line in, in the line infantry in particular because troops were rotated to sort of equalize, death rates. So French high command wanted to avoid some revolt in some area of France. And this was the fear that if you decimate the, the guys from this area, then people would, might, might rebel.
00:18:07 - So in that other paper, what we, the, the question I wanted to answer is, is it true that the French nation was made? Because you had people from Brittany who went to the same battle as people from Ika, and they met and they, you know, this is another myth I learned at school, that the French nation was built in the trenches of World War I. So I took our data and a bunch of other data that we have, and actually, if you look at the data, it's not at all what happened. But what happened is that people within, a military recruitment bureau became much more homogeneous from the point of view of vote, right? So you have a vote in 1914 in the parliament
00:19:23 - and you see this great homogenization within a recruitment bureau, but not necessarily across, right? Some people go left, some people go right? And so this paper is kind of explaining why, you know, some went, went left. and so kind of explaining the, the emergence of the socialist movements in France and the role that military recruitment, military rotation, and in particular the, the characteristics of your officers. You know, whether they were these reserve officers who were teachers who themselves or socialists, and then like diffuse these values within those regimens or not. So we, so this is what the other paper is doing.

Faculti AI can make mistakes

Research

Policy

Analysis