I think this map highlights some fundamental flaw with wikipedia or what we highlight in history or something similar to this.
Or possibly it's a perfect example of what is wrong with summarising a complicated entity. I hope in the future that people will move away from the Top Ten things about X or the 5 things you need to know about Y.
Seems to me that your second idea is probably pretty accurate. History books (general purpose ones, that is) cover wars extensively. On the one hand, wars cause rapid and massive political and social changes, so they are obviously very important to understanding history. But at the same time, understanding what caused a war, what the political and social aftermath looked like, and the cultural perceptions of the war are also extremely important. Not to mention all the non-war history. So there is probably a good argument that history books focus too heavily on wars. So it is not surprising that Wikipedia would also focus heavily (perhaps too heavily) on wars.
> But at the same time, understanding what caused a war, what the political and social aftermath looked like, and the cultural perceptions of the war are also extremely important.
And those paragraphs will contain the word "war" many times.
I guess there's a few reasons why. A big reason for learning history is learning from exceptional events, and wars are exceptional. Learning how to avoid them (if possible) and win them (if necessary) isn't something you want each generation to learn by trial and error.
Also, wars are kind of interesting. A good guy (us) and bad guy (them) is a fail-proof way to make a story mildly interesting.
I guess there's other lessons we should learn from history (as opposed to trial and error, or observation). Political systems are a good one - there's things we could learn from the Greeks which could apply today (like sortition, as a way to stop career politicians getting too powerful, but still giving the decision makers enough power that they have an incentive to properly study the issue). We could also learn about dealing with natural disasters, resource exploitation, and the history of moral panics.
I suspect if you had lived though a significant war you would probably think history was overly focused on social changes and should spend more time on technology and tactics.
Even the people who lived through actual battles (where technology and tactics would be relevant) tend to tell stories that are mostly about hunger, deprivation and being cold. Shooting people and executing strategic plans hardly ever figure in them.
And this makes sense: most soldiers spend most of their time waiting around for battles or marching from one battle to the next. The actual battles themselves take up very little of their time (even if they're extremely stressful).
The stories of those that lived through wars as civilians (the vast majority of people) of course never talk about technology and tactics as they had nothing to do with them.
War is often vary 'hands on' for at least one civilian population. Did William Tecumseh Sherman decide to burn your barn? Well now you care about tactics.
But you're still not likely to tell stories of the battlefield tactics and grand strategy that led to the burning of your barn, but rather than deprivation and hardship that not having that barn caused.
It'd probably be better if it used a better measure than raw term frequency in the articles. For example, most of Europe is "War". If it used something like inverse document frequency (tf-idf) to rank the terms, it might get more distinct words for each country.
That might be so if your goal is to hilight what sets the countries apart.
The author of the map chose for 'history of __' as the topic. And when doing that with countries it is not surprising to find that countries are defined by war, nobility, imperialism and rivalry.
We like to think countries are formed because of their people, and their culture, but in reality countries are formed because of their resources, defendability and their interested parties.
If there is one thing that defines borders in europe, it is war, and this map shows that quite laconically :)
Yeah, I'd be interested in doing an associative version of this (sort of like what we did for this[1]), though my mapping skills are not quite up to the author's/
I was hoping for more words when you zoomed, like a hierarchical wordle, but I realise that that was not the purpose of the map. Interesting way to get students thinking about history.
That sounds like a great idea. I hope the original poster picks up on your idea. Although that may still leave some of the offensive terms listed elsewhere in the comments (Kuwait <- Iraq, Ethiopia <- Eritrea)
This would be an amazing demonstration of cultural relativism.
You used the "en.wikipedia.org" site for the source. Good. Now put a loop around the whole thing, and after you're done making the English-"history-of ____" map (which you have just done), repeat with the French - "history of___ map", and so forth for every language that has entries (some may be blank for some countries).
Next, you can take your 40-80 or whatever maps (whichever languages you decided to complete this for), and translate them via automated translation into English.
So, we get "the different language Wikipedia's different views of the world", with English (untranslated) being only the first source, and additional sources both native and translated directly into English.
For example, while for Russia, "soviet" won out from the English-language
Eh... I'm from Malta. Representative or not, I'm thinking "a" or "I" might be the only words which fit there...
Reminds me of a department manager who insisted everybody place a push pin on a world map he hung up in the office, to show where everybody was originally from. I obliterated my own country in so doing. After that most people assumed I was born on a cruise ship or something...
I was puzzled by this, particularly by the number of countries that were marked with "war". So, I did what any programmer might do, I wrote some code to check it out. Here it is:
I just threw it together quickly. It's got some warts (not the best code for grabbing data from Wikipedia). Of course, feel free to do with it as you wish.
I only sampled a couple dozen countries or so. It seems that the author of the graphic decided to take a little license here and there rather than making strict use of the word with the most usage.
For example, if I didn't screw-up the code, in Canada the most frequent words are:
[ 143] books
[ 118] canadian
[ 95] british
[ 82] war
However, the author decided to use "british".
In Argentina:
[ 53] president
[ 50] argentine
[ 47] war
[ 46] military
Yet he used "war"
Brazil:
[ 28] military
[ 22] portuguese
[ 19] portugal
[ 16] first
[ 15] war
The choice was "military"
Saudi Arabia:
[ 114] saudi
[ 76] arabia
[ 37] king
[ 31] oil
[ 28] arab
Chosen: "oil"
I am not passing judgement at all, just pointing out what I learned from a very limited test. At some level It would be interesting to see data for all countries. At the same time, a better metric than single-page word frequency would probably be more interesting than this. For example: http://en.wikipedia.org/wiki/Tf%E2%80%93idf
In the interest of completeness I updated the code based on some on and off-list feedback. The new results are much cleaner as they ignore all anchor, image, navigation and other text. I've updated the gist with the new code. Here's a sample output:
United_States
[ 118] war
[ 108] states
[ 96] american
[ 78] united
[ 42] first
[ 37] president
[ 35] rights
[ 35] world
[ 34] government
[ 33] economic
England
[ 57] henry
[ 57] king
[ 54] english
[ 43] government
[ 41] britain
[ 39] century
[ 38] war
[ 37] edward
[ 34] local
[ 34] kingdom
Brazil
[ 23] military
[ 17] portugal
[ 15] first
[ 14] portuguese
[ 13] years
[ 13] war
[ 13] president
[ 12] political
[ 12] country
[ 10] called
Argentina
[ 51] president
[ 43] military
[ 41] argentine
[ 40] government
[ 37] war
[ 36] political
[ 30] buenos
[ 30] aires
[ 24] against
[ 23] national
Canada
[ 71] british
[ 62] war
[ 53] canadian
[ 38] north
[ 37] french
[ 35] quebec
[ 34] government
[ 34] first
[ 32] american
[ 27] france
Spain
[ 103] spanish
[ 62] war
[ 46] century
[ 39] empire
[ 38] king
[ 34] under
[ 29] peninsula
[ 29] first
[ 28] roman
[ 27] country
Portugal
[ 99] portuguese
[ 28] first
[ 25] war
[ 22] king
[ 21] spanish
[ 20] independence
[ 19] east
[ 19] republic
[ 18] between
[ 18] main
Russia
[ 105] soviet
[ 105] russian
[ 68] war
[ 52] union
[ 43] citation
[ 41] century
[ 39] europe
[ 39] state
[ 38] government
[ 37] moscow
Germany
[ 240] german
[ 88] war
[ 74] empire
[ 68] east
[ 63] states
[ 57] west
[ 55] france
[ 55] prussia
[ 46] austria
[ 42] between
France
[ 221] french
[ 151] war
[ 114] louis
[ 95] king
[ 68] paris
[ 56] against
[ 52] government
[ 52] republic
[ 49] would
[ 49] empire
Japan
[ 89] japanese
[ 58] war
[ 50] emperor
[ 40] main
[ 37] military
[ 37] china
[ 35] imperial
[ 35] ashikaga
[ 35] power
[ 34] government
China
[ 127] dynasty
[ 62] chinese
[ 29] han
[ 26] emperor
[ 26] shang
[ 25] zhou
[ 25] ming
[ 22] empire
[ 20] military
[ 20] main
Saudi_Arabia
[ 83] saudi
[ 48] arabia
[ 28] oil
[ 28] king
[ 21] hejaz
[ 21] arab
[ 19] mecca
[ 18] first
[ 16] saud
[ 16] state
Ah, yes. I didn't do a full cleanup job before processing and you caught one of those words. If you open that page with Chrome and do a "View Source" you'll see that "books" actually appears 597 times in a whole pile of links on the page. Some of them got through.
If you only search what a human can see, yeah, it's somewhere in the order of 3 to 5 times.
I might have a look at the code and see why the full <a> tag contents are not being removed. Funny enough, that's the only bit of code I did not write. I got lazy and grabbed something right out of the PHP manual pages. Lesson learned.
Using a text-mode browser to render HTML to text, strip off headers, and avoid summarizing links (lynx works great for this) might be a fast route to cleaner results.
You'd probably also want to exclude the "See also" and "References" sections.
For Canada, not excluding stopwords ("the", "of", "and", etc.) or the country name itself, I get:
1 869 the
2 449 of
3 326 and
4 288 in
5 214 to
6 142 a
7 135 canada
8 106 was
9 87 by
10 72 as
11 71 british
12 70 with
13 69 s
14 62 war
15 61 on
16 53 canadian
17 52 from
18 49 for
19 44 were
20 42 new
Yup, in case you missed it, I posted updated code that really cleans-up the scraped page data very well. With the new code "british" is on top of the Canada list. It was a fun little exercise. There are some issues accessing Wikipedia this way but no big deal for a quick little test.
It's interesting how Nicaragua is "defined" by a family name (Somoza). Other family names I see include Kim for North Korea and Khaan for Mongolia. They're all analogous to Dynasty for China.
‘No war’ for Sweden is only defined in terms of not declaring war on another country; Sweden has still sent soldiers to Congo, Afghanistan and Libya to take part in military action in more recent times: https://en.wikipedia.org/wiki/List_of_Swedish_wars
There is a big difference between being at war and sending peace keeping troops. Defining the use of military in an area to prevent further violence and war as being at war is very wrong.
The author has stated[1] that the names of the countries were excluded for consideration in the choice of most frequent word.
@atsaloti "New" really should have been excluded as the most frequent word. It appears a lot in the names of New Guinea, New South Wales, New Holland... It would be interesting to know what word came afterwards in frequency.
That's what many people say - but on the other way, that's how I make my living! And surprisingly, it works (it's not $9.99 per wallpaper, as you might think - it's for entire account, unlocking everything I have to offer).
Yes I thought you might say that. But I personally feel like this is a psychological issue - I am quite unwilling to pay 9.99 for an account, to get access to all the stuff you make(no offense, I think your art is great). I like this wallpaper, I want it now,and I would be more than happy to pay a single-time fee for it - 1.99,2.99 or even 4.99 - that would be psychologically easier to stomach than getting an entire account made,which I don't want or need.
I'd love to see the results for other text queries. E.g. use the cuisine of X range of articles as a source. Maybe filter the results to only include actual foodstuffs. The typography is completely custom, though, so there's no way to automate the process.
I would bet that if you did this same things with many history books, similar words would come up, but I do not think that it is a fair summary.
It would be more interesting to see the most searched Google queries coming from these countries (assuming that many use Google). Finland is Ilta Sanomat (The Evening News), for example: http://www.google.com/zeitgeist/2012/#the-world
Unfortunately, Antarctica is not represented in Google Zeitgeist, nor are many others, but perhaps you could contact Google for that data.
Spend enough on Adwords or google appss and you get a dedicated rep. Buy a nexus and you get a 24/7 phone number. Post something about google virtually anywhere on the Internet and Matt Cutts will probably reply to it.
As confronting as it is, most of these seem to me spot on. Most of European history is indeed about war. Most of the African and Asian history that is written about is indeed about who colonized what.
It's a great map: it provides an easily digestible visualization of the limitations of history writing.
The comments seem to prove why the map is useful. What would you like your countries word to be? Work to that end. Or find a word you like and emigrate.
The entry for Eritrea shows that the word "Ethiopia" comes up a lot in the wikipedia article "History of Eritrea". Given the history of the two countries, that's to be expected. What exactly are you taking offence at?
Mainland China would most definitely not be happy with this label. The government that was exiled to what we call Taiwan considers themselves to be the legitimate government of China. The mainland government considers them to be illegitimate. This tension still exists politically and culturally.
I have had many Taiwanese friends young and old and most of them prefer themselves and their culture to be referred to as Chinese.
I have also had some young (mainland) Chinese friends and they, too, are still very aware of the situation. Even though some of them have Taiwanese friends themselves, they still look with disdain on Taiwan/RoC as a political entity.
I visited China about a year ago and was on a train when a local started talking to me. He asked why I was there and what cities I've seen, mainly small talk. I had lived in Taiwan for about a year prior so when he told me I should live in China I said, "well, I lived in Taiwan for a year". His face went cold and his tone became somewhat ominous. His response was "Taiwan IS China" and in a threatening manner said "RIGHT?". That really made an impression on me. The Taiwanese people don't consider themselves part of China, but would someday like to see the ROC regain control of continental China. When they think of China they think of the People's Republic of China, the communist China.
Or possibly it's a perfect example of what is wrong with summarising a complicated entity. I hope in the future that people will move away from the Top Ten things about X or the 5 things you need to know about Y.
This map is ignorance presenting as a summary.
(Yes I know it's meant to be a cute map)