Last week I discovered Unsuck-it and quipped, “When will somebody create a de-jargon-izer for International Development? The storytelling project is starting to demonstrate that people working at non-governmental organizations all over the world speak a different language from those in the communities they help. And The Economist recently explained why “NGOish” matters.
I searched around for a dictionary, glossary, concordance, or lexicon of the specific technical words NGOs use – but alas what’s out there isn’t complete. DFID provides one glossary, Development Frameworks another, an Canadian Geographic provides a third one – but nothing complete – so I made one.
By pulling all the text from 4970 projects on GlobalGiving, I was able to rank the 2000 most common words used by 1500 organizations working in 100 countries. For contrast, I ran the same python script on 3500 community stories gathered in our GlobalGiving Storytelling Project (summarized in this PDF). Complete data files for both sets are found below.
By comparing this development lexicon to the story lexicon, I think we will better understand how the language that trained NGO staff people speak differs from that of the communities they serve – at least in East Africa. I encourage you to download and analyze both text documents.
The Feedback Switchboard
This is all Interesting – but not my real purpose. These two lexicons actually allow me to do something useful.
My real purpose is to deliver each one of these organizations a personalized digest of the most relevant stories from the towns and villages where they work. To do that, we needed a set of words that reflect the NGO to match against a set of words that represent each story. This is a bit like OK Cupid’s analysis of dating sites – for the NGO world. Matched keywords, combined with story locations and some direct attribution by the storytellers themselves should allow us to direct every incoming story to the NGO that ought to be listening – turning GlobalGiving into a sort of feedback switchboard. Or maybe I it is like an Facebook profile for a NGO, in which case this matching method allows us to recommend people they should be following in their community.
It’s just another experiment in seeing what ways we can inform and entice local organizations to consume local knowledge that we’re collecting on their behalf.
Download the development lexicon (2000 most used NGO words. This is a plain text file: rename to .txt)
Download the story Lexicon (2000 most used words in communities. Plain text file: rename to .txt)
Mapping story words:
Mapping the “when” (timeline) and the “where” (geocoding) of information is much easier than visualizing the what. This Google Motion Chart (coded by GapMinder.Org) is a feeble start. I’d prefer to be able to sort all the words by NGO and Location, to combine Who-What-Where:
Note: The “height” on the Y-axis is a calculation of reliability of the stories that share each keyword. “Reliability” is based on diversity of story sources: If many locations, NGOs, storytellers, and scribes are involved, divided by the number of stories in that keyword bubble, then those stories rise to the top. As you might expect, keywords at the top are universal to most stories.
Sample: First 100 of 2000 common NGO words
(format: word, number of instances):
children,9224 education,7026 school,5930 women,5422 health,4894 community,4538 training,3810 students,3482 water,3385 people,3354 families,3164 help,2780 schools,2775 girls,2658 rural,2578 youth,2509 food,2372 care,2326 skills,2205 local,2130 poverty,2128 access,1985 poor,1858 social,1812 sustainable,1618 life,1563 work,1558 income,1548 new,1546 economic,1516 services,1482 high,1446 need,1432 rights,1330 medical,1325 educational,1279 public,1279 young,1271 improve,1250 learning,1244 living,1141 orphans,1129 hivaids,1122 family,1120 africa,1118 environment,1052 hiv,1040 building,1011 lives,1009 opportunities,1000 activities,998 basic,983 quality,979 human,968 technology,960 safe,929 international,916 materials,914 business,913 resources,900 use,897 village,894 fund,878 district,878 areas,876 change,868 supplies,863 working,855 build,851 lack,849 teachers,847 vulnerable,843 literacy,840 awareness,832 home,826 center,825 years,823 small,810 create,810 violence,809 sanitation,803 information,801 increase,787 better,779 farmers,768 free,766 future,764 parents,757 primary,757 relief,744 villages,738 arts,738 aids,728 empowerment,724 clean,722 live,711 leadership,696 sports,696 environmental,695 learn,691
“International Development Buzz”
Reprinting an old game. Use this chart to generate thousands of jargony NGO phrases: