I started with a search problem
When I first thought about Culture Atlas, I assumed the problem was finding information.
That seemed obvious. India is enormous and culturally complicated, so I thought the useful thing would be to collect more information in one place.
But the more I searched, the less true that seemed. There is already a huge amount to discover about India's languages, places, communities, food, music, art, architecture, festivals, histories and stories. The problem was not simply that the information did not exist.
The problem I kept running into was what happened after I found it.
An answer often created a better question
If I discovered a language, I wanted to know where it was spoken, which communities used it, what literature or oral traditions existed in it, and what other languages had influenced it.
If I discovered a tradition, I wanted to know who practised it, where it was practised, whether it had changed over time, what materials or skills were involved, and where I could see or hear it.
If I discovered a person, I wanted to understand the places, events, institutions, communities and ideas around that person.
Each answer was creating a new direction. But most websites made me stop, go back to search, and begin again.
I realised that what I wanted was not just information. I wanted continuity of curiosity.
A list can be organised and still hide relationships
One obvious way to build Culture Atlas would be to make a page for every state and then create categories underneath it: food, language, festivals, art, people, monuments and so on.
That would be tidy. It would also create a problem.
The structure would quietly suggest that every cultural thing belongs inside one administrative box. It would make the categories more visible than the relationships between them.
I could also make a huge directory: hundreds or thousands of entries arranged alphabetically or by topic. That would contain information, but it would still require the visitor to know what to look for.
I want Culture Atlas to work even when you do not know what to search for yet.
The connection itself can carry information
This was a new idea for me.
Suppose the Atlas knows that a language is associated with several places. That relationship tells you something. Suppose a tradition is practised by a community and documented by a museum, an archive and an oral-history recording. Those relationships also tell you something.
The links should not just be buttons saying “related content.” They should have meaning.
That means I need to think about two things every time I add a record: what is this? and how is it connected?
The second question is often harder.
Connections also create responsibility
Once I started thinking this way, I noticed another problem. It is easy to create a connection that sounds reasonable but is too vague.
“Related to” is not always enough. Is a person born in a place, working in a place, associated with a movement there, or simply mentioned in a source about it? Is a language historically present in a region, currently spoken there, or officially recognised there? Is a tradition practised by a community, originated by it, or just popular in the same location?
I am learning that if the relationship is important enough to show, it should eventually be clear enough to explain.
This is one reason I do not want the Atlas to become a giant web of random links. More connections are not automatically better. Useful connections need structure.
What this changes in the product
If Culture Atlas is really about connections, then the product has to support exploration in more than one direction.
A place should not only list things “inside” it. A language should be able to lead to places and communities. A community should lead to traditions, people and stories. A story should lead back to the people, places and sources connected to it.
Search still matters. Categories still matter. Maps still matter. But none of them should become a dead end.
I am beginning to think of every record as a possible starting point rather than a final page.
A useful test I am giving myself
When I add something to Culture Atlas, I now try not to ask only, “What information belongs on this page?”
I ask:
What is this connected to?
What kind of connection is it?
Would following that connection help someone understand more?
Can I support the connection with a source?
If I cannot answer those questions yet, that may mean the record needs more research.
But connections can mislead too
I also have to argue against my own idea. A connection on a screen can look important simply because I drew a line between two things.
That means Culture Atlas should not reward me for making the biggest possible web. A smaller number of clear, supported relationships may teach more than hundreds of vague “related to” links. The connection has to earn its place.
Seeing a pattern is only the start
Maths has made me more alert to patterns, but it has also made me ask whether a pattern is actually supported. Two things appearing together does not automatically tell me why they are connected. That matters here too. In I&S, I need historical or geographical context; in English, I need to explain the relationship clearly enough that another person can follow my reasoning. A Culture Atlas connection should survive all three tests.
A connection has to explain something
I began with the idea that Culture Atlas needed to organise information better. I now think organisation is only part of the problem.
The more interesting challenge is to preserve the paths between pieces of knowledge. That is what can turn a collection into an atlas you can actually explore.
What I'm figuring out next
If the relationships are this important, they cannot live only as links inside paragraphs. They need a structure underneath the site.
That leads to the next part of the build: deciding what the main “things” in Culture Atlas are, and how a cultural graph could connect them without becoming impossible to understand.
Follow the build
I am documenting Culture Atlas while I build it — including the ideas that change and the things that do not work.