The Books on
My Shelf

A bookshelf is data before we call it data.

It contains information about what we read, what we are interested in, what languages we use, what we choose to keep, and what we have bought or been given.

Usually, we simply see a collection of books. I wanted to see what would happen if I started treating that collection as a dataset.

How does changing the way we look change what we see?

I recorded a set of attributes for each book: subject, language, owner, reading status, how it was acquired, approximate size, and my own evaluation of it.

Once the books had been described in this way, the same collection could be reorganised and viewed through different dimensions.

The process sounds simple, but it immediately involved choices.

What counts as a subject? How should a book be classified? What does “read” mean? How do you turn a personal judgement into a number?

Even a small dataset is shaped by the decisions made while creating it.

Changing the view changes the patterns that become visible.

A shelf organised by ownership tells a different story from one organised by language, reading status or subject. The books remain the same, but the relationships between them become easier to notice.

This is what I find interesting about data: it does not always begin with a spreadsheet, a survey or a large statistical database.

Sometimes it is already sitting in front of us, hidden in an ordinary collection of things.

The bookshelf was familiar to me before I recorded it. But once I turned it into data, relationships that were difficult to see at a glance became possible to explore.

The experiment is small, personal and inevitably subjective. But that is precisely the point.

Data can be found almost anywhere. The interesting part is deciding what to notice.