At Lux for Predictive Anthropology, we understand culture by decoding topics of conversation to determine what those topics mean to people.

But we’re not interested in the dictionary definition of these words. We want to figure out how those topics play a role in shaping meaning for people, when they put those topics in the contextual web of meaning that makes up their daily life.

We analyze topics to decode the role that topic of conversation plays in defining a culture.

Let’s dive into a simplified example, before we get to a definition: an example is worth hundreds of pages of explanation.

In our work, the AI Anthropologist analyzes hundreds of thousands of consumer verbatims before the data is presented. But in this case, lets look at three consumer verbatims on the topic of chocolate cake: a number much more suitable to humans!

When we perform cultural analysis, we understand this conversation about chocolate cake as portal into the culture of chocolate cake.

By analyzing this conversation, we determine what chocolate cake means to these people.

We call the totality of these meanings the macroculture: the broadest culture of chocolate cake and all those engaged with it.

We call the distinct subgroups of meaning the microcultures: the smaller, sub-cultures that share a unique meaning of chocolate cake.

To really make this clear, we need to talk about meanings.

Let’s go back to those consumer verbatims about chocolate cake, through the lens of cultural anthropology.

We can see a bunch of different meanings that highlight the role chocolate cake plays in people’s lives.

Note that none of these have to do with the dictionary definition of chocolate cake (a “sweet food made by baking a mixture of flour, eggs, sugar, fat, and cocoa in an oven”).

Instead, these five meanings are distinct understandings of the role cake plays in people’s lives. In our custom reports and Trends tool, we study these meanings as microcultures.

Another important thing to note is that the second and third comment never refer to chocolate cake.

Rather, we rely on contextual understanding to know that “chocolate one” and “it” refer to chocolate cake.

We also use contextual meaning to understand the “Yes!” in verbatim as a recapitulation that “it’s not a birthday without chocolate cake.”

When we crunch our data, our AI engine extracts these contextual meanings to ensure we’re capturing a full picture of culture, the maturity of ideas, and the population of consumers participating in that culture.

The above culture is comprised of three comments. Now scale that up to millions of data points in a vector data base and you have an idea of what our big data cultural analysis involves.