# Are there “6 degrees of separation” for ideas?

Sites like linkedin.com are built on the idea that you can be related to any person in the world though around 6 people. One person knows another one, that one another one, and in the end you are connected through intermediate connections.

I'm interested if something similar exists for ideas - does human mind naturally connect and infer from different ideas to find a relationship between ideas? I'm thinking about this in the context of Freudian dream interpretation. In his book on the analysis of dreams, Freud repeatedly lists examples where he uses analysis and logic to find meaning in dream imagery and experiences.

This makes me curious if what Freud has discovered is some function of memory or thinking that establishes connections between unrelated ideas through intermediate ones.

For example, there is no connection between cats and space flight. But ones mind may try to connect them through a cartoon episode which involved super intelligent cats from the outer space invading earth in spaceships. Now there is a connection between cats and space flight.

What I'm trying to ask is if it is natural for the mind to bridge gaps between unrelated ideas through intermediate ones, especially if/when a person is actively looking for a connection? if there are 6 jumps between two unrelated ideas/concepts, then there are 6 degrees of separation.

-
Could you further explain how you see the 6 degrees of separation idea relating to the general issue of the brain or mind implicitly connecting ideas? –  Jeromy Anglim May 9 '13 at 4:55
"6 degrees of separation" is the closest term I could think of for the idea connection phenomenon I'm trying to understand. –  Alex Stone May 9 '13 at 12:42
@AlexStone I don't think "6 degrees of seperation" is related to your question, unfortunately. I think that what you are looking for is the more general "intermediate ideas" or "intermediate links" or "network structure". I tried to explain that in my answer. –  Artem Kaznatcheev May 9 '13 at 13:24

You are asking two different questions in your title and the text. I will answer the text question, explain why you are asking two different questions, and comment on the title question.

The human mind seems to have some sort of relational structure for how it works with ideas. This is usually study in the domain of memory via free recall. In a free recall experiment, a participant will memorize a list of items and then be asked to recall them in any order. Alternatively, there might be no list to memorize and instead the participant may be asked to name as many of some common things as they can quickly (for instance: name all the animals you know). The experimenter tracks the time delays between recall of items, and uses that to infer the "mental distance" between concepts.

It a typical experiment, you will see clustering effects with participants quickly naming a bunch of objects of one type before taking slightly longer to transition to another type and then name many objects of that type quickly. As an example, I might be asked to name cartoon characters and I would start listing:

"Bart" [short delay] "Homer" [short delay] "Marge" [short delay] "Lisa" [short delay] "Maggie" [long delay while my mind switches from 'The Simpsons' cluster to 'Futurama' cluster] "Fry" [short delay] Leela [short delay] Dr. Zoidberg [etc]

This suggests that I naturally connect and infer different relationships between ideas. In particular, the temporal structure of my recall reveals that I known concepts like "The Simpsons" and "Futurama" and I connect certain names to each other because they share a relationship of "being on the same show". Of course, how I structure my categories and recall is shaped by my expertise.

As an experimenter, I could represent the way people structure their ideas as a graph/network, with individual ideas represented as nodes and closely related ideas (or ones between which people usually transition often or quickly) connected by edges. Free recall is then a random walk on this network. The netwok is similar to a social network, except there nodes are people and edges exist between friends. However, just because what you are studying has a network structure, doesn't mean that it has "6 degrees of separation".

The "6 degrees of separation" idea is a subset of all networks that are known as small-world networks. Small-world networks are defined as networks on $n$ nodes where the longest shortest path (i.e. the fewest edges you need to follow to get from any arbitrary node to any other arbitrary node) is of length about $\log n$. Many networks, say a regular grid, are not small-world (in the case of the regular grid, the length of the longest shortest path is proportional to $\sqrt{n}$ which is much bigger than $\log n$). Thus, the question in the body of you post is not necessarily related to the title.

However, if we take a highly clustered small-world network and perform a random walk on it, then the results will be similar to what we see in free recall experiments. In particular, the random walk will tend to wander for a while in one cluster until taking a jump to another cluster where it would again linger. Thus, the relational structure of ideas might be consistent with small-world networks.

-