I am working on a project to try and come up with (either by finding it or creating it) a system for categorizing human relationships in a purely descriptive sense.
Previous attempts to classify or categorize human relationships have mostly focused on the personality of the individual, rather than the objective, observable details that make one relationship different from another. (See Bowlby 1969 & 1973, Ainsworth et al. 1978, Bretherton & Waters 1985, Hazan & Shaver 1987). However, some have attempted to address these more descriptive differences. (See Kelley et al. 1983, Hendrick & Hendrick 1986) But those focused solely on romantic or sexual relationships.
One of the best attempts on the subject that I have found is Kayser, Schwinger, & Cohen (1984) which examined loving relationships, friendships, and work relationships separately.
Foa and Foa’s “Resource Theory” developed in the early 1990’s has a lot of promise, and I may start there. (The theory is that we fundamentally exchange 6 types of resources: love, status, service, information, goods, money. These types of resources vary by the degree of their concreteness/abstractness and by the degree of their unversalisticness/particularlisticness. If a type of resource is universalistic, it means that it maintains the same value regardless who the giver is. Likewise, it is considered as particularlistic if its value depends on who gave the resource.)
Many of these relationship models fail to consider a broader view of relationships. While the discussion of the exchange of status and service in the context of a romantic relationship can be easily tweaked to address D/s dynamics, but the question of an open vs closed relationship is somewhat more difficult to address, especially when love is considered a resource/commodotiy, and therefore is finite or zero-sum like goods or money.
Previous research aside, I am bouncing back and forth between how an categorization system should be approached. Should it be taxonomic (like how we organize species of lifeforms) or should it be more like computer specs, with each feature being independent of each other one?
I would be curious to know if any readers have ideas? Either for places to go for research or an idea for how to approach the problem? I’ll probably be working on this for a while.
|X|C|BDSM| 
I think your categorization depends on the objective of the research.
Are you trying to help people with the dynamics of their relationships?
Are you trying to increase awareness of non-traditional models?
Is it purely ivory-tower academic? (doesn’t seem like you.)
Or something else.
I highly recommend starting with the end in mind, and then the categorization and approach will fall neatly into place.
Great idea, btw.
There are a few purposes.
I came up with this when trying to make an illustrative graphic of our large and growing poly network to use in our “intro to poly” classes. At first, we were trying to color code the lines connecting people, but we quickly realized that relationship types were too diverse (especially in alternative communities) to be represented by just a few different colors. I considered using a lot more colors, but was then struggling to succinctly (2-5 words) label each color. I found that I didn’t have the language (and seemingly no one else does, either) to easily describe the different relationships.
As I was thinking about this problem, I expanded the idea to cover all possible relationships between any two people.
I can imagine a lot of applications for this, once it is developed. One is to solve my problem of being able to map out my current relationship (as well as all of my partners’ relationships) in a way that is understandable and clear. I think that, in a broader sense, having a shared language would aid communication between partners, help them to define where they are or where they would like to be. It could help in a psych setting, where therapists would have their patients list out the significant people in their life and then work out what the designation for that person is in a unified, standard way.
Also, once there is a standardized categorization system, it would allow us to do some population surveys and measure how many people engage in different relationship types. That information would be very useful.
And I’m sure there is much more I haven’t thought of, yet.