Social Process Models

We open this step with one main observation – that social processes are quite hard to predict and are thus difficult to model. Sometimes changes happen where we wouldn’t expect them to, or with a surprising speed or scale. In other cases there is no change even though we would like it to happen, like in case of some ecology-related choices.

All the stories, examples and exercises in the People, Networks and Neighbours: Understanding Social Dynamics course are designed to help you:

• experience that social processes are complex and there is more to them than meets the eye;
• explore how computational models can help us deal with these complications.

Social Processes are Complicated

So, let’s start with the first one. In fact, social processes are quite complicated to predict, for a number of reasons:

• large number of people are involved
• people are complicated themselves and differ from each other
• people interact with each other and influence one another
• and it all changes in time

When you start thinking about social processes in a way that embraces all these statements and puts them into focus of attention, it means that you’re studying social dynamics.

The Study of Social Processes

Social dynamics is a study of social processes that links micro-behaviours with the final result on a social level. In order to understand such processes, we have to dig deeper and instead of thinking about a mass of people, we need to think about the mechanism behind the process. So, the first major point from all the examples and stories within the course is that social processes are complicated and it’s not only the people that count but also what happens between them.

There are some major points from deciphering social processes that relate not only to our protest-organisation example, but also to social processes in general:

• it’s not only characteristics of individuals that count (e.g. average threshold level), but also some specific features of the group (e.g. the size of the early goers group) and structure (e.g. shape of a network, spatial arrangements of a neighbourhood);
• social processes are rarely linear, we can expect “jumps” and flat lines and all different shapes of the lines;
• sometimes a small change can lead to a huge change in the final result (Butterfly Effect);
• sometimes an orderly pattern appears because of uncoordinated actions of individuals (it emerges).

Computational Social Process Models

Because of this complicated nature of social phenomena, we need a tool that helps us decipher them and this is where computational models come in handy. As Joshua Epstein, one of the most famous experts within the field, said:

Grow it and you’ll know it.

When we are modelling, we need to make precise choices and pick the most important elements of a certain process and then figure out how they work together. By trying out different examples (whether pen-and-paper or with the help of a digital tool) we can simulate how the process would unfold.

Simulations can help us uncover the basic properties of a process and give us insights about what we need to pay attention to. For example, in case of the protest-story, they helped us see the importance of initiators and early-goers group, the role of blocking clusters in the network and that we can expect big clusters in a city just like it were segregated – even though it wasn’t.

There is one more feature of computational models worth mentioning. Modelling requires precise language and forces people to translate some of their observations about humans or about their environment to a set of simple numbers and rules. This is extremely helpful when we want to deal with some multidisciplinary challenges and we need experts from different fields to work together. For example, when dealing with energy transitions we’d probably need some engineers and sociologists, and psychologists, and economists in order to fully understand how we can change energy use in a certain area. While creating a model, all those experts have to translate their concepts into one language.

Social Process Models: Simple to Sophisticated

Models and simulations can have different purposes and be less or more detailed. On the one end of the scale there are very simple ones that illuminate the main mechanism of a process (just like a simple threshold model from Week 1). On the other side of the spectrum, they can be very specific and sophisticated and help analyse a specific process in a specific environment (e.g. evacuation strategies for a certain area or traffic models for a city). Within the broad world of Computational Social Sciences a whole range of different strategies and applications of models and simulations can be found and we encourage you to explore it further. This approach is really helpful in tackling the difficult consequences of our fascinating social nature.