For many of us, swearing is part of our everyday existence. The ability to let off your dissatisfaction at another person or inanimate object in a linguistically flamboyant fashion is one of the smaller pleasures in life. Yet little is understood about the factors that influence an individual\’s inclination to swear.

However, the opening up of social media data provides researchers with an opportunity to gain new insight into these previously hard-to-measure patterns in social behaviour. During March 2012 I collected all geo-located tweets for the London area and using this data I set about identifying simply how swearing patterns change over time and space.

Let\’s first look at how swearing on Twitter varies over the course of a day. I won\’t publish here the list of swear words I used during this analysis, but be assured that I was quite comprehensive in covering all possible English-language configurations.

The graph appears to suggest a growing tendency to swear during the latter part of the day, culminating at a peak at around 10pm. But isn\’t this meant to be our most happy time of the day, when we\’re at home and relaxing? Well, as I\’ve speculated elsewhere, perhaps what this shows us is that swearing actually occurs more when we are relaxed, when the work persona is taken off and we are chatting with friends or shouting at the TV.

Click here to see the full size map

Spatial patterns in swearing propensity across London indicate some alternative trends at play. Looking at the map, central London would appear relatively well-mannered in comparison to the outer and eastern boroughs. Yet the influences driving these trends are less readily identifiable. I have laid out some hypotheses on my blog – mainly I suspect these trends are influenced by a mixture of working patterns, as well as demographic and local cultural variation. However, I would welcome and encourage additional perspectives.