Another fun experimental paradigm is to connect people sensors that measure their galvanic skin response (roughly speaking, how sweaty they get) in order to determine their stress levels. Of course, the whole point of this paradigm is then to stress the poor subjects in order to see how they respond.

A team at the University of Bristol in the UK decided to see if swearwords caused a significant amount of stress compared to two different categories of non-swearwords: euphemisms (“f-word” “c-word”), and neutral words (“glue” and “drum”).  They asked 24 volunteers to take part in a study where they read words from a screen and had to say if they were swearwords (or euphemisms for swearwords) or not.

“Euphemistic” swearwords
Neutral words
“Euphemistic” neutral words

The researchers measured the levels of stress that the volunteers experienced.

Electrodermal activity

Skin conductance - a proxy measure for stress - measured over time. From Bowers & Pleydell-Pearce (2011)

The results showed that the participants experienced the highest levels of stress when saying the swearwords aloud: their levels of stress when saying swearwords were considerably higher than when they said “f-word” or “c-word” instead. Although stress levels are a bit higher for euphemisms for swearing than for neutral words, and neutral “euphemisms”, there was no significant difference between the groups. Saying “f-word” is not significantly more stressful than saying “drum”.

One thing that stands out as interesting for me is the pattern of responses to the neutral “euphemisms” (“d-word” and “g-word”). Initially, the levels of stress rise more sharply for these terms than they do for the complete words, glue and drum. Could it be that seeing “-word” is enough to make us alert for the possibility of swearing, however euphemistically delivered?
Bowers JS, & Pleydell-Pearce CW (2011). Swearing, euphemisms, and linguistic relativity. PloS one, 6 (7) PMID: 21799832