Altmetric Blog

Interactions: Expletive-Laden Tweets (for Science!)

Jean Liu, 5th June 2013

arrrghPart 1: Swearing for pain relief

[altmetric doi=”10.1097/WNR.0b013e32832e64b1″ float=”right”]

Have you ever blurted out a string of angry four-letter words in reaction to a painful injury? If so, what effect did uttering swear words have on your pain? In 2009, psychologists Richard Stephens, John Atkins, and Andrew Kingston of Keele University (Staffordshire, UK) published a quirky paper entitled “Swearing as a response to pain”. The paper, which appeared in NeuroReports, described an experiment in which 67 undergraduate students undertook the “cold pressor test” (it involves sticking one hand in cold water until the pain is unbearable) whilst repeating expletives or neutral words. Interestingly, the researchers found that the students who repeated curse words were able to leave their hands in the cold water for a longer amount of time, suggesting that pain tolerance levels were increased. The findings were later covered and replicated in Discovery Channel’s Mythbusters, and the paper even ended up winning an Ig Nobel Peace Prize in 2010.

Despite being nearly 3 years old now, the Stephens et al. paper has consistently appeared on social media, from expletive-laden tweets to Reddit posts to numerous blog posts (see Altmetric details). It’s even cropped up on news sites years later, such in as this article from Forbes. For many, the paper was an excuse crack jokes and swear profusely:

It works! After only two hours of swearing, my lumbago pain went from please-kill-me-now to a mere excruciating…

— Björn Brembs (@brembs) March 17, 2012

However, Altmetric wasn’t collecting mentions from most sources of attention back when the paper was actually published (our data, especially those from Twitter, is most accurate after mid-2011), and so I can’t accurately describe the early responses to the paper. However, it’s still evident from some blog posts (see the excellent posts at NeurophilosophyNeuroanthropology, and Neurotic Physiology) that quite a bit of online buzz was generated around the paper around its publication in 2009, and again after it won the 2010 Ig Nobel Peace Prize. Moreover, academics seemed to find the paper important, as it was saved to Mendeley 65 times.

The paper actually experienced something of a revival this week, after Max Fisher, foreign affairs blogger at the Washington Post, single-handedly boosted the popularity of the paper with the following tweet:


Part 2: Incremental advances

[altmetric doi=”10.1016/j.jpain.2011.09.004″ float=”right”]

Two years after the publication of the first paper on swearing and pain, the same research group replicated and extended their findings in another study (published in The Journal of Pain) and entitled “Swearing as a Response to Pain—Effect of Daily Swearing Frequency”. Not only did the researchers confirm their previous findings, they also found that people who generally swore more frequently in day-to-day life received less benefit during the cold pressor test, implying that “overusing” swearing dampens its pain-relieving effects.

As with the first paper, this new one was mentioned in the mainstream media (e.g., a story in The Independent) and was popular on social media, especially after being shared by high profile science communicators such as Vaughan Bell (see Altmetric details):

The more often you swear in everyday life, the less it reduces pain when you’re hurting

— Vaughan Bell (@vaughanbell) November 16, 2011

Some people weren’t as impressed with this advance in the research. Stewart B. Leavitt wrote the following scathing commentary on his site, Pain Treatment Topics:

“Our disappointment with this sort of research is that the first time around it was a rather novel, somewhat entertaining, look at an everyday occurrence — cursing at a painful mishap. Now, with this second published report on the same basic concept we are wondering if this is what passes as serious science these days in the pain management field.”

Online audiences seemed somewhat divided about whether or not the Stephens et al. papers constituted “useful” or “serious” science. Certainly, the Improbable Research blog (responsible for awarding Ig Nobel Prizes) gave the initial paper a particular non-serious label, and many social media users and bloggers echoed this sentiment in their comments. Interestingly, although the 2 papers by Stephens’ lab group were quite similar, each one apparently received quite a lot of independent attention (i.e., most people who mentioned the second study did not appear to have read the first).

Both papers received reasonably high levels of online attention, with many mentions consisting of enthusiastic reactions to the study findings. (Notably, the second paper has the highest score amongst all papers in The Journal of Pain.) It’s quite possible that those who had read the original study just had no interest in mentioning the new one. Or, perhaps this time around, the paper happened to reach different audiences. According to the Twitter demographics (which should be taken with a grain of salt, considering the fact that Altmetric lacks accurate pre-2011 Twitter data), 77% and 12% of tweeters of the first paper were members of the public and scientists, respectively. In comparison, 59% and 29% of tweeters of the second paper were members of the public and scientists, respectively.


Tracking research uptake in multiple papers of the same project

What is the ideal point at which to stop an experiment and package the results into a complete paper? In the case of the Stephens et al. papers, perhaps the findings of the second paper could have been incorporated into the first; however, the ability to replicate the first experiment would not have been shown. Logistical issues and a number of other reasons might have prevented the researchers from conducting more experimental stages for their first study. However, it is still important to be able to observe the overall impact of a research project, in addition to the impact of a single paper. Ultimately, when focusing on article-level metrics and assessing research uptake, we need to be able to track the uptake of a multi-paper project, and follow the natural progression of research as it takes one small step forward at a time.

Leave a Reply

Your email address will not be published. Required fields are marked *