A research project I’ve been working on with Huiwen Lian, Michal Kosinski, and David Stillwell, has been accepted for publication at the Social Psychological and Personality Science journal in October 2016. I then uploaded a pre-print of the paper to my website and several scholar network websites (e.g., Researchgate).
To my surprise, in December I started getting email notices from Researchgate about hundreds of reads for this preprint as well as emails asking me for the paper, and a quick check revealed that Rolf Degen spotted this preprint and tweeted about it. That tweet eventually received over 28000 views and kicked off a viral spread of this article resulting in hundreds of news mentions all over the world (see partial list here, I eventually stopped tracking it). In mid January, the article finally went live on the journal website (open access). Over the course of a month+ and about 4 waves of this article going viral I received many requests for interviews or clarifications, most of which I was unable to respond to as this was already a very hectic time for me with other projects.
This post is meant to try and clarify some points about the article I’ve been frequently asked about. Every headline, video, and/or forum discussions about the article resulted in very mixed feelings. On the one hand, it was nice to get noticed. On the other hand, I was very surprised and sometimes worried by some of the coverage and discussions about our article. This experience made me realize just how great the divide is between what we mean to report in scientific articles and what is reported by the media and understood by the public. Headlines interpreting our findings as “swearing makes you more honest!” and advice columns suggesting profanity as means to increase honesty or appearance of honesty were especially troubling. Perhaps in a later post I’ll write some of my thoughts about that.
For now, I’ll first share a series of replies to frequently asked questions that I received from the media and interested readers. I will update this list as I receive more, and will probably continue to edit these as to avoid any possible misinterpretations of our findings.
We set out to examine two conflicting perspectives on the relationship between profanity and dishonesty. On the one hand, swearing is considered a norm-violating behavior, and therefore those who curse might be more inclined to also exhibit other norm-violating behavior and deceive. On the other hand, profanity is often used to express one’s genuine feelings, and could therefore mean that people who curse care more about genuinity and authenticity than they do about what others think of them, and will therefore be likely to lie less. Across three studies, we found a consistent positive relationship between profanity and honesty; profanity was associated with less lying and deception for individuals, and higher integrity at the society level.
What is the use of this study?
Psychology scholars aim to understand human behavior, and both profanity and dishonesty are generally behaviors that are not well understood. Here, the idea for this study started from reading a passionate online debate on debate.org (http://www.debate.org/opinions/are-people-who-swear-more-honest) and we thought that the 50%-50% split deserves an in-depth academic discussion and an empirical investigation. Interestingly, the public opinion debate was also reflected in an academic discussion regarding the link between profanity and honesty, with competing views arguing it can be both positive and negative. Our study takes the first step to offer some insights on this relationship.
We generally found that people use profanity in their everyday life interactions as a tool for authenticity and genuine expression of self, and for the most part – not to hurt or abuse others. We found that those who use profanity, despite the prevalent social norms against it, simply prioritize authenticity and ‘telling it like it is’ more than they do the need to adhere to social norms for linguistic use. So, if we were to try and frame that as practical conclusions, it is that we can perhaps better understand why some of our friends and family use profanity, and that often it is not that they wish us harm but rather that this is a form of self-expression and their way of communicating to us what they really feel.
Does honesty mean unethical behavior? Are people who curse also less likely to cheat, steal, etc.?
In this study we examined dishonesty as a generalized personal inclination to obscure the truth in natural, everyday life situations, mostly represented by “white lies” or “social lies” that people tell themselves or others in order to appear more desirable or positive. These are also the most common lies, and are often harmless and easily rationalized. For example, in the first study we examined lying on a lie-scale which scholars often term as a “social-desirability” scale, and in the second study we examined linguistic markers that indicate language use likely to be deceptive.
We did not measure blunt unethical or immoral behavior, and therefore cautioned in the article that the findings should not be interpreted to mean that the more a person uses profanity the less likely he or she will engage in more serious unethical or immoral behaviors. Therefore, honesty in our article is mainly about the authentic genuine expression of the self in interactions with others.
Does honesty lead to profanity? Does swearing cause honesty?
The three studies are correlational, not experimental. Scholars often say that “correlation does not imply causation”, meaning that we cannot draw conclusions on whether honesty is driving profanity or profanity is driving more honest interactions, and it is possible that both are simply happening simultaneously and are related to a third factor. However, theoretically, it may seem more plausible that caring about authenticity drives less use of filters in social interactions and a higher use of profanity as an expression of a genuine self.
How strong is the relationship? Does this mean that everyone who curses is more authentic and genuine?
In social psychology the effect found (around 0.2) is typical, between weak to medium, and overall only explains about 4% of the overall behavior. Therefore, we can say that there’s a general tendency for people who curse to be less dishonest in their social interactions with others, but this is far from certainty and there are many other individual differences and contextual factors that can affect honest behavior.
Should we encourage cursing as means of increasing honesty?
I strongly advise against it. We found that profanity and honesty often go together, meaning that it is likely that co-occur, but we could not conclude a causal relationship and therefore there is no way of telling what impact encouraging profanity would have.
Profanity and honesty are both very complex social behaviors and any change in these two behaviors can lead to either positive or negative consequences. For example, the profanity rates we observed in normal behavior on Facebook was typically between 0.25% to 0.5%, meaning less than 5 curse words in every 1000 words. It would seem that there is a certain level of tolerance in society for the use of profanity, but that higher rates of profanity may be unacceptable, and may even seem abusive.
Also, using honesty as a justification for higher overall profanity use could prove counterproductive as it is likely decrease the potency of profanity in the genuine expression of expression in unique situations of surprise, anger, and frustration. Therefore, it is very unlikely that swearing more will make you more honest, as it is very unlikely that you will be perceived as more honest if you now start swearing more.
Does your use of the term dishonesty refer to unethical behavior? Are people who curse also less likely to cheat, steal, etc.?
In this study we examined dishonesty as a generalized personal inclination to obscure the truth in natural, everyday life situations, mostly represented by “white lies” or “social lies” that people tell themselves or others in order to appear more desirable or positive. These are also the most common lies, and are often harmless and easily rationalized. For example, in the first study we examined lying on a lie-scale which scholars often term as a “social-desirability” scale, and in the second study we examined linguistic markers that indicate language use likely to be deceptive. We did not measure blunt unethical or immoral behavior, and therefore cautioned in the article that the findings should not be interpreted to mean that the more a person uses profanity the less likely he or she will engage in more serious unethical or immoral behaviors. Therefore, honesty in our article is mainly about the authentic genuine expression of the self in interactions with others.
I was wondering if you could tell me more about the profanity measured in this study?
[Related questions: How did you figure out which words were considered profane? Were there varying degrees of profanity? If so, how did the varying degrees of profanity correlate to honesty? If someone used a very taboo swear word rather than a mild swear word, is that person more likely to be more or less honest?]
We had two studies looking at individual level profanity.
In the first study we asked participants an open question about their favorite and most used profanity without indicating a specific number, and then two raters simply coded the number of curses as a measure of natural tendency to use profanity. We also included another measure where we asked participants to self-report their everyday life profanity. All three measures (count of liked profanity words, count of most used profanity words, and self-report profanity) had a significant negative relationship (correlation) with lying on a lie scale.
In the second study we used real everyday life interactions on Facebook by coding status updates. This was done automatically using a widely used software in academic for linguistic analysis called Linguistic Inquiry and Word Count (LIWC). As the name suggests, it simply count the number of swear words and calculate profanity rate, meaning the proportion of swear words out of all words. The average profanity rate we observed was 0.37%, meaning 3.7 words in every 1000 words on the Facebook status updates.
We also used this software for dishonesty by measuring certain linguistic markers which have been shown to detect deception better than chance. This is a published paper by Newman and colleagues called “Lying Words: Predicting Deception from Linguistic Styles” in Personality and Social Psychology Bulletin 2003. This is not a perfect measure of dishonesty, it only means that for people who curse we observed lower likelihood of using linguistic markers indicative of deception.
What are the next steps, if any, that are needed to further examine the relationship between honesty and profanity usage?
The three studies were only a first step in the study of the relationship between profanity and honesty, as one article alone does not provide sufficient evidence for a definite conclusion on any topic. Both profanity and honesty are hard to study, because these are behaviors that people either try to hide or behaviors that are hard to observe in a well-controlled lab setting. We need more studies finding innovative ways to further examine this relationship.
A next step, would be to examine whether the relationship between profanity and honesty in social interactions found in this study would also show when examining the relationship between profanity and blunt unethical or immoral behavior. There is a big different between authenticity and ethicality/morality, and so it is a very interesting research question to try and extend these findings to morality more broadly.
Is swearing not a sign of weakness/fear/impotence?
Swearing is a complex behavior and it still not very well understood. We tried to gain a better understanding regarding one of the paradoxes of profanity, about the relationship between profanity and honesty in people’s interactions with others. On the one hand, swearing is considered a norm-violating behavior, and therefore those who curse might be more inclined to also exhibit other norm-violating behavior and deceive others. On the other hand, profanity is often used to express one’s genuine feelings, and could therefore mean that people who curse care more about genuinity and authenticity than they do about what others think of them, and might therefore be less likely to lie.
There are other paradoxes, and swearing can serve many purposes. Depending on the person or the context, it can both be a sign of high linguistic complexity, creativity, and even intelligence, or – of lower linguistic capacity and taking linguistic shortcuts in making an argument. It can be used for abuse and threatening others, and it can be used to entertain and make others laugh. I think that any of us can think of instances that highlight the paradoxes of profanity. So, we cannot summarize profanity as simply being about one thing, but we can try and understand why people are using profanity and how they use it in different contexts.
There are a lot of different kinds of cursing and swearing. Which ones are good or bad?
Categorizing behavior into good and bad is tricky, and in psychology research we often find that things we thought were very bad actually have some good in them, and things that we thought were all good have a dark side.
Generally, mild profanity as an expression of authenticity and as genuine expressions of emotions in specific contexts where it will not be perceived as inappropriate or offensive may contribute to better interactions and even more intimacy and understanding. Abusive profanity aimed to hurt and offend others or using profanity uncontrollably or inappropriately can generally be categorized as bad, as it violates social norms, disrupts relationships, and tends to backfire.
Most of the studies were done in the US. Are the conclusions universal?
The second study using Facebook was done using an international sample of English speaking Facebook users, and includes non-US countries like the UK. We also ran first study in Hong Kong using Cantonese-English bilinguals and found similar results. So, there is some evidence that this is not simply about the US.
However, the three studies are only a first step in the study of the relationship between profanity and honesty, as one article alone does not provide sufficient evidence for a definite conclusion on any topic. Both profanity and honesty are hard to study, because these are behaviors that people either try to hide or behaviors that are hard to observe in a well-controlled lab setting. We need more studies finding innovative ways to further examine this relationship.
We also need more studies in international settings, and using languages other than English. It is very plausible that profanity is used differently in different cultures, different contexts, and different languages.
Should I try coaxing myself to swear more and recommend swearing to others?
One study, on any topic, is insufficient for drawing strong conclusions, as we warned in the article at the general discussion. Another big issue, which we also warned about, is that correlational data should not be inferred as a causal relationship, and observational data should not be taken to mean that interventions are needed or would work. One last major point, there is very little in the social world that that is all good, and things that seem all positive often come in a package of good and bad, so even if academia starts off by pointing out some advantages, down the road you can expect that scholars would also explore the disadvantages. Therefore, before forcing yourself and recommending to others to do anything, you need an accumulation of comprehensive findings over time with replications, and both correlational and experimental data. This study is just a first step to encourage more studies in this complex area of inquiry.
You support profanity abuse / your work is bolstering the self-esteem of abusers and creating a dangerous environment for their victims
I’m sympathetic to people who have suffered from abuse, and I do understand that profanity is sometimes used as a tool in abuse, but this is not what this article is about and we have repeatedly and explicitly discussed and warned against any such interpretations of these findings. We were very careful not to use any judgment regarding the issue of profanity, nor to take any positive or negative stance on this issue. Abuse can occur through profanity, but does not require profanity, nor does the use of profanity always indicate abuse. We do not study abuse situations, but rather everyday life interactions (on Facebook or in the lab), and in the paper we discuss the two conflicting views on profanity. From our samples, people self-report to using profanity not for abuse, but as a self-expression of their true feelings, which is only to say that there are other uses to profanity that do not involve abuse. We cautioned against interpretations of the findings as if we encourage the use of profanity, and warned that we do not mean that suggest it as a tool. We also warned that the study is correlational and involves many limitations and so much more research is needed before any strong conclusions are drawn. Our preliminary findings merely point out that profanity can be viewed as a way for people to express their true selves.
Catchy title, but poorly executed research, there are all kinds of problems with your methods
Skepticism and criticism of scientific research is important, welcomed, and encouraged, We tried to openly address the limitations of our research in the article’s limitations section, the longer limitations section in the supplementary materials of the article, this FAQ, and a follow-up article currently in-press. We explained in the article and repeatedly emphasized that this study is merely a first step and that “Future studies could build on our findings to further study the profanity–honesty relationship using experimental methods to establish causality and incorporating real-life behavioral measures with a wider range of dishonest conduct including unethical behavior.”
What is important here is for others, like yourself, to conduct follow-up studies in the most rigorous way you can and share your findings with the world. The best way forward is for more researchers to pick this up and conduct further studies. These are complex constructs that are hard to study, and so I would be excited to see more innovative studies examining this research question. I would be happy to give my thoughts and review your work, should you decide to study this further.
Other questions? please comment below and I’ll add this to the FAQ.