A study looks at the effect of subtle wording changes on the performance-boosting power of motivational self-talk.

I’m an instinctive skeptic, so the widespread claim from sports psychologists that the semi-random babble of words bouncing around in my head can influence my 10K time has always seemed... improbable, to put it politely. But a series of studies has changed my mind, and over the past few years I’ve written and spoken about self-talk frequently—which has sparked a common question that I’ve been totally unprepared to answer: once you believe it, how do you actually do it?

Fortunately, that’s the question a recent [2019] study in the Journal of Sports Sciences starts to wrestle with. Of course, there’s lots of folk wisdom out there on how to deploy self-talk, and some of it is undoubtedly good. But the skeptic in me still wants evidence to back whatever advice I offer, and there hasn’t been much systematic testing of different self-talk approaches for endurance. So it’s neat to see this experiment, from a research team led by James Hardy of Bangor University’s Institute for the Psychology of Elite Performance, on the critical role of something I never would have considered: grammar.

Hardy and his colleagues decided to compare the effectiveness of self-talk using first-person or second-person pronouns—that is, the difference between telling yourself “I can do this!” or “You can do this!” They didn’t just pluck this idea out of thin air. Previous research, for example, has suggested that second-person self-talk enhances public speaking performance and reduces the associated stress, possibly because it enhances “self-distancing.” Stepping outside your immediate experiences and emotions, and viewing them instead from the detached perspective of a supportive onlooker, allows you to take the fear of failure less personally and to make better decisions. Scott Douglas’s excellent article on the new study, for Runner’s World, gets into some of these ideas in greater depth, but I’d like to focus here on the practical side of how to actually implement this.

First, some basic details on the experiment: to test the grammar idea in an endurance sports context, the researchers asked 16 volunteers to do a series of 10K cycling time trials. The first one was basically a practice trial, and after each kilometer the cyclists were asked to say out loud some of the internal thoughts they’d had during that kilometer. This gave each subject a baseline set of self-talk statements that was personal and relevant to them. Using a guided workbook, they then analyzed their statements and figured out ways of making them more positive and functional—from “Hang in there” to “I’m hanging in well,” for example.

Then, on separate days, they did two more 10K time trials, using their updated self-talk statements in either first-person or second-person format, in random order. The results were significantly faster in the second-person condition, averaging 17:25 compared to 17:48, a 2.4 percent difference that was statistically significant with a p-value of 0.014. It’s worth noting that without a control condition (i.e., stick to the whatever you do naturally), we don’t know if both forms of modified self-talk improved performance to different degrees, or if one made it better and the other had no effect or made it worse.

Anyway, the main point is that sticking to second-person seems to have helped. Hardy and his colleagues are fairly diffident about the implications: “We are cautiously optimistic,” they write, “that [the findings] represent an untapped branch of self-talk worthy of further consideration by researchers and practitioners.” That’s a good attitude to take. This is, after all, one small study in a field where replication of results is notoriously hard. No one should stake their career on the importance of pronouns just yet. But as an addition to the growing pile of research on self-talk for endurance, there are a few points worth digging into.

The first is individual variation. In this particular study, 13 of the 16 subjects did better with second-person rather than first-person pronouns. That’s a pretty good batting average. Are the three misses just random chance—they had a bad burrito the night before their second-person pronoun trial or something—or is there some way of predicting who would or wouldn’t respond? In a way, given how different people are and how nebulous the proposed reasons for the second-person advantage are, it would be shocking if everyone responded in the same way. The researchers note, for example, that people with more narcissistic tendencies use first-person pronouns more frequently, which might make them either more or less sensitive to the effect. That’s one obvious avenue for further research, and the broader lesson probably applies to pretty much all advice about self-talk: what works on average may not work for you.

And that brings us to the second point, which is the details of how they customized self-talk for each athlete. While this process is often described in general terms, the new paper includes a couple of specific examples. They take the phrases the athletes reported during their practice trial, then adjust or replace them so that they have five pre-planned self-statements, one for each two-kilometer section of the race.

Here’s an extract of what that looked like for one participant:

Distance: old phrase -> new phrase

0-2K: I can do it -> I (You) can do it
2-4K: I am determined -> I’m (You’re) determined
4-6K: I need to keep going -> I (You) can keep going
6-8K: No pain, no gain -> I (You) can work through the pain
8-10K: I’ve done it -> I (You) will succeed
In some cases (the first two), there’s no change at all. In other cases, the changes are pretty subtle: “I need to keep going” becomes the more optimistic “I can keep going.” The list has also been pruned down to eliminate some of the other less useful phrases this volunteer reported, like “Feeling motivated.” But overall, this particular person already had a pretty good internal monologue.

The other sample workbook features a lot more phrases that they end up culling out—stuff like “keep grinding,” “keep pushing,” and perhaps worst of all, “forget about the pain,” which, as Dostoevsky would no doubt have pointed out, is like trying not to think about a polar bear. The replacements tend to be encouraging and actionable: “I (You) can keep going.”

What’s the actual peer-reviewed, double-blinded evidence for replacing “I need to keep going” with “I can keep going”? Nothing that I’m aware of. This part of the self-talk picture, as far as I can tell, still relies mostly on experience, theory, and intuition. So take it for what it’s worth, as an example of what practitioners in the field are currently doing. If you want to try self-talk, give this approach a try, noting your current phrases and making suitable adjustments. And hopefully, over time, more studies like this one will test the various ways of tweaking self-talk, so we can zero in on the most effective approaches.

Alex Hutchinson is the author of Endure: Mind, Body, and the Curiously Elastic Limits of Human Performance, with a foreword by Malcolm Gladwell. Find him on Twitter and Facebook.