If you’re familiar with the internet, you know there’s a problem with the word “actually”. After initially gaining recognition in 2012 as “the worst word on the planet”, it quickly rose to an unpopularity stratospheric enough to justify think pieces in The New Republic and The Atlantic. Its ill repute transcends the English language: last year, in an annual contest in the Netherlands, it placed in the Top 10 most irritating words.
Researchers at the Max Planck Center for Psycholinguistics in Nijmegen were intrigued, so they subjected the word “actually” to Max Planckian levels of scientific scrutiny it can’t possibly deserve. However, the results of their recently published study provide surprising new insights into why the word drives us so crazy. You might even say the research goes some way to rehabilitating actually’s busted reputation.
Geertje van Bergen and Hans Rutger Bosker were trying to understand what being #actually’d does to a person. For a long time linguists thought “actually” was like any of the other small throwaway words known as discourse markers – meaningless interjections like “of course”, “you know”, “well”, “I mean”, and “indeed”.
Except anecdotally we all know that an “actually” makes us feel different than any of those others. Here’s how Roxane Gay, an English professor at Purdue University, explained it to the New Republic:
“When people use the word actually in many contexts, they are implying that they have exclusive access to a font of incontrovertible knowledge. When they actually you, they are offering you a gift.”
So how do you quantify this feeling? That’s to say, how do you scientifically measure the difference in effect on a listener between an “actually” and a more anodyne discourse marker? To find out, van Bergen and Bosker designed an eye-tracking study to compare how much of a cognitive delay the word “actually” introduced, compared to another similar word.
For comparison they chose the word “indeed” – inderdaad in Dutch (as compared to “actually”/eigenlijk). Next they designed an impossibly rigorous series of tests around them to understand what happens in real time as a person experiences a conversation partner unloading an “actually” on them. Previous research had indicated that “actually” is a marker of realness, indicating to a listener that the speaker is about to introduce an unexpected element into the conversation, signalling that this was something to look out for. “Indeed” on the other hand, was thought to be a far less complicated signal, alerting the listener to relax; no additional cognitive demands were about to be made by the forthcoming sentence.
In one of the experiments, participants listened to 48 simulated conversations in which a crucial word was redacted with a *beep*. The big clue to what the word was going to be was contained in the speaker’s use of the word “indeed” or “actually”.
All the conversations were structured something like this:
Narrator: Despite her fear of animals, Mary went to the circus.
Mary’s friend: “You must have been terrified by the animal act?”
Mary: “I was INDEED/ACTUALLY scared by the running *BEEP* at the end.”
What’s under the beep? As they listened, the participants were also shown a screen with images pertinent to the conversation. After hearing the snippet, they were instructed to click the one that most likely fit with the use of the discourse marker (indeed should go with the lions, where actually should go with the clowns.)
As it turned out, participants reacted very differently when a speaker used “indeed” and “actually”. Long before they clicked an image, the indeed or actually was telling their brain what’s up: the eye tracking software showed exactly when during the sentence their eyes drifted. Right when Mary said indeed, eyes drifted to the correct answer, before the sentence had even finished. With actually, they dithered a bit more.
Actually takes a fair bit of cognitive work. A speaker needs to think about what their addressee is expecting from the conversation. Then they use “actually” to acknowledge that there is a problem, to signal that what’s coming next is a realignment of what they have perceive as the two mismatched sets of expectations. The listener needs to understand all of that, and then use it to reframe everything they hear next.
But even taking that into account, the researchers still found it still took a surprisingly long time for people’s eyes to settle on what should have been the “correct” option – sometimes they ended up not even clicking the correct picture!
To understand why, consider this exchange:
Narrator: Before her job interview, Theresa had a pesto sandwich.
Theresa: “Do I have something in my teeth?”
Theresa’s friend: “Actually, there is something right there in your *BEEP*.”
Participants were pretty stuck on this one, and here’s why. If everyone agreed on how the word “actually” is supposed to function – that is, as a straightforward signal that a person is about to lay something unexpected on you – participants would have seen the counterfactual “actually” and known that the sentence was supposed to be, “Actually, there is something right there in your hair.”
But not everyone does agree on how the word actually is supposed to function. In recent years we have developed an off-label use for the word – as a social softener. You’ve probably done this, and I know I have. “Actually, I ordered the chicken,” you might say with a cringe face to signal an apology for being difficult. “Actually, that colour does make you look hideous,” you might say with the same face when trying to be honest without hurting your friend’s feelings. Actually has actually become a way to pull our punches.
Which is why participants in the study couldn’t figure out if Theresa’s friend was telling her that actually, she did have something, but it was in her hair, not her teeth – or if she was saying, actually, I’m sorry to tell you this, but there is a colossal leaf of unmasticated basil flapping next to your left incisor, and I really hope you haven’t had that job interview yet.
But even this social softening interpretation has been diluted, because now a lot of people perceive “actually” – however the user is deploying it – as a deliberate weapon of passive aggression. “Actually, you’ve been misusing the word actually throughout this entire article.”
van Bergen says this evolution of “actually” shouldn’t come as a surprise. “Languages are dynamic systems that change all the time,” she says. “Words and constructions constantly acquire new meanings and uses over time, while losing other meanings and uses; this is a natural process.”
However, she does reckon that this particular polyfunctionality contributes generously to actually‘s bad reputation, especially when a speaker uses it in one way and a listener understands it in another. “My guess is that such mismatches underlie its bad reputation,” she says.
Being online makes it all worse – as it always does. “When eigenlijk/actually is used online, it is typically directed at multiple readers rather than to a single listener,” she says. “I can imagine that the chances of misunderstanding increase because not every reader has the same beliefs and expectations.” (Not recommended: telling someone to stop being offended by your use of the word actually because “actually, you misinterpreted my use of the word actually.”)
Is it kind of impressive for one little filler word to be doing triple duty in our fucked up ambiguous internet communication vortex? Or does this just make it more annoying? Incredibly, we will need even more scientific analysis of the word “actually” to find out.
Photo credits:
Actually sphinx: Originally a political cartoon starring Benjamin D’Israeli, at the Wellcome Collection. Sorry, everyone.
Lion: a lioness hunting worthogs in the Serengeti, by Schuyler Shepherd
Clown: A clown participating in a Memorial Day parade, by Rick Dikeman
Hair: Medium blond hair, by Xight
Teeth: The teeth of a model, by David Shankbone
