A few years back I wrote a psychology thesis on the effects aggressive music has on emotion, which outlined how aggressive lyrics can potentially lead to aggressive behaviour in people. In a study of 110 participants, I found statistical evidence that aggressive words do lead to negative affect with reports of aggressive emotions. These aggressive emotions can then prime people into subsequent aggressive behavior. This was found from just words without the accompaniment of music. So, of course, this got me thinking about how we digest words on a daily basis; from reading street signs, timetables, using websites and apps, to being barraged by sensationalist news.

Do we create a better user experience if our choice of words does not accurately evoke the behaviour we want from our target user? In short, no we don’t.


Take for instance the number of times we find websites or apps where we want to do something, but get confused to where the ‘call to action’ button lives. There is a myriad of material out there regarding this subject, but I would like to share with you, my take on the topic:

If we can elicit emotion and behaviour from just words, then it stands to reason that a ‘call to action’ that requires an actionable behaviour needs behaviour-heavy meaning in the word/s on the call to action button itself. Icons can work, but in the context of understanding the power of words, let’s focus on the user experience of language. For instance, if you wanted to ask a user what they wanted to do i.e. perhaps it is a website that requires them to download, go to a page or open a document, what trigger words should you use?

Taken from an article in UX Booth, a great discussion was outlined on how a ‘call to action’ must start with a verb in order to push actionable behavior from users.

“Buttons are for actions, like “Get a quote,” “Download,” “Open an account,” “Go to checkout.” The text on the button should begin with a verb. Otherwise, it’s not a call-to-action, just a button with some text on it. “More information” for example, is not a call-to-action.” – David Hamill

This simple yet logical approach to the user experience of language is mirrored in the now famous statement by UX Author, Steve Krug.

Don’t make me think.

Steve Krug

Steve Krug: Don’t make me think book cover

The mechanisms behind correct wording goes much further than this, for example, we are aware that certain words and phrases can evoke powerful mental images. When researching language for my psychology thesis, I found that graphic words like ‘slaughter’ and ‘murder’ evoke strong mental images in people’s minds. Although they are similar words in meaning, they evoke different mental images. Another extreme example is the word ‘assassinate’ as opposed to the word ‘slay’. They can be used to explain the same thing, but again evoke different mental images which are often an automatic response based on people’s knowledge or cognitive schemas. All we have to do is read the news headlines to see how effective these are in getting people to click.

So what does this all mean when designing a user experience with words?

Well, it must be said that our choice of words should be carefully picked for our audience, which is a no brainer, but often overlooked. Take for instance language that is appropriate in meaning, imagery and discourse. Like if you were to use the word ‘elegant’ to describe a product. This can evoke a very different mental image and meaning than using words like ‘neat’ or ‘nice’. However, the word ‘nice’ might be appropriate for a different target audience. Imagine a user demographic that does not use the term ‘elegant’ in their vocabulary, they might appreciate ‘nice’ as it is more recognisable, perhaps less alienating or more comfortable. Or say you wanted to grab a user demographic that appreciates or looks for class; in this instance, ‘elegant’ might be a better choice of word. This type of language analysis involves language discourse (ie. the conversation style). Again, when designing what words and phrases to use, this is especially true when understanding the actual audience you are communicating with. Legal or political discourse uses a very different communication style than social or playful discourse.

Furthermore, if we are mindful of who the user is we can also take a lesson in some general psychology. Take for example navigation menus that often use less than standard words and phrase like ‘Get in Touch’ or ‘Let’s Talk’ when in reality ‘Contact’ or ‘Contact Us’ are more widely accepted or recognised terms. Obviously, branding also dictates this and is an important consideration when making these decisions (i.e. when formal is used for corporate brands and conversational for fun brands). However, tone of voice or discourse style needs to be balanced with recognition.

Recognition is a conditioned (mostly learned) response people have when consciously or subconsciously navigating websites. So again, ‘Don’t make me think’ applies here as well, as choosing the most recognised words and phrases for the correct audience is the quickest way to avoid confusion. This is strongly reflected in Jakob Nielsen’s heuristics – most notably second on the list ‘Match between system and the real world’.


Making sure the choice of language is correct can be confirmed quickly with user testing; asking the user every step of the way if the terminology makes sense. This approach also supports assumptions with real data. If 80% of people prefer the functionality of certain wording, over 20% who don’t or are indifferent, then the majority rules. If it is more divided (like 60/40 or 50/50), then further testing of different terminology is required, until the user feels they no longer need to hit the ‘translate’ button for the language to be intuitive.

In conclusion, this topic is a big one and I have only touched on a few points here. However, it is always worth noting that our choice of words will have a serious impact on the functionality of our designs. Whether you are designing apps, websites or other interfaces. The call to action or general content must adhere to the language that guides users efficiently. This also needs a scientific, human-centred approach to be correct. That way testing our claims about functionality can be validated and we are never just designing on flimsy assumptions.