Welcome to a scientific resource on fonts.
Do you spend WAY too long looking for fonts? Do you have trouble finding fonts that “feel right” for your context?
Well, this article is for you.
I read 75+ academic articles on typefaces. And I compiled the actionable findings into this article. By the end, you’ll know how to decipher the “personality” of any font, so that you can choose the best font in any context.
In this part, you’ll learn the step-by-step cognitive process. You’ll learn why people associate personality traits with fonts (and how to identify those traits).
In this part, you’ll learn specific findings from research. You’ll learn which font traits (e.g., serif vs. sans-serif) are more more effective in certain contexts.
Let’s play a game.
Among the fonts above, which is better for:
Like most people, you probably chose C, A, then B.
They felt right? Seemed fitting? Looked good?
Sure. But WHY did those fonts feel right? Most people can’t articulate the reason because the mechanism occurs subconsciously.
So here’s the answer…
If you follow my content, you’ve heard me explain your brain’s associative network.
If you’re a newb here, then watch my quirky video about spreading activation:
Your associative network plays a role in font perception. How? I summarized the steps in the following model:
Let’s look at each step…
You see a font…and that’s it. Pretty straightforward. If you want a deeper understanding, Koch (2011) explains the biological components of font perception (see pgs 17-27).
Fonts contain various components (e.g., line, weight, size, orientation). When you see a font, your brain disentangles those perceptual components:
Big whoop, right?
Well…yeah. It IS a big deal. To appreciate the importance, you need to understand a crucial concept.
Look at the traits from the previous image:
Notice something? Those traits are general adjectives. They describe stimuli outside of the font world. And that’s crucial.
Fonts share visual characteristics from the real world. If you want to choose an appropriate font, then choose a font that visually resembles your context:
Here’s an example.
Kang and Choi (2013) created ads for a cell phone. When ads emphasized the “slim” nature of the phone, condensed typefaces performed better:
However, some ads referenced the elegant nature of the phone. In those cases, the opposing font performed better:
In both cases, the font matched the visual qualities of the product.
However, font traits can also be metaphorical. Suppose the ad emphasized the phone’s speed. Even though technological speed is intangible, we associate visual traits with general speed — such as a forward tilt:
“What artistic conventions are used to convey the motion of animate and inanimate items in still images, such as drawings and photographs? One graphic convention involves depicting items leaning forward into their movement, with greater leaning conveying greater speed.” (Walker, 2015, pp. 111)
Need to convey technological speed? Then tilt your font forward. Maybe add a slight blur. Incorporate traits that are visually related to speed.
It might sound far-fetched, but Lewis and Walker (1989) found that people identify “fast” related words more easily in slanted fonts.
Conversely, direct associations refer to the aggregate combination of font traits — usually the font family.
“…[direct] associations refer to the influence of historical precedence on affective response to typography. The typeface Fraktur has many associations with Nazi Germany, and Helvetica is commonly associated with the U.S. government since it is used by the IRS on tax forms.” (Shaikh, 2007, pp. 21)
Those fonts acquire meaning through your semantic network. Whenever you see a font (e.g., Fraktur), you associate meaning — based on the context. That includes semantic meaning (e.g., Nazi Germany) and emotional meaning (e.g., disgust).
Whenever you encounter that font in a future context, you modify the original connections in your network:
That’s how fonts acquire meaning (see Shaikh, 2007). It’s a never-ending process that we’ve been performing our entire lives.
When you encounter a font, you activate the perceptual and direct associations in your network. In turn, the activation spreads to related nodes.
Suppose you see the logo for Avon — a women’s beauty company:
You’ll activate the node for that overall typeface and logo. Thanks to spreading activation, you’ll activate direct associations:
In addition, you’ll disaggregate the perceptual font traits. Since the typeface is tall and thin, for example, you’ll activate those nodes in your network.
But wait…tall and thin? Aren’t those traits usually associated with beauty? Aha! Yes they are. When you see those perceptual traits, you’ll trigger an extra wave of activation toward the node for beauty.
Good job, Avon.
However, don’t jump the gun. At this point, you might be thinking: Well, if their node for beauty is activated, then people will perceive the font — or product — to be more beautiful.
It’s tempting. And I used to believe that explanation. When I published my article on advertising, I said:
“Activation spreads toward your node for beauty. That activation gives you a new temporary lens. With your concept of beauty more prevalent, you perceive stimuli in the immediate environment to be more beautiful.”
But that’s wrong. My bad. I deserve a punch in the face.
Font evaluation is more nuanced. In a few steps, you’ll see why that explanation is wrong.
In the previous step, the font activated related nodes in your network. At this point, you combine those activated concepts into a collective meaning for the font.
The collective meaning is a combination of semantic concepts (e.g., beauty) and emotional feelings (e.g., pleasantness). Because of the concoction of meaning, you often can’t articulate it. The font just “feels right” or it doesn’t…which is the next step.
Let’s revisit the explanation from my advertising article.
I explained that we associate beauty with tall and thin traits. Because of that connection, tall and thin fonts activate the concept of beauty (which will cause you to perceive stimuli to be more beautiful).
So…why is that explanation wrong? Well, consider the font, Fraktur — which was used for Nazi propaganda:
With the previous explanation, you should NEVER use Fraktur. Since people associate it with Nazi Germany, spreading activation would trigger negative emotions. And those negative emotions would transfer to the immediate stimuli.
But it doesn’t work that way. When you evaluate fonts, you consider the appropriateness of the font (see Doyle & Bottomley, 2004).
Once we generate the collective meaning, we compare that meaning to the context:
That’s why you can use Fraktur in certain contexts (e.g., documentaries). It doesn’t matter if people associate negative emotions with it. Those negative emotions won’t tarnish their evaluation, as long as the font is appropriate for the context.
It sounds like common sense. And it is. But I needed to verbalize that step because the underlying mechanism is important.
So…what’s the mechanism? Why does appropriateness lead to a positive or negative evaluation? That’s our final step…
We evaluate fonts —positively or negatively — through processing fluency.
When you process stimuli quickly and easily, it feels good. And you misattribute those positive emotions to the stimulus. And that’s the answer.
Suppose that you see the logo for Avon. That exposure will activate beauty-related concepts in your network (due to the perceptual and direct associations).
You’ll then compare that collective meaning to the context. Aha! Here, the context is beauty products. It’s congruent with the activated nodes. Because the concept of beauty is already activated, you’ll experience higher fluency:
In turn, that higher fluency will trigger positive emotions that you’ll misattribute to the context. The font will “feel right.”
Now, in that example, you processed the font and THEN the context. But the mechanism also works in reverse.
If you’re watching a documentary on World War II, Nazi-related concepts are activated in your network. If you THEN see the font Fraktur, you’ll process the font more easily because of the overlapping connections. That ease will make the font feel right.
And that’s it. That’s how we evaluate fonts. Still with me? Good.
Now that you understand the cognitive process, let’s apply it…
Fonts differ in their perceptual traits. For example:
Serifs are slight projections at the end of typefaces.
Designers love debating serifs vs. sans-serif fonts. Based on the research, though, it seems counterproductive. Other traits play a bigger role.
Nonetheless, here are some findings:
Well…supposedly. Some designers claim that serifs help guide the eye flow:
“Roman typefaces are more legible because the theory states that serifs assist in the horizontal flow of reading and eye movements.” (De Lange et al., 1993, pp. 246)
But I’m not convinced. We don’t read text in smooth patterns. Our eyes jump across lines of text— in movements called saccades (see Becker & Fuchs, 1969).
Since the movement is jerky, the “eye flow” claim seems like a myth.
This finding seems more plausible.
Computer screens display information through a pixelated grid. Due to that box-like structure, a serif may be less identifiable.
With today’s technology, it shouldn’t be an issue. But it might play a role if you’re working with very small fonts.
Generally, serif fonts are more effective for formal or scientific contexts.
In one study, people evaluated scientific text more favorably when the font contained serifs (Kaspar et al., 2015). Other research suggests that serif fonts seem more elegant and beautiful (Tantillo, Lorenzo-Aiss, & Mathisen, 1995).
Conversely, people perceive sans-serif fonts to be more informal and innovative (Tantillo, Lorenzo-Aiss, & Mathisen, 1995).
I explained the reason in this section.
Luckiesh and Moss (1940) researched the optimal weight for readability. They displayed “Memphis” in different weights, and they found that medium weights were most readable:
Researchers argue that bold fonts convey an extreme connotation:
“Bold can be made to mean ‘daring’, ‘assertive’, or ‘solid’ and ‘substantial’, for instance, and its opposite can be made to mean ‘timid’, or ‘insubstantial’. But the values may also be reversed. Boldness may have a more negative meaning. It may be made to mean ‘domineering’, ‘overbearing’.” (Van Leeuwen, 2006, pp. 148)
Lieven et al. (2015) also explain a connection between heavy typefaces and masculinity. Since people associate a bulky stature with men, bold fonts match that visual trait.
Bar and Neta (2006) published a paper called, “Humans Prefer Visual Curved Objects.” They found that…well…humans prefer visual curved objects. Angular shapes trigger an evolutionary threat:
“…sharp transitions in contour might convey a sense of threat, and therefore trigger a negative bias…” (Bar & Neta, 2006, pp. 645)
Now, if context were irrelevant, then people would prefer rounded fonts over angular fonts. However, context DOES matter. As I explained, appropriateness is important.
That said, rounded fonts are more effective for domains related to:
Conversely, angular fonts perform better for:
If your message is direct and straightforward, use a rigid typeface without any ornaments (Li, 2009). The simplistic font will match the simplistic nature of the context — thus increasing fluency.
Processing fluency conveys abundance and familiarity. Usually, those traits are beneficial. But not always.
With unique products, familiarity is detrimental. You want some disfluency because it’ll reinforce the distinctiveness of your product.
In one study, people perceived more value in a gourmet cheese when the font was difficult to read (Pocheptsova, Labroo, & Dhar, 2010). The complex font became a signal for uniqueness — which was congruent and appropriate with the context.
Earlier, I explained why slanted fonts convey movement and speed (see Walker, 2015).
Likewise, straight fonts — with their rigid structure — convey stability and durability.
Oosterhout (2013) examined branding and perceptual characteristics. She found that lowercase letters are particularly effective for “caregiver” brands that promote compassion and altruism. She also found that thin lowercase letters are congruent with “creator” brands that emphasize innovation.
Oosterhout (2013) also found that uppercase letters are effective for “hero” brands that convey qualities related to energy, courageousness, and focus:
“BWM, Diesel, Duracell, Nike and Sony are also using capitals in their word marks, to express their power and strength.” (pp. 39)
Garvey, Pietucha, and Meeker (1997) studied legibility in road signs. They found that mixed case letters are most readable.
They proposed two reasons:
When letters are separated, fonts convey the perception of individuality:
“[connectivity] has its own metaphoric potential. External disconnection can suggest ‘atomisation’, or ‘fragmentation’, and external connection ‘wholeness’, or ‘integration.” (Van Leeuwen, 2006, pp. 149)
Likewise, connected letters portray the perception of closeness
Connected fonts aren’t necessarily cursive. The letters simply need to touch.
Earlier, I mentioned that a condensed font performed better for an ad promoting a “slim” cell phone (Kang & Choi, 2013). That’s because of the perceptual overlap:
“Maximally condensed typefaces make maximal use of limited space. They are precise, economical, packing the page with content. Wide typefaces, by contrast, spread themselves around, using space as if it is in unlimited supply.” (pp. 148)
“Wide typefaces may also be seen in a positive light, as providing room to breathe, room to move, while condensed typefaces may, by contrast, be seen as cramped, overcrowded, restrictive of movement.” (Kang & Choi, 2013, pp. 148)
Font height resembles our metaphorical conception of gravity:
“The meaning potential of horizontality and verticality is ultimately based on our experience of gravity, and of walking upright. Horizontal orientation, for instance, could suggest ‘heaviness’, ‘solidity’, but also ‘inertia’, ‘self-satisfaction’” (Kang & Choi, 2013, pp. 149)
If you need to portray something as durable or immovable, a short font might work well.
Likewise, tall fonts convey lightness and quickness. The meaning can also extend to metaphorical associations of aspiration and ambition (Kang & Choi, 2013).
Other research has linked verticality with luxury (Van Rompay et al. 2012). If you need to portray a luxurious product, a tall font might be a good choice.
I threw a ton of info at you. But you can use this article as a reference guide. Feel free to download the PDF so that you can reference it moving forward.
Or if you need to choose a font quickly, then use the table below (which I adapted from Henderson, Giese, & Cote, 2004). Just look for a group of traits that describe your context. Then choose a font with similar visual characteristics.
Want some more content? My other articles expand on the cognitive aspects of design and linguistics. You might enjoy: