Fonts and Authority Perception (Psychology)

thomasv1's picture

I'm doing my Bachelor Thesis in English Language and Culture on the perception of typefaces. I intend to establish the existence and extent of the phenomenon that authority is transferred upon the reader through reading the font. So, by having the participants rate texts on authority, I intend to establish that serifs are more authoritative than sans serifs.

So, my question is, have you perhaps come across past research dealing with the subject of fonts and perception? Currently, I'm looking for any info I can get. Any websites, papers, journals... I'd love to hear from you.



Don McCahill's picture

There have been some similar studies in the Psych journals. Sorry, but I can't remember the references. None get too deeply into type design (of the ones I have seen, they are often very basic character recognition studies.)

There is a good article on Bouma on the Microsoft site at

You might also want to look at the writings of Aries Arditi, many of which are available at

blank's picture

So, by having the participants rate texts on authority, I intend to establish that serifs are more authoritative than sans serifs.

Maybe I’m just unfamiliar with the grammar of psych academia, but shouldn’t you intend to determine whether or not a typeface with serifs can enhance the authoritative tone of text? And isn’t authoritativeness a product of the nature of the text? Serifs may be able to enhance the credibility of an authoritative piece of writing in the context of a society where figures perceived as authorities tend to use serifs, but serifs are not able to, for example, impart authority to a comedy that is intended to have none.

Chris Dean's picture

@ Thomas: Sounds like an interesting study. This could make a great thread. Can you tell us a little more about yourself and your experimental design?

1. How are you operationalizing the term "authority"
2. How are you asking participants (Ss) to rate authority?
3. Your independent variables (IV's) are typeface (serif and sans). Have you selected any? If so, which ones and why?
4. What text are you having Ss read?
5. What school do you go to?
6. Who is supervising your study?
7. Do you have any background in cognitive/experimental psychology?

As far as research goes, you need to get access to science journals and databases. PsychINFO and Web of Science are probably going to be of most use. If your school does not have access to them, I would strongly ask your supervisor/department head to find someone who does and put you in touch with them. Conducting a lit search without access to such databases &c is asking you to operate with a significant disadvantage. It is a very important skill to develop in the field of experimental psychology.

Chris Dean's picture

In addition, how many Ss do you plan to run?

Nick Shinn's picture

This is a bad idea.

Typefaces are not perceived, only documents.

It is the designer/typographer's task to match form with content, to create an authoritative document.

paragraph's picture

I intend to establish that serifs are more authoritative than sans serifs

You have arrived at your conclusion before you started :)

Chris Dean's picture

Paragraph is correct. You should begin with a hypothesis.

CameronWilliams's picture

To James Puckett, Christopher Dean, Nick Shinn and paragraph: thank you.

To thomasv1: what sorts of serifs? Oldstyle, transitional, modern; cupped, bracketed or slab? What will be your methodology for making this decision? Will this study be double-blind?

thomasv1's picture

My main approach will be to present the participants with four names (in different fonts) of lawyer firms, museums, or dictionaries, and have them rate them on authority. Of course, I will rotate the fonts used, in order to eliminate the possibility that one firm/museum/dictionary will have a better-sounding name.

@Don McCahill: Thanks for the links!

@James Puckett: You are correct. My phrasing of the problem is problematic, and because I am very much in the early stages, I will refine it later on.

@Christopher Dean: (1) I haven't found any literature on that, so anything would be more than welcome.
(2) I'll have the participants 'assign' authority through rating the four firms/museums/dictionaries one to four, with one being most authoritative.
(3) The typefaces will have to be representative of a real-world situation, so I will probably make a selection between Adobe Jensen Pro, a Libertine, Adobe Caslon, Georgia, Times New Roman, a Garamond, etc. Any suggestions, maybe?
(4) As said, business logos. I will probably add a question asking the participants to rate a small paragraph of text, as well.
(5) I go to St. Radboud's University in Nijmegen, the Netherlands (
(6) M.J. van der Haagen.
(7) No background. This is my first venture into the field of cognitive studies, which is why I'm looking to gather any information I possibly can. I have done a similar study on the effectiveness of adjectives in film descriptions.

@Nick Shinn: Well, that's what I'm trying to find out, really; whether your statement is correct, or whether there's more to it than meets the eye.

@CameronWilliams: The scope of the survey will not allow me to get to much into different categories, although I surely intend to highlight some fonts. Of course, the study will be double-blind.

@Everyone: Thanks for the feedback.

Nick Shinn's picture

My main approach will be to present the participants with four names (in different fonts) of lawyer firms, museums, or dictionaries, and have them rate them on authority.

So, there's no more to a corporate identity than picking a typeface?
That's bad news for the graphic design profession!

dtw's picture

What about the difference in authority between the same font well- and poorly-spaced? Or in all caps vs title case? Or roman vs italic? Or colour? Or small caps... or... or...

Sorry :-D
Ever since I chose to block pop-ups, my toaster's stopped working.

will powers's picture




thomasv1's picture

@Nick Shinn: You're absolutely right in pointing that out. Having only the name means I'm conducting the survey in an unrealistic environment (i.e. without the complete logo and the rest of the design choices), but my focus is on providing graphic designers with info on typographic decisions only. A survey always has a margin of error, because it simply cannot encompass the entirety of reality. Maybe I could replicate a logo or somehow attain a more realistic representation of a lawyer's office logo. Any help would be more than welcome.

@dtw: Accounting for the scope of my thesis, I will be limited to using regular title case only.

Pablo R's picture

If you reefer to the expression of traditional lawyer firms, museums, and dictionaries corporate image, It probably has more to do with communicating a conservative, solemn, formal, serious or traditional personality than with authority.

For me authority would have more to do with power or respect. Like the army, the police or a stop sign.

The message is not just in the type but it is one of the ingredients of the expression.

eliason's picture

"Operationalization" really is a word, God help us. :-)

CameronWilliams's picture

Perhaps Joseph Goebbels, arguably the greatest ad man of the twentieth century, did research on which Nazi propaganda posters were most effective; those using sans, fraktur, or serif faces. An outrageous and even offensive statement and suggestion; I know, but there may be some useful information there.

dtw makes a very good point.

Pablo R also makes a very good point, especially in regard to stop signs. I'm sure various Departments of Transportation have information on which types used for road and warning signs have been most effective. Adrian Frutiger has also written extensively on this subject.

Alessandro Segalini's picture

Reminds me of my grandma saying speak like you eat.
Yes, Thomas, we don't read fonts, we read words—patterns of blacks and whites.
No piquing anyhow, knowing that you have such interest.

Nick Shinn's picture focus is on providing graphic designers with info on typographic decisions only.

Readers can't tell you that.
As Dave notes, the "authority" of a typeface will change with the way it's set.
Depending on what a company symbol and page layout looks like, the same typeface may look dignified or silly, progressive or conservative.

Symbolic forms such as letters and type have no absolute meaning, but only have meaning in relation to other symbols.
For the reader, these other symbols are the rest of the document at hand.
However, for the designer, the realm of typographic significance is more abstract, and encompasses our knowledge of type culture: it is this meta-language we work with to create meaning for readers.

thomasv1's picture

@CameronWilliams: Thanks for your pointers.

@Nick Shinn+Alessandro Segalini: I appreciate your vision. Your concerns of the survey not taking place in a realistic setting is entirely valid. I will discuss this further with my supervisor and reach a compromise between the scope of my thesis and its relevance as a completed paper.

paragraph's picture

So, serif is more authoritative than sans?

guifa's picture

Paragraph, it seems like serif might be authoritative and sans authoritarian :)

paragraph's picture


1985's picture

I looked up operationalizing also.


I think the conditions of the experiment are much much more complex than you are anticipating.

"The timeline is between love and fear… "

1985's picture

What am I doing wrong with the quote marks?

CameronWilliams's picture

If you’re on a Mac, type option-left bracket for an open quote, option-shift-left bracket for a close quote; on a PC, I believe you just substitute the alt key for the option key. But this sites’ "SmartyPants" feature will let you type ASCII quotes and substitute the real ones reliably.

eliason's picture

(Except if you have a space before your intended end-quote.)

paragraph's picture

Or if 'you' prefer single ones.

nora g's picture

To open your mind the following book might be useful: Das Gesetz und seine visuellen Folgen/La loi et ses conséquences visuelles (German and french) edited by Ruedi Baur at Lars Müller Publishers/Baden.

And the link to Denis Pelli – Professor of Psychology and Neural Science:
He did a lot research on the topic of letter identification ... perhaps you can find something of interest here too.

Ed_Aranda's picture

@paragraph: Interesting demonstration. You hardly ever see serif fonts being used for government agencies, or law enforcement for that matter. But is the typeface itself communicating that authority, or are we conditioned to respond to that font in such a way because, through repeated exposure, we now associate it with the qualities and values represented by the Federal Bureau of Investigation? In other words, are those qualities imparted on the typeface by the context in which it is used, or is there something innate about that font that lends itself to communicate authority?

I would venture to guess that government agencies choose sans serif fonts not because they communicate authority, but because many of them tend to be colder and less personal than many serif fonts. However, I may be way off.

It's going to be very difficult to generalize the findings of this study across all applications. If you will be using law firms, for example, I think that your conclusions, however significant, will only be applicable to law firms.

CameronWilliams's picture

@Ed_Aranda: “I would venture to guess that government agencies choose sans serif fonts not because they communicate authority, but because many of them tend to be colder and less personal than many serif fonts. However, I may be way off.”

I was once told by a non-designer boss that I had to use Helvetica, because it was the “industry standard”. Who knows where he got that idea; needless to say, I didn’t work for him for very long. The choice may often be due to simple laziness, too.

Thomas Levine's picture

Thomas, before you can come up with a precise and useful experiment, you'll have to do a decent amount of background research, and this is what you asked for here. Nobody's really giving you many references, and this thread is becoming more brainstorming. I think this is partly because few rigorous studies in the area exist.

In order to study this, you'll probably want to review a bunch of existing documents and see how authoritative they seem to you and analyze their typography in general. You'll probably want to try to find old accounts of type too.

I've had the same question as you but about small caps in particular. People seem to love them at my school, and I think it's because they look fancy and official to these people. (Long ago, I used to be one of them.) I read a lot about the history of small caps, but I'm stuck at figuring out exactly why this attitude came about. I think I'd have to find old letters to explain this.

Del's picture

Check out political campaigns and the thought that goes into their design.
Their should be some good studies on what typefaces were used
for certain candidates and why.

blank's picture

Check out political campaigns and the thought that goes into their design.

Don’t. Politicians are notoriously inept when it comes to designs, and will gladly pick type and colors with no thinking and reject any thoughtful design that comes their way. I know political consultants who go through this over and over and it drives them nuts.

chimerical's picture

thomasv1: I intend to establish that serifs are more authoritative than sans serifs
paragraph: You have arrived at your conclusion before you started :)
Christopher Dean: Paragraph is correct. You should begin with a hypothesis.

I agree with Paragraph. What happens if your research suggests the opposite - that sans-serifs are actually more authoritative than serifs? Do we defend serifs anyway? The only way in that case would be to purposely only seek and present data that supports the serifs argument. I hadn't noticed that a lot of us tend to fall into this trap of making up our minds before we've begun (including myself) until a professor pointed it out. Interesting thesis by the way!

Chris Dean's picture

Observer expectancy effect. This is why it's risky to start with a bias. Solution = have a research assistant who does not know what your hypothesis is run Ss.

In order to conduct research like this you will probably need support from a supervisor with a strong background in experimental psychology to help you, and the equivalent of a second year psychology course (usually called Psych 2000), some stats and a few labs. If you plan on being published in a peer-review journal as opposed to a design magazine &c you may need to go through ethics approval (nightmare) as you're using human Ss.

@ Thomas: I encourage you to explore this as I believe scientific research of this nature is an important part of the future of our professions.

@ chimerical: What happens if your research suggests the opposite?. You raise a good point. A null hypothesis. This is where the math starts to get tricky.

Chris Dean's picture

Correction. Null = no effect. I misread chimerical.

Pablo R's picture

First we would have to define what kind of authority we refer to.

The actual message is not in the sum of elements but in the way they interact.


blank's picture

First we would have to define what kind of authority we refer to.

This is why I suggested that the experiments explore how typeface choice affects the credibility of a message perceived to have authority rather than assume that a typeface imparts authority directly. To further your example, a stop sign is both authoritative and a warning that one obeys for the sake of self-preservation. Using a wispy serif, or even comic sans, probably has no more impact on the credibility of the message than a heavy sans, because most people still don’t desire to blow through an intersection and get sideswiped by a truck. But I’m not sure that this would be the case with highway signs restricting carpool lanes to a car with three occupants—in that case drivers are already resentful of the authority behind the sign and its policies, and a break with the sans norm might discredit the message and lead to it being ignored. And of course this is all culturally relative!

Pablo R's picture

Exactly James, not very sure about the carpool.

Nick Shinn's picture

I'm thinking some kind of a grunge script...

Chris Rugen's picture

Both the FBI and Stop sign examples were what I thought of first as well. Type as voice is important, but as Nick points out: context is everything.

I think that the comparison of serif vs sans would work best with words that are out of context or are perhaps not even words. Would "glakbp" be more authoritative in Times or Arial? But then we run into the other big problem: which serif and which sans? Perhaps a family with both, such as Meta and Meta Serif or Scala and Scala Sans (to name two), would help to minimize the problem? Unfortunately, serif type and sans serif type both come in 1,000s of variations. So, really, your study would be dealing with one serif style vs. one sans serif style. I don't think you can make a definitive distinction with such broad terms.

I suspect, as I think many others here do, that this study is subject to so many fluid cultural factors that the data would only really be valuable as a continuum of samples taken regularly over an extended period of time and in multiple geographic locations. Many of us here are designers and we all have robust toolkits for manipulating perception that could probably subvert the outcomes of the study pretty easily. In eliminating the variables designers use to inject meaning and provide context to content, you would be removing the study so far from the reality of type as it's used, that I'm not sure if the data would yield information of much value beyond an extremely narrow case.

I'm happy to hear that you're interested in examining type's effects, but I think a further refinement of the study's parameters is in order. I recommend that you educate yourself about type, then pick the aspects you want to study. I believe serif vs. sans is simultaneously too granular and too broad to allow for conclusive and meaningful results, because type performs such a unique role in the world.

blank's picture

It occurred to me that Kai Bernau’s thesis about his Neutral typeface is somewhat relevant. Check out the PDF and consider ordering a copy of the book.

Chris Dean's picture

@ Puckett: Nice link

To help get the ball rolling…

paragraph's picture

Nick is right, context is everything:

David Rault's picture

Funny, I always thought that generally, authority is conveyed by capitals. The 'serif' or 'sans' choice influences my perception of who this authority might be.


Don McCahill's picture

The stop and similar graphics are a little beyond what the OP is aiming at. Sure bold will be more authoritative than hairline. But when other typographic features are similar, does the serif lead to a more authoritative impression from viewers.

I suggest that matching faces could be used. For instance Lubalin Graph and Avant Garde, where serifs are the key differences. And many other families have been made with serif and sans that otherwise complement.

No wonder there is such a dearth of true scientific study done on type ... typographers keep rolling the question into opinion.

Chris Dean's picture

And we still haven't operationalized our primary DV. Until we do, this is conversation is mostly intellectual (and entertaining).

@ Thomas: Check out scientific method.

Nick Shinn's picture

I recall c.1980 working on various accounts.
Some were industrial business-to-business advertisers, and for those one used Helvetica, Futura, Univers, Franklin Gothic...
On the other hand, for the National Ballet of Canada, classic serifed faces such as Bodoni, Garamond, Goudy, and the then new Pontifex.
Now, the corporate face of the National Ballet is Helvetica 55.

Also back then, typeface choice was considered to be an important element of brand differentiation, so at the typical small-town intersection there would be a bank on each corner, each with a different corporate face: Century Oldstyle (National Trust), Palatino (CIBC), Plantin (Scotiabank), etc. Now all the banks here apparently use Frutiger/Myriad.

So, as a cultural study (rather than an ill-conceived quiz posing as meaningful research), it appears that this is a complex issue involving time and place.

Why is it that 30 years ago art directors considered a variety of serifed faces suitable for bank advertising, and now a homogenous sans?
Is type choice a decision of clients, creative directors (with an eye on status/awards within the creative metaculture), or an accurate reflection of what consumers best respond to?

All of the above, in a dynamically evolving system with multiple feedback loops, an element of chance and perhaps some technological determinism.

This proposed research may be useful as a snapshot, to compare with future snapshots, but IMO a cultural study should study culture, not masquerade as science, because the only way behavioral science is of any use to creative design is in product development.

Chris Dean's picture

@ Shinn: You raise some good points. Especially cultural. I think it may be possible to take this into account by controlling the population of participants (age, language, ethnicity, vision &c) but this would lower this studies ecological validity. A trade-off in experimental research should you wish your results to be reproducible.

thomasv1's picture

Thanks for engaging in the discussion and in doing so, providing me with things to consider.

I've decided on defining authority/status as a construct made up of believability (not credible/credible, not convincing/convincing, etc.), expertise (unskilled/skilled, unknowledgeable/knowledgeable, etc.), and trustworthiness (unreliable/reliable, untrustworthy/trustworthy, etc.). Participants will be questioned on their perception of the author of the text, one group receiving a serif version, the other a sans serif. I haven't decided on the content of the text itself, but it will be about half a page in length, as opposed to my earlier idea of doing law firms, etc.

Syndicate content Syndicate content