Look at the lowercase letter a in your handwriting, then compare it with the same letter in a book. There is a good chance they do not look alike. The version most people write by hand resembles a simple circle with a stem on the right, while many printed fonts use a much more complicated form with a bowl below and another structure above it. The lowercase g is even stranger. The version most of us write is relatively simple, yet many books and websites use a form with two loops, an ear and a connecting stroke that few people would ever naturally draw with a pen.
These are not decorative oddities or mistakes. Both are established forms of the same letters. Typographers generally describe them as single-storey and double-storey forms. Monotype’s typography glossary defines a single-storey a or g as having one counter, compared with the two-counter construction of their double-storey counterparts. For the letter g, you may also encounter the terms opentail and looptail.
Once you notice these differences, they become one of the easiest ways to distinguish typefaces. A geometric sans serif can completely change character when its double-storey a is replaced by a simple circular form. A traditional serif can feel more calligraphic when its complicated printed g is replaced with an open handwritten form. These two small letters reveal a surprising amount about how a typeface was designed.
Why Do We Have Two Versions of the Same Letter?
Writing and printing did not develop as one perfectly consistent system. Letterforms changed as handwriting styles, calligraphy, movable type and later mechanical and digital typesetting evolved. As a result, the Latin alphabet contains several characters whose printed forms do not perfectly match the versions many people learn to write by hand. The lowercase a and g are particularly obvious examples because their alternate forms have survived side by side rather than one version completely replacing the other.
The single-storey a has a relatively simple construction and is strongly associated with handwriting and many geometric typefaces. Its double-storey counterpart has a more complex structure and is extremely common in text fonts. The lowercase g follows a similar pattern. The open-tail g resembles the form commonly used in handwriting, while the loop-tail version appears frequently in traditional upright typefaces. A useful overview of these variations appears in CreativePro’s guide to two-storey type.
There are no absolute rules about which type category must use which form. Serif fonts can use single-storey letters, sans serifs can use double-storey letters, and italic styles sometimes switch forms even when their upright counterpart does not. This flexibility is part of what makes the letters interesting. Browse FontLark’s serif fonts and sans serif fonts and pay particular attention to the lowercase a and g. Two fonts that initially appear similar can suddenly feel very different once you compare those characters.
Designers also use these forms to control personality. A single-storey a often feels simpler because its basic structure is closer to a circle. That can complement geometric fonts built from apparently simple shapes. Double-storey forms introduce more internal detail and can make a typeface feel more closely connected to conventional reading typography. In handwritten and informal designs, simpler forms can help preserve the impression that the letters could plausibly have been made with a pen. You can see this especially clearly across handwritten fonts, where strict printed structures would sometimes conflict with the overall style.
The Strange Case of the Lowercase g
The lowercase g presents one of the most interesting examples of the difference between recognising a letter and actually knowing its shape. Researchers from Johns Hopkins University investigated people’s knowledge of the double-storey, or loop-tail, lowercase g in a study published in the Journal of Experimental Psychology: Human Perception and Performance. Their findings were surprisingly dramatic. Many participants were unaware that printed lowercase g commonly appears in two distinct forms, and almost none could accurately draw the loop-tail version from memory. You can read the study abstract through PubMed.
In another part of the experiment, participants searched through text containing multiple examples of the loop-tail g. Immediately afterward they were asked to reproduce the form they had just been looking at. Half instead drew the open-tail version normally associated with handwriting, and only one participant accurately reproduced the loop-tail g. The researchers also found that participants had difficulty identifying the correct g when it was placed among incorrect versions with structural details moved or reversed.
That result seems strange until you consider how reading works. Fluent readers do not normally inspect every stroke of every letter. We become extremely efficient at recognising patterns. A person can therefore identify a printed g thousands of times without developing a detailed mental blueprint that would allow them to draw it. The letter is familiar enough to read instantly but unfamiliar enough that, when asked to construct one, many people suddenly realise they do not actually know where all its parts go.
This helps explain something important about font design. Letters do not need to perfectly resemble the forms we produce in handwriting in order to be readable. Readers learn to recognise multiple visual representations of the same underlying character. At the same time, recognisability still matters. A display typeface can push the shapes of letters much further than a font intended for long paragraphs because the reading conditions are different. This is one reason display fonts often contain more unusual interpretations of familiar characters.
Some Fonts Let You Choose Which a or g You Want
Modern fonts do not always force designers to accept only one version. An OpenType font can contain multiple glyphs representing alternative shapes for the same character. The default a might be double-storey, for example, while a single-storey version is included as an alternate. Applications that support the relevant OpenType features can then allow the designer to switch between them without changing fonts.
This functionality is implemented through OpenType glyph substitution. Microsoft’s official GSUB specification explains that a font can contain functionally equivalent but visually different glyph alternatives. OpenType also defines a feature called salt, or Stylistic Alternates, specifically for alternative glyph designs used for aesthetic reasons. Microsoft’s registered OpenType feature documentation even uses an alternative lowercase g as an example of how the feature can work.
Fonts can additionally contain numbered stylistic sets such as ss01, ss02 and ss03. Instead of changing a single character independently, a stylistic set can replace several glyphs so they share a consistent design idea. One real-world example comes from the Antenna family. Type Network’s documentation of Antenna’s stylistic sets shows separate options that switch its a and g to single-storey forms.
That can make a remarkable difference to branding. Imagine a company name containing three lowercase a characters. Changing only that one glyph from double-storey to single-storey could make the entire wordmark appear simpler and more geometric without changing the typeface itself. The reverse can also happen. Introducing a double-storey form into an otherwise minimal design can add detail and make particular letters more distinctive.
These alternatives are also why looking at a font preview image does not always reveal everything the typeface contains. The visible alphabet normally shows the default glyphs. A well-developed OpenType font may hide alternate letters, ligatures, different numeral styles, swashes and other options that only appear when the relevant typography controls are enabled. When evaluating a font for branding, editorial work or packaging, it is worth checking its complete glyph set rather than judging it solely from the first preview.
Which Form Is Better?
There is no universally correct answer. Double-storey letters provide distinctive shapes that readers encounter constantly in conventional printed text, while single-storey forms can feel cleaner, more geometric or more closely connected to handwriting. Context matters more than simply counting how many levels a letter has. The rest of the alphabet, spacing, x-height, stroke weight and intended reading size all contribute to whether the typeface works successfully.
Single-storey a and g forms are sometimes called “infant characters” because they have been used in typography aimed at young readers, partly because they resemble handwritten forms children may already know. That does not mean research has established a universal rule that simpler forms are automatically easier for children to read. A recent discussion of children’s typography by Type.Today, drawing on research from the University of Reading and the history of children’s typesetting, notes that the assumption is considerably more complicated than it first appears.
For designers, the more useful question is whether the form suits the project. A geometric technology logo might benefit from a circular single-storey a. A newspaper or publishing project might prefer the greater differentiation of traditional double-storey forms. A handwritten identity may look strange if one character suddenly becomes much more mechanically constructed than everything surrounding it. There is no need to treat one version as more correct than the other.
The interesting part is that most people spend their entire lives switching effortlessly between these forms without thinking about them. We write one g and read another. We recognise a double-storey a in a book and then immediately use a single-storey version when writing a note. The underlying letter remains stable even while its visual construction changes substantially.
That is one of the reasons type design offers so much variation despite working with the same alphabet. A designer does not need to invent a completely new character to change the personality of a font. Sometimes changing the way a single lowercase a is constructed can alter the rhythm of an entire typeface.
Next time you compare fonts, ignore the headline specimen for a moment and type something containing plenty of lowercase a and g characters. Look at which versions the designer chose, how much space appears inside them and whether the font offers alternatives. Those two seemingly ordinary letters can tell you far more about a typeface than you might expect.
Explore fonts on FontLark or compare our serif, sans serif, display and handwritten font collections and see how many different versions of a and g you can find.
Research Sources
- Wong et al.: The Devil’s in the g-Tails
Journal of Experimental Psychology research investigating readers’ recognition, awareness and ability to reproduce the loop-tail lowercase g. - Monotype: Typography Terms and Definitions
Typography reference defining single-storey forms and other common terminology used in type anatomy. - Microsoft OpenType Specification: GSUB
Primary technical documentation explaining how OpenType fonts substitute alternate glyph forms. - Microsoft OpenType Registered Features: Stylistic Alternates
Official documentation for stylistic alternates and stylistic sets, including alternative glyph forms such as different lowercase g designs. - Type Network: Antenna Stylistic Sets
A practical example of a professional font family offering optional single-storey a and g forms through OpenType stylistic sets. - CreativePro: Two-Story Type
A visual explanation of the single-storey and double-storey forms of lowercase a and g across different typefaces. - Type.Today: Questions About Children’s Book Typography
Discussion of single-storey letterforms in children’s typography and research surrounding assumptions about their readability.