Uncharacteristic Characters

How the virtues (and pitfalls) of type, calligraphy, and handwriting came together in the design of a new typeface: a lecture at Typographics 2017.

I recently had the privilege of speaking at the third annual Typographics conference, an event organized by the Type@Cooper program at The Cooper Union, to share a new project with fellow designers. While visually this new typeface would be unlike anything we’ve ever done before, in many ways it’s the quintessential H&Co project, taking on many of the themes that have characterized our work over the past twenty-eight years. It pokes at the perimeter of the “type family” as it’s commonly understood (much like Hoefler Text, The Proteus Project, Knockout, Numbers, and Nitro), it meditates on the relationship between formality and informality in typography, like Ideal Sans, The Historical Allsorts, Archer, and Operator; it looks for clues in the observable world about how people interpret letterforms, like all of our typefaces, from Shades to Gotham.

The conference has just posted the video of the talk, which I’m pleased to share with those who were unable to attend. And attached below are my lecture notes, which may make up for my double-time performance: already someone who speaks too quickly, I’d mistakenly planned a forty minute lecture for a twenty-five minute slot! —JH


Sixteen years ago this September, I got married at the Brooklyn Botanic Gardens. Like all designers, I designed my own wedding invitation. The invitation itself was set in type, and the outer envelopes were addressed by a calligrapher. But the thing I had the hardest time with was the map, which would show our out-of-town guests how to find the venue. Anything set in type seemed too institutional, as if I was creating the map for the Botanic Gardens itself, rather than my own wedding. But anything rendered in calligraphy seemed absurd: to me at least, calligraphy seems like the stuff of historical maps, not contemporary ones; the only contemporary maps I can think of that have a genuine claim on calligraphy are from the fantasy realm, and alas I was getting married in Prospect Park, not Mordor. So neither typography nor calligraphy had the right tone. I wanted the versatility of type — after all, I had to create an elaborate map with parking instructions, nearby subways, and so on — but I wanted it to have the warm, natural, genuine feeling of something that came from people — something that was recognizably not from an institution, but rather from the two of us.

Around that time, my friend Jesse Sheidlower called me to ask a piece of typographic advice. Jesse has one of the most fantastic jobs in the world, and if you think being a typeface designer is esoteric enough to rope you into long conversations on airplanes, Jesse has it far worse, because he does something even more exotic: he’s a lexicographer, who writes dictionary entries for a living. At the time, Jesse was the Editor at Large of the Oxford English Dictionary, specializing in North American English, which meant that he supervised a staff who pored over TV scripts, hip-hop albums, and Usenet posts to come up with things like an accurate definition for chillax, or the definitive early citation for BFD.

Jesse was writing a book called The F-Word, which would be a dictionary dedicated to — well, the F-word — and its many colorful variations. He’d called to ask my advice about a suitable typeface for the project, and a lot of this conversation was also about tone: he and his editors wanted the tone to be casually blithe — “insouciant” was the word he used — but also to be underpinned by a foundation of scholarship that readers would recognize by the typography. While the book was inherently entertaining, it was also going to be a dictionary, and needed to present as such. This conversation about tone echoed many of the open questions I had about my wedding invitations, and what sort of typography might be at once casual and serious. But it also raised the question of whether such a typeface could be up to the functional requirements of serious content: dictionaries after all are incredibly sophisticated typographic projects, that use a surprising number of different typographic dimensions to articulate different kinds of information.

Sophisticated typography helps educate people about how things work.

Thinking about Jesse’s dictionary, and our wedding map, as likely applications for such a typeface, was intriguing. Like dictionaries, maps are sophisticated objects, and one of my favorite examples of how design can use typography as a vital part of the user interface. Maps use different styles of lettering to connote different kinds of information, helping users to intuit what each data point means. Spending time with a map like this [slide], it soon becomes clear that towns and bodies of water use different styles, and the distinctions run deeper still. One lovely mapmaking tradition that I always like is the way sites from antiquity are identified using blackletter, which blends right in with its background unless you’re specifically looking for it.

I often use printed maps as an example of how sophisticated typography can educate users about how something works, without the need to refer people elsewhere for instructions. Certainly every map has a legend that resolves any questions, but it’s rarely necessary to consult it. At a certain point, simply by using a map, you just figure it out.

So, like all the most interesting projects, it seems, Jesse’s dictionary didn’t have the time to consider a custom solution. But the idea of creating a typeface that was both informal and highly functional stuck with me — I could see it being useful to have a typeface that was serious in the sense of thorough, but not serious in the sense of earnest; to have a typeface whose voice was unmistakably personal, rather than institutional.


When I started designing typefaces professionally, back in 1989, I spent a lot of time advocating for custom typography as a way to end the monoculture of early digital fonts. An analogy I used a lot was to compare the potential for typography to the abilities of scribes, who could adapt their lettering at will. I wasn’t just thinking about formal calligraphy — Renaissance writing masters, for example, who could move effortlessly from one hand to another — but also contemporary artists who use letterforms in their work. One that I always loved growing up — and for all I know, this may be the very thing that got me into typeface design — is the work of Charles Schulz. Everyone recognizes the iconic lettering of Peanuts, but one of the things that I especially loved as a kid was what would happen when Snoopy sat down at the typewriter: the lettering would shift to something typewriter-inspired — which is funny, and unexpected, and a really effective way of immediately signaling that what you’re reading is no longer dialog, but something being written.

I loved the way Schulz used different styles of lettering to gently, almost subliminally, connote different kinds of language. And on reflection, I realize that it’s something that he had in common with another artist who I grew up with, the inimitable Edward Gorey. (For those of you who don’t know Gorey’s work, you’re in for a treat. His stories are great: they’re elegant and deeply subversive, marvelous dark fantasies set in a vaguely Victorian universe full of mustached gentlemen and ladies with parasols. They generally involve people being driven insane, or hapless infants coming to some grisly fate.)

Gorey was a master lettering artist, and I’m not talking so much about his skill as a draftsman, but rather his perceptiveness about what letters can mean. The title pages for his stories always reveal a deep love for lettering, and he works these different styles of letter into the stories themselves, in ways that are always apropos, from the slab serif [in this slide] that signals “old-time baseball uniforms” to the blackletter that identifies a proverb.

Illustrators and artists instinctively use the traditions of typography to confer legitimacy on the fantasy, to speak to readers in the same ways that graphic designers do.

But I especially love the way he uses lettering not only for display, but for text. When we designers at a type conference talk about “lettering,” we’re almost always talking about large, solitary things: logos, headlines, things that are grand, and have a great opportunity for pyrotechnics. And don’t get me wrong, I love this kind of lettering, but I also love the rarer species of lettering that’s used for text. Gorey’s illustrations use all kinds of tropes and clichés to good effect, and his lettering does too. He uses the traditions of typography to confer legitimacy on the fantasy, whether it’s the bookish presentation of a caption that uses romans for text and italics for titles, or an epigram whose attribution is set in small caps, in just the way that a graphic designer might do. Gorey was obviously a keen student of typography on three different levels: what letters look like, what different kinds of letterforms connote, and how typography can be used to articulate content. Typeface designers think constantly about these three different perspectives, and how their unity can yield typefaces that are not only interesting, but also useful.

I find myself using lettering to shape content every day, and not just in my role as a designer. Rather, it’s in the proofs that I exchange with the designers at Hoefler&Co: my job is to oversee the design of typefaces at the company, and one of the ways that our team communicates is through marked up proofs.

There are times I find myself indulging in lettering that’s more enthusiastic than it really needs to be, sometimes to relieve the tedium of working through a 240-page kerning proof. But I also do what many of us do instinctively, which is to reach for natural ways to highlight different kinds of information in distinct ways. There’s not really an organized system at work here [in this slide], but clearly there’s some attempt being made to ensure that key details don’t go unnoticed.

Style is incidental to what handwriting is. Principally, informal writing is simply the record of an author’s words.

This [slide] was a note to Sara Soskolne, one of our senior typeface designers, who has always had kind things to say about my scribblings. And perhaps as with Edward Gorey, it’s less about the lettering itself (which is pretty ghastly) and more about my affection for using different typographic styles together. I suppose you could dissect a note like this to say that it’s generally set in script, with occasional highlights set in sans serif small caps; a sort of wide, decorated inline; and even a pen-drawn bold. But I think it presents to the reader differently, as simply a single statement that uses noticeable styles the way a typographer might employ bold or italic. And to me, that’s the most interesting thing about handwriting, as opposed to calligraphy: calligraphy formalizes its mannerisms, and the qualities of its tools, and refines them into a particular style. But handwriting is less conscious, and therefore has more freedom to move between different styles. Style is incidental to what handwriting is: handwriting is principally the record of an author’s words, and only incidentally possessed of style in the typographic sense. I realized that this is what I wanted all those years ago, when designing that map, and also what I wish I’d had to recommend to my friend’s dictionary: something that had the style-less-ness of handwriting, but the versatility of a modern type family.

Or to put it another way, I wanted something that borrowed the best parts of type, calligraphy, and handwriting. I wanted the versatility of type, but none of typography’s artificiality: type after all is synthetic, institutional, more anonymous than personal, which as a communication tool is generally its strength. But instead of artificiality, what I wanted was something from calligraphy: its organic quality. Calligraphy is genuinely hand-made, warm and natural — it’s the product of the wrist, the elbow, and the shoulder — but it’s too mannered for what I had in mind, since by design it coalesces into a formal style. An antidote to this formality is found in handwriting, which is as informal as letters can be — and I really love the way handwriting reveals authorship, instead of concealing it, but then handwriting can often be disorderly, if not downright sloppy. So my question was whether it might be possible to strain out the virtues of these three kinds of lettering, and eliminate their shortcomings, and that’s what we’ve been trying to do with a new project called Inkwell.


Inkwell is an attempt to create a family of fonts that feels like hand-written text: how you might letter the entire contents of a book, if you had infinite time, patience, and skill. It began with a set of drawings that I created around 2004: initially sketches on paper, and later a set of digital fonts, to serve as a “proof of concept” for the idea, and see if it was worth undertaking what I imagined could be a much larger project. In addition to wanting the design to feel personal — authorial — I wanted the design to have the versatility of a hard-working type family. This meant obvious things like a sophisticated character set, but also a set of varied, expressive styles that would have recognizable relationships for writers and designers, whatever handwriting’s equivalent of “bold” and “italic” might be.

Like handwriting, I wanted to make sure it wasn’t related to any particular typeface, or even a particular style. The last thing I wanted this to be was “hand-drawn Gotham,” but I also hoped that it would escape classification altogether, so that it didn’t feel like an Old Style, a Venetian, or a Slab, but simply felt like “neatly printed letters.” These are some of my drawings from 2005, to explore how the basic design might develop. As with handwriting, I found myself reaching for styles that aren’t technically connected with what a proper typeface would do: here after all is a seriffed text face, whose small caps happen to be sans serif. As a typeface designer, beginning this way means one of two things: either convincing everyone that it’s OK for these two morphologically dissimilar styles to share the same typeface, or doubling the work to create both a serif and a sans.

I found myself wanting to do more with the typeface than just a handwritten serif or a handwritten sans — let alone both — and noodled out a couple of blackletter characters to see how they might work. And I liked the idea, that this typeface could conceivably have both roman and gothic letterforms, that are related only because they’re made by the same hand, and with the same tools. I hoped this might go in even more directions, maybe including the kind of open capitals that my father used to use on blueprints — he was a theatrical set designer, and was always lettering the headings on shop drawings in a style like this one. It’s an engineer’s letter that you’ll find on blueprints and patent diagrams, and while he used to draw his freehand, I think they were probably an approximation of what a lettering template could produce.

If the typeface was going to have not just sans serif letters, but a more stylized degree of sans in the form of these open capitals, then perhaps the serif ought to have a fancy cousin as well, perhaps a Tuscan, one of my guilty typographic pleasures. I always really liked the capital H [in this slide]; the A was more of a placeholder for an idea that I wanted to come back to later. The thought that all these styles might work together reminded me of the way engravers — or Edward Gorey, for that matter — could create these riotous collages of letterforms that always felt unified, because they were all the product of the same tools.

Inkwell advances the idea that the same tools can help unify divergent styles of letterform.

And I think there’s a real value in being able to work with disparate styles in the same piece of design, without it feeling cacophonous. Here’s an example [in this slide] I snapped years ago: I’m certainly seduced by what this piece of typography is going for, but it’s somewhat undermined by the divergent traditions of its fonts. The Spencerian script at the top is pretty solid, and below that is an engravers’ letter from turn-of-the-century business cards; fair enough. But the third typeface is inspired by classical roman inscriptions, and the fourth one is an especially Victorian interpretation of an Anglo-German fraktur. It seems to me that this lettering might have been more attractive in the mockup, when everything was drawn with the same marker, and by the same wrist. And that’s what I hoped Inkwell might do: it might unify these different styles by rendering them with a common formal vocabulary.

I put this project away in 2006, to focus on other things. Matthew Carter once commented that he’s never regretted stepping away from a typeface, and I agree: time confers a valuable critical perspective on our work, and can exert an evolutionary force on typefaces, too, to thin out the herd.

Two of the best ways to improve a typeface are to set it aside, and to share it with another typeface designer.

But besides time, the other thing that makes every typeface better is sharing it with another designer, and that opportunity presented itself last year, in the person of Jordan Bell. Jordan’s a typeface designer who joined us three years ago, and among other things, I think it’s safe to say that he’s a compulsive, and chronic graphomaniac. Here [in this slide] are some of his more exuberant notes from the Inkwell proofs. Some of them are of notes to me, others are notes to himself, others are perhaps notes for posterity.

When Jordan started, we put him through the traditional hazing of drawing a difficult project. Jordan worked with Andy Clymer and me to create three heavier weights of Archer, of which one of the most excruciating parts was extending the fonts’ tabular figures into this new territory. Tabular figures of course need to have the same widths throughout a range of weights, which fights against the natural tendency of letters to get wider as they get bolder. But Jordan came through with flying colors, making these terrific additions to Archer.

Something that’s great about working with another designer is that it forces you to articulate your own position, which in turn can reveal inconsistencies in your own thinking, or tendencies that you hadn’t thought to examine. When Jordan and I began working together on Inkwell, one of our first conversations was about how serifs should connect, which meant dissecting my drawings to see how they worked, then abstracting from these a set of rules, and then seeing if any characters didn’t fit the mold. This was the beginning of the “Inkwell Bible,” a long and growing collection of guidelines about what Inkwell should do. (Of course, it never looked this good until I put these slides together — most of what we recorded were one-off Illustrator files, or notes in e-mail, or Slack, or Trello.)

One of the ways we tried to keep things from becoming monotonous was to try and apply these strategies differently throughout the character set, so that things like caps and small caps don’t follow the same underlying design. In many ways, Inkwell became about doing the things that you typically don’t do in a typeface, since our goal was to introduce not consistency but variation, up to the point where it becomes noticeable.

Another thing I love about collaborating with other designers is that they inject unexpected and valuable ideas into every project. Part of this is the natural tendency to grow tired of your own work and see the work of others with fresh eyes, and part of it is that every designer has different talents, instincts, and interests. I’d had this idea that even though the design was a serif, it might have sans serif small caps, because this is how I write. Of course, this in turn would have meant having a secondary set of sans serif caps, to accompany these small caps, which in turn would have meant two set of caps in the font, one with serifs for the lowercase, the other without, for the small caps. This started to feel complicated, and felt likely to either confuse people or actually thwart their expectations — in any case we weren’t sure it would be possible to engineer things so that they’d behave as people expected, so in the end we did the thing that would be better for designers, if more labor-intensive for us, which was to do two typefaces, both a serif and a sans.

Of course, since there was going to be a serif italic, we’d presumably need to do a sans serif italic as well, and this wasn’t part of the family that I’d explored. And this was another place that it was great to collaborate with another designer, because while a sans serif italic isn’t a style that I naturally write with, it’s very much one of Jordan’s favorites. So here’s the sans serif italic that Jordan designed, which is has a real flavor of signwriting that I love. This one quickly became one of my favorite parts of the family.

I was really happy that Jordan was enthusiastic about the blackletter that I’d been tinkering with, especially after the ridiculous amount of time I’d spent trying to figure out what “informal blackletter caps” might look like, most of it scrawled on the back of restaurant checks and junk mail. We spent last summer trying to find the right way of dressing down the gothic style, so that it still felt ceremonial, but not twee. I really wanted it to feel more special than the roman, but not necessarily more buttoned-up — and that really emerged as a key part of the Inkwell family, that all the styles should feel as close to the center as a traditional family’s bold or italic.

Dressing down the gothic style, so it’s ceremonial, not twee.

The line quality was a big part of this, as were the constructions we chose for the caps, which borrow pretty freely from Regency Blackletter, Kanzlei, and Engravers’ styles. For instance, since the traditional gothic N and E are a little too archaic, even indecipherable these days, we dropped them in favor of a more contemporary construction, the kind that we all know from newspaper flags.

Part of the fun of a typeface like this is figuring out what the back end of the character set might look like, and we had a lot of fun here. The figures alone are always a challenge, since they have neither a historical nor a cursive foundation. Jordan drew these, and I really love them. And either because I traumatized him with Archer, or just to flex a little bit because he could, he designed these on tabular widths. I say with both fondness and admiration that Jordan’s a sick bastard.

In the middle of all this, we started to talk about the idea of weight. I’d originally imagined Inkwell as a collection of alphabets that could all be the product of the same pen, and therefore all with a consistent stroke width. But a companion boldface is obviously a valuable thing for a designer to have. And even if the design were to stick to a single weight, it seems to me that choosing that weight is important — I wouldn’t be surprised if Inkwell brought out the kinds of red-hot feelings that designers have about writing instruments. So we decided to expand the family, from a single weight into a range of weights, bounded by a technical pen Thin and a graffiti marker Black.

As useful as these weights are in the more sober styles — and they really do introduce some nice flavor in their more cursive moments, especially — for me it’s the blackletter where things really start to cook. I mean, I’m a blackletter fan in general, but this typeface is really just having a good time.

The style we hadn’t really spent much time with was the Tuscan. My drawings were still pretty embryonic, and though I’d been sketching these incessantly for more than a decade, there were key characters that were starting to feel impossible to resolve — like the capital O. (No Tuscan typeface has a nice O.) Now that we’d added a weight axis to the project, it seemed likely that a style this intricate simply wouldn’t survive getting so much extra weight, so we shelved it. I found myself doodling all kinds of things in an Inkwell style — here’s a little monogram [see slide] that, for a time, I thought we might include in the fonts — and at some point, we both realized that this sort of construction might be a more fertile direction for the Tuscan. It delivered on the underlying premise that Tuscan alphabets have bifurcated serifs, and it seemed as if this sort of design might work in a range of weights. It also invited, at last, a solution to that impossible letter O. I did some sketches to propose the basic morphology for the character set, and drew a couple of control characters in Robofont, and Jordan took things from there, creating a really beautiful and expressive family of fonts. This one interpolates really interestingly: it goes full-on circus poster in the Black weight, but as it gets lighter, it takes on a sort of rustic, woodsy feeling, which I’ve been finding really useful.

By last summer it felt like the more Inkwell we drew, the more Inkwell we wanted to draw — a bracing change from the usual pattern, in which you get sick of a typeface three quarters of the way through. Here [in this slide] is the Open that I’d wanted to do, which we redesigned as two extreme weights that would interpolate. Jordan brought some of that signwriters’ fluidity to these letters that I really like.

Inkwell’s broader mission is to expand the very idea of the type family.

Early in the project, I’d wondered what the italic should look like: whether it should be a fully cursive script, as handwriting probably should be, or whether it should be a carefully lettered typographic italic in the Edward Gorey tradition. We could make arguments for both, so we decided to do both, wrapping up the project with a script. People have been asking us for a script for ages, and I’m really happy that our first one isn’t just a typeface, but has the broader mission of being part of an experiment in growing the type family.

The notion of the type family, to me, is really what type design is about. Some of the most significant moments in type history have been not about typefaces, but about the expansion of the type family as we understand it. Aldus’s italic isn’t just significant for being a new kind of typeface, it’s important because of what it made possible: the marriage of romans and italics that have endured for five centuries. Italics aren’t just part of typography, they’re part of language — especially English, in which it’s common to write sentences that can’t be properly understood without them. Here’s a good one [in this slide], in which the presence of italic clarifies the meaning, and the placement of the italics gives rise to nearly as many meanings as there are words in the sentence. Even The Chicago Manual of Style, which urges authors to avoid italics, admits that they are often crucial to comprehension, and irreplaceable parts of the language.

So this is the significance of Aldus, I think, not just making a font, but helping the language and the culture grow. And nineteenth century slab serifs had a similar legacy: these are typefaces that grew out of the tradition of novelty typography for display sizes, and like the Aldine italic, they were imagined as standalone typefaces. But once they were produced in text sizes, they found a role in dictionaries, being used to highlight headwords, a role they still play today. So here again the definition of the type family expanded, to include an entirely new typographic dimension — boldface — with which authors could articulate content, and which upon which designers have relied ever since.

The typographic family tree is forever growing. In The Elements of Typographic Style, Robert Bringhurst diagrammed its relationships this way [shown in this slide], which reveals the perspective of a book designer about how styles relate. My hope is that Inkwell helps to push this definition further, in order to better serve both designers and readers.

I’m especially curious to see how designers will use Inkwell’s unusual family tree to render different kinds of information — especially new kinds of information. Wikipedia’s a great example of how text can be subtly shaded — but right now, typography isn’t one of the tools it uses to assist readers. It uses bolds and italics for all the usual purposes, but doesn’t use type to distinguish between “articles,” “past versions of articles,” and “commentary.” But what if it used a typeface that had built-in voices that could help readers distinguish these three kinds of content, styles with strong associations — even for non-designers — that are understood to mean “writing,” “old things,” and “notes?”

And for that matter, why not use typography to separate the site’s content and its architecture, or find a typographic way, instead of a color, to identify a link? In discussing Inkwell with Barbara Glauber last week, I raised Wikipedia as an example of a place where typography wasn’t being used to really provide depth to these new kinds of content, and she wondered, tongue-in-cheek, what Wikipedia might look like in Inkwell. I couldn’t help but oblige, just to see if the fonts are up to the task. And I do like the idea that I can finally see the difference between articles, commentary on articles, and history.

I want a special mode for people who aren’t intimidated by typography, that looks like this. (But that’s probably just for us.) In the meantime, I’ll settle for seeing the fonts wherever designers can use them, hopefully one day in dictionaries, and perhaps, someday, someone’s wedding map. Thank you very much! —JH

Jonathan Hoefler
16 June 2017
The Cooper Union, New York


Your project exceeds the 1,000k limit, so your changes have not been saved.

Try adding fewer fonts, fewer styles, or configuring the fonts with fewer features.