A short paragraph, typeset in Input Sans, a proportionally spaced typeface by David Jonathan Ross

Same look, less pain:
Proportionally spaced typefaces with a monospaced appearance

76 people pressed and 12 people found it worth a retweet when Ivo Gabrowitsch wrote: “If I had one wish, I’d wish for less ‘inspired by’ and more ‘solves this problem’ typefaces.” It is hard to disagree with what he says – but it is also hard to disagree with Ksenya Samarskaya’s reply: “Best ones do both, no?” This post is about what I consider to be a successful example of doing both.

Many designers are in love with the aesthetic of monospaced typefaces (i.e., typefaces in which each character occupies the same amount of horizontal space, as opposed to proportionally spaced typefaces in which character widths are independent of one another). The appeal of these typefaces may stem from some kind of nostalgic association with typewriters or from their use in technical environments (computer terminals, programming etc.). Maybe it’s both. In any case, the fascination is there. As a consequence, monospaced (or fixed-width) typefaces are used everywhere – not least in contexts in which their defining feature (fixed spacing) is irrelevant. But this feature keeps on having an impact, mainly on the experience of the reader. In monospaced typefaces, glyphs have to be wider or narrower than in typefaces without the fixed-width constraint in order to fit everything in the space that is available while keeping the spacing optically balanced. The result is not what most designers would like to set a novel in and it’s not what most readers would like to read a novel in. Actually, anything beyond a short stretch of text is a pain to read in a monospaced typeface, however nice things may look at first blush.

A short paragraph, typeset in Input Mono, a monospaced typeface by David Jonathan Ross

Type designers have found a solution to this problem. It’s not a new solution, but it is a solution that I think deserves more attention (hence this post). Type designers have created typefaces that look like monospaced typefaces, but actually use proportional spacing. The benefit: Designers get to keep the look they love, but readers don’t have to go through the pain of a true mono. Such typefaces – some call them ‘monofaked’ or ‘fauxnospaced’ – are easier on the eyes than the ‘real thing’. It is a balancing act for type designers to keep enough elements of typical typewriter fonts in order to avoid losing the appearance, while at the same time making substantial improvements to reading ease.

A short paragraph, typeset in Input Sans, a proportionally spaced typeface by David Jonathan Ross

The first typeface that fits this description was apparently one called Bulletin Typewriter: Released in metal as a monospaced font, it became available with proportional spacing in phototype and transfer lettering formats. The earliest in-use example of the proportionally spaced Bulletin Typewriter I am aware of is from 1973. American Typewriter was released not much later. In the following list of other typefaces in this category, I will – as usual – apply a liberal interpretation of any relevant criteria: Any typeface that vaguely looks like a console or typewriter typeface (read more about this term in an article by María Ramos) has a ‘monospaced appearance’ in my book. For a change, the sorting will be chronological rather than alphabetical to emphasise developments in this genre.

Proportionally spaced typefaces with a monospaced appearance

  • 1973: Bulletin Typewriter by Morris Fuller Benton (Mecanorma)
    (The link refers to a digital version of the metal monospaced typeface, originally released in 1933 by ATF. A proportionally spaced version was only available in phototype and transfer lettering formats. It’s in the list mainly for reference.)
  • 1974: American Typewriter by Joel Kaden & Tony Stan (ITC)
  • 1989: Officina Sans & Serif by Erik Spiekermann & Just van Rossum (ITC)
  • 1996–98: Letter Gothic Text by Albert Pinggera (FontFont)
  • 1999: TypeStar by Steffen Sauerteig (FontFont)
  • 2000: Bs Monofaked by Mário Feliciano (Feliciano Type Foundry)
  • 2000: New Letter Gothic by Gayaneh Bagdasaryan (Paratype)
  • 2001: Courier Sans by James Goggin (Lineto)
  • 2007: Newsletter by Ingo Krepinsky (Die Typonauten)
  • 2008: Generika by Alexander Colby (Milieu Grotesque)
  • 2008: Lacrima Senza & Serif by Alexander Colby (Milieu Grotesque)
  • 2008–2010: Lekton by Luciano Perondi and students at ISIA Urbino (Google Fonts)
  • 2010: Typewriter by Henrik Kubel (A2-TYPE)
  • 2011: Hellschreiber Sans & Serif by Jörg Schmitt
  • 2011: Relative Faux by Stephen Gill & The Entente (Colophon Foundry)
  • 2011: Signika by Anna Giedryś (Google Fonts)
  • 2012: Anaheim by Vernon Adams (Google Fonts)
  • 2012: Executive by Gavillet & Rust (Optimo)
  • 2013: Documan by Martin Vácha (Displaay Type Foundry)
  • 2014: Input Sans & Serif by David Jonathan Ross (DJR)
  • 2014: Queue by Tal Leming (Typesupply)
  • 2014: Triplicate by Matthew Butterick
  • 2015: Clone by Lasko Dzurovski (Rosetta)
  • 2016: Millimetre by Jérémy Landes (Velvetyne Type Foundry)
  • 2016: Operator by Andy Clymer (Hoefler & Co.)
  • 2016: Proportional by George Triantafyllakos (Atypical)
  • 2017: Attribute Text by Viktor Nübel (FontFont)
  • 2017: Bitcount Prop by Petr van Blokland (TYPETR)
  • 2017: Comspot Tec by Nils Thomsen (TypeMates)
  • 2017: iA Writer Duospace (based on IBM Plex Mono by Mike Abbink & Bold Monday)
  • 2017: Ultraproxi by Ray Larabie (Typodermic Fonts)
  • 2018: Clincher Duo by Alexander Lubovenko (ParaType)
  • 2018: Covik Sans Mono by James Edmondson (OH no Type Co)
  • 2018: Drive Prop by Elliott Amblard & Jérémie Hornus (Black[Foundry])
  • 2018: Tuner by Simon Renaud (Production Type)

This list started on Twitter. Thanks to Reed Reibstein, David Jonathan Ross, George Triantafyllakos, Daniel H., Dave Coleman, Anthony Masure, Martin Wenzel, Frank Adebiaye, Michael Piotrowski, Akira Yoshino, Eric Mellenbruch, Max Phillips as well as the people from Studio Het Mes, Fonts In Use and Displaay Type Foundry for suggesting typefaces. If you know of any additional typefaces that may qualify, please get in touch.

Olicana, a script typeface with optical sizes

Hands for all sizes:
Script typefaces with size-specific variants

Size matters – at least in type. With digital type, we can use any font at whatever size we desire. But more often than not, we should not. Typefaces that work well at a certain size usually do not work equally well at other sizes. A typeface with pronounced contrast, sharp details and tight spacing that looks great on a billboard will fail at typical body copy sizes. Conversely, a typeface with coarser details and sturdier proportions, optimised for small sizes, is likely to look clunky when set large.
FF Tisa and Orpheus
But what if you would like to use the same typeface for headings and text? In that case, you may want to consider a typeface with ‘optical sizes’ (i.e., variants with size-specific adjustments). Many such typefaces exist (Identifont has an alphabetical list). Tim Ahrens and Shoko Mugikura have written about some of them and provide a list of all typefaces covered in their book on their website. Browsing through these lists, I noticed that few are script typefaces (i.e., typefaces based on calligraphy or other types of handwriting). That surprised me, given that handwriting is also markedly size-specific. The release of a new typeface in that category was the reason for me to compile the following list:


Welcome to @isoletters

On 1 January 2018, I will start sharing pictures of letters on a new Twitter account: @isoletters. I have been taking letter pics for years and occasionally shared them, for example on my @‌isoglosse Twitter account (which I will continue to use for anything except letter pics) or on Flickr. Browsing through my collection, I decided that I would like to share pictures from it more regularly. A collection is not worth much when nobody looks at the items and maybe gets inspired by them. For that reason, I will try to post one picture on each day of 2018. You are welcome to follow. I also hope that you will comment, correct me if I am wrong and share your own pictures. What kind of pictures can you expect?
All pictures I will share on @isoletters are of some typographic interest (or at least I think so). Given my own background, I may also slip in some pictures that are interesting from a linguistic perspective. When trying to classify the pictures I collected, one may end up with roughly the following categories:


Old type in new fonts? First impressions of Fontstore

A few days ago, Font­store launched – a new player in the market for subscription-based font services. To be accurate, Font­store is both a foundry and a font service provider: You can buy fonts there like at any other foundry, but you can also become a subscriber. For $ 15 per month (or $ 150 per year), you get access to their entire library of desktop fonts. Having used Typekit for a few years and having reviewed the Monotype Library Subscription (MLS) last year, I was curious to see what Font­store has to offer. The very first impression was not too favourable: Their website is slow and cluttered up with animations and effects, making it difficult to get an overview of their collection (others loved their site or also hated it). Font­store does not excel at transparency either: The name of its founder, Satya Rajpurohit, is nowhere to be found on the website (though he has promised to remedy this). In order to access the Font­store Library (the contents of which will be discussed below) through the subscription service, I subscribed and payed by credit card; both went without problems. Installing the app syncing the fonts on Windows was not exactly a seamless experience, though: I got a number of warnings that are typically not shown upon installing software, but just going ahead did not crash the system. Once installed, you go to the website, select the fonts you want to use, and they are available just seconds later.

Slim pickings for type designers

As I noted in my review of the MLS, low-priced font subscription services do not only raise technical questions, but ethical ones as well. Creating fonts is a time-consuming task requiring specialised skills that are gained through extensive training and practice. Few type designers create fonts just for the fun of it. Most do so to make (part of) a living – but you need a lot of subscribers before you are able to provide decent payment to all type designers. As a customer, you have to decide: Is subscribing to such a service tantamount to supporting exploitative working conditions and fuelling the downward price spiral in the font market? In the case of Font­store, the decision may be a little easier: All typefaces were commissioned for that service with (I presume) designers knowing what they signed for. In a post on TypeDrawers, Satya Rajpurohit, the founder of Font­store, explained that Font­store offers different payment models to type designers, neither of them based on royalties. In any case, the business risk mainly rests with Font­store, rather than with individual designers. That seems friendlier than royalty-based models, but $ 15 per month is still not much to go round, so time will tell if the system is sustainable. Also, even the best offer in a given situation can be a bad one: Font­store carries fonts from quite a few young type designers, some of them releasing their first retail fonts. Many senior designers may have been reluctant to accept Font­store’s payment models and have their fonts offered at a bargain price.


On cognates and friends, true and false

Learning a language is hard, so every little bit helps: If English is your mother tongue and you start learning German, you will be delighted to hear that the German word for hand is Hand. You may be less delighted to hear that the English word gift does not translate to its German orthographic counterpart Gift (which means ‘poison’), but to Geschenk. Then again, if you like linguistics, you will be delighted to hear that not just the former, but also the latter pair of words is etymologically related (both gift in English and Gift in German refer to something that is given to someone, the verb give also being related to the nouns). So, we have three aspects of relatedness in words: the origin (where words come from); the form (how they are written); the meaning (how we define them). For some clusters of relatedness, there are established terms: The term I have always used for words that share the same spelling, but not the same meaning is ‘false friends’ (e.g., gift/Gift). Today I learned through a blog post by Jonathon Owen that some English speakers (erroneously) refer to such false friends as ‘false cognates’ – a term that would properly be applied to words that share the same meaning and the same form across languages, but that are in fact etymologically unrelated. Inspired by said blog post, I made an overview of different types of relations between word pairs across languages, based on origin, form, meaning or some combination thereof. You can download a PDF version of the chart by clicking on the image below:

An overview of words that share the same origin, form, meaning or some combination thereof

Die englische ›type foundry‹ und ihre deutschen Äquivalente


How do you translate type foundry into German? Gießerei (literally castery), the traditional term from the era of metal typesetting, is not used for digital foundries and seems unlikely to be revived. This text is supposed to ignite a debate about possible alternative terms. These include Schriftanbieter (literally type offerer), Schriftenhaus (type house), Schriftherausgeber (type editor), Schrifthersteller (type producer), Schriftlabel (type label), Schriftverlag (type publisher) and the English loan word Foundry.


Wie übersetzt man type foundry ins Deutsche? Bei einer Diskussion auf Twitter, die sich um die Suche nach einem Begriff für die deutsche Wikipedia drehte, wurde klar, dass es hierzu keinen Konsens gibt. Grund genug, sich das englische Original und einige mögliche Übersetzungen genauer anzusehen.

Den englischen Begriff definiert Wiktionary als »a company that designs and/or distributes typefaces« (»ein Unternehmen, das Schriftarten gestaltet und/oder verbreitet«). Erster Bestandteil ist type – ein Wort mit griechischen Wurzeln (altgr. τύπος ›Schlag, Stoß; Abdruck; Gepräge, Relief; Abbild, Vorbild, Vorlage; Form, Gestalt‹ zu τύπτειν ›schlagen, stoßen‹), das, übers Lateinische und Französische, Ende des 15. Jahrhunderts ins Englische kam. Ausgehend von älteren Bedeutungen wie ›(Vor-)Bild, Vorlage; charakteristische Form‹ wird es seit dem frühen 18. Jahrhundert für Drucklettern verwendet. In dieser Bedeutung wird es im Deutschen mühelos mit ›Schrift‹ wiedergegeben, so auch in einigen Wörtern, die im Weiteren besprochen werden. Der zweite Teil der Begriffs ist es, der bei der Übertragung ins Deutsche Schwierigkeiten bereitet. Drucklettern entstanden bzw. entstehen in einer foundry. Auch dieses Wort, das im Englischen seit dem 16. Jahrhundert belegt ist, stammt aus dem Französischen: Das Verb fondre, das wiederum auf das lateinische fundere zurückgeht, bedeutet unter anderem ›schmelzen‹. Eine fonderie ist eine Werkstatt, in der Metall geschmolzen und in neue Formen gebracht wird, zum Beispiel Buchstabenformen (deren Kombination zum Zeichensatz einer Schriftart als font bezeichnet wird). Im Deutschen bezeichnet man solche Werkstätten seit dem frühen 17. Jahrhundert als Gießerei bzw., wenn sie sich mit der Herstellung von Buchstaben beschäftigen, als Schriftgießerei.

Der deutsche Begriff Gießerei ist, anders als sein englisches Gegenstück, in der Zeit verblieben, in der Buchstaben aus Metall waren. Für Unternehmen, die sich mit digitaler Schrift beschäftigen, hat er sich nicht eingebürgert. Aber was soll man dann verwenden? Dass sich die Frage überhaupt stellt, dürfte auch daran liegen, dass ein Großteil des Geschäfts in dieser Branche auf Englisch abgewickelt wird. Auch auf Deutsch wird über Schrift geschrieben, aber häufiger noch auf Englisch. Die Zahl der type foundries, die eine deutschsprachige Website unterhalten, ist niedrig. In der Diskussion und bei der Recherche tauchten dann doch sechs Wörter auf, die die nähere Betrachtung lohnen: Schriftanbieter; Schrifthaus; Schriftherausgeber; Schrifthersteller; Schriftlabel; Schriftverlag (bei allen kann der erste Teil auch ›Schriften-‹ lauten, aber die Verteilung der Formen ist nicht bei allen Begriffen dieselbe).


Auf ein Wort (24): Céline Dion

Céline Dion frz. (QC) [seˌlɪn ˈd͡zjɒ̃]; frz. (FR) [seˌlin ˈdjɔ̃]; engl. (AE) [səˌlɪn diˈɑn]


Angesichts der Bekanntheit von Céline Dion war ich überrascht, wenig Diskussionen über die Aussprache ihres Namens zu finden. Als ›ursprüngliche‹ Lautung kann man wohl die betrachten, die der quebecischen Aussprache des Französischen folgt. Mit dieser Varietät ist die Sängerin aufgewachsen. Im Vergleich zur Aussprache in Frankreich, die hier als Zweites angegeben ist, fallen einige Unterschiede auf.

Zum einen ist der Vokal in der zweiten Silbe des Vornamens in der kanadischen Aussprache zentralisiert. Zu den geschlossenen Vokalen [i], [y] und [u] in Frankreich kommen in Québec [ɪ], [ʏ] und [ʊ] hinzu, die erstere in vielen geschlossenen Silben ersetzen und sowohl lang als auch, wie hier, kurz sein können. In ihrer englischen Aussprache des eigenen Namens übernimmt Céline Dion diesen zentralisierten Vokal. Es sei darauf hingewiesen, dass die Lautung mit [iː] – also [səˈliːn] oder auch [seˈliːn] mit Vollvokal in der ersten Silbe – im Englischen von anderen Sprechern am häufigsten zu hören ist.

Ein weiterer Unterschied zwischen den zwei französischen Varietäten findet sich im Nachnamen: Im quebecischen Französisch werden [t] und [d] vor Vordervokalen affriziert, also [t͜s] bzw. [d͜z] gesprochen. Außerdem ist der Nasalvokal am Ende im Vergleich mit der Aussprache in Frankreich offener. In Québec werden die Laute, die in Frankreich nasale Monophthonge sind, zudem von einigen Sprechern diphthongiert – im Fall von [ɔ̃] zu [ɒ̃ʊ̯̃]. Die Aussprache des Nachnamens in einer Silbe wird im Englischen zu zwei Silben aufgebrochen, wobei in der zweiten, betonten Silbe meist ein Vokal aus dem ›lexical set‹ von lot verwendet wird, also [ɒ] im britischen und [ɑ(ː)] im amerikanischen Englisch.


Für deutsche Sprecher ist der Name keine allzu große Herausforderung. Als Eindeutschung zu empfehlen wäre [seˌliːn ˈdi̯õː] (ungefähr: ßee-LIHN DIÕ). Der Vorname reimt auf ›Berlin‹. Der Nachname kann auch zweisilbig – also [diˈjɔ̃] – ausgesprochen werden. Wer kann, sollte der im Deutschen verbreiteten Tendenz widerstehen, den Nasalvokal denasaliert und gefolgt von [ŋ] auszusprechen.


Vacuüm in het woordenboek

Onderzoek in de media: discussie in een vacuüm

Gister stond het in de krant: ‘Lager loon voor platprater’. Het artikel in het AD ging over onderzoek dat Jan van Ours, hoogleraar arbeidseconomie aan de Universiteit van Tilburg, samen met Yuxin Yao heeft uitgevoerd. Op 16 september is het onderzoek gepresenteerd op het EALE-congres in Gent. Naar aanleiding van het krantenartikel was er enige discussie over het onderzoek in de (sociale) media.* Die discussie was in de eerste plaats gekenmerkt door een opmerkelijk gebrek, namelijk een gebrek aan details. Het artikel in het AD maakte geen gewag van zo’n beetje alles wat je over een dergelijke analyse wilt weten. En in eerste instantie bleek er ook geen onderzoeksrapport, manuscript of artikel beschikbaar met de nodige informatie. Pas later op de dag vond ik een rapport van een paar maanden geleden, maar we weten niet of dat de huidige stand van de analyses weergeeft. Desondanks schroomde menigeen niet om zijn zegje te doen over het onderzoek, gebaseerd op een krantenartikel.

Dat kan volgens mij beter. Hier lees je wat ik journalisten en wetenschappers zou willen aanbevelen in het belang van een serieuze discussie over de resultaten van wetenschappelijk onderzoek:

Vacuüm in het woordenboek

Identical zero and oh glyphs

Zero vs. oh: Strategies of glyph differentiation

When characters look similar to one another, misreadings or even misunderstandings may result. In such cases, type designers often try to avoid glyphs looking completely identical. One example of a similarity that can get too close for comfort is between the digit ‘0’ (henceforth: zero) and the letter ‘O’ (oh). Differentiation strategies between lining zero and upper-case oh have been shown to be fairly uniform across typefaces: Zero is almost always narrower and often less tall than Oh. These and other strategies have been discussed in a 2013 article by Charles Bigelow. He focuses on lining figures in typewriter typefaces and only briefly mentions contemporary approaches to non-lining zero and lower-case oh – a pair that is also prone to confusion and seems to be treated with less uniformity in recent typefaces. Inspired by a tweet by Shiva Nallaperumal, I have compiled typical and not so typical ways of distinguishing non-lining zero from lower-case oh. My observations are mainly based on old-style roman text typefaces in the Typekit library; this sample is not representative of anything but the Typekit library itself, which, however, contains a bunch of well-known, widely used typefaces.


Charis SIL, used in the Longman Pronunciation Dictionary

Fonts for phonetic transcriptions: An overview

In 2014, I started compiling a list of fonts for typesetting phonetic transcriptions using symbols of the International Phonetic Alphabet (IPA). It is probably the most comprehensive and up-to-date list of such fonts, providing short, yet detailed reviews of the typefaces and the quality of their symbols – but it has two disadvantages: First, the reviews are written in German, which most people do not read. Second, the list – featuring more than 40 typefaces – has grown quite long and maybe even a bit confusing. If you are looking for a decent sans-serif typeface that includes phonetic symbols in its bold style, the long list will not be much help. That is why created a table of all fonts for phonetic transcriptions I am aware of (thanks to Friedrich Althausen, the designer of the Vollkorn typeface, for the suggestion!).
Charis SIL, used in the Longman Pronunciation Dictionary by John C. Wells
For each font family, I have indicated whether the roman and italic styles in the regular and bold weights contain phonetic symbols (R: Regular Roman; I: Regular Italic; B: Bold Roman; BI: Bold Italic).¹ Whenever a typeface family includes more than these four styles, this is noted in the ‘More styles’ column (but you’ll have to look up the details for yourself). This is also true when a typeface has more than one bold weight (e.g., Semibold and Bold) or more than one italic style (e.g., ‘true’ italics and oblique). If you want to read the reviews of the typefaces, click on their names (warning: 🇩🇪 content ahead).

In each category,  means that a (more or less) complete set of phonetic symbols is available. Even if a font is marked that way, some symbols may be missing. (✓) means that a style or weight is present in the typeface, but that it does not contain phonetic symbols.  means that a style or weight is missing entirely from the typeface in question.² The last column contains a rating: This is not about whether a typeface is nice in general, but only refers to the completeness, drawing quality and functioning of the phonetic symbols. Please note that a five-star rating, used for the best typefaces in this area, does not imply that the typeface is absolutely complete and utterly flawless; minor deficiencies may remain. In any case, the rating is subjective, of course, so feel free to comment if your evaluation differs from mine. Also, please let me know if you know of any other typefaces with phonetic symbols.