Archives for category: Typesetting

My Modern Met shows us a typewriter which could type music for you. (Link via Shelf Awareness for Readers.) It looks horrendously intimidating.

The Keaton Music Typewriter was first patented in 1936 with a 14 key-arrangement, which was raised to 33 with a new 1953 patent. There were two keyboards: one a rotating keyboard with notes, and another, which didn’t move, with symbols like bar lines whose vertical alignment was constant. The fixed keyboard consists of the nine keys at the right hand end of the ring. The curved bar at the side enabled you to move up and down to place your note at the right height on the staff lines. Each notch on the bar moves the print head 1/24″ — the system requires you to have preprinted paper with the staff lines 1/12″ apart. A needle arrangement enables you to see where the next note is going to be typed.

Here’s a one-minute video showing the machine in (tentative) operation.

If you do not see a video here, please click on the title of this post in order to view it in your browser.

It is assumed that most of these machines would have been used by music publishers and printers rather than composers, who would surely have found it much easier to write their music out by hand.

Music Printing History is a trove of information about the different techniques which have been used in reproducing music over the last 1,000 years.

See also my earlier posts Music engraving, and Typewriter music.

A fleuron is an ornament on a printed page in the shape of a small flower or leaf. It is sometimes called a printer’s flower or floret. It can also be found in discussions among aesthetic architects and numismatists. The Oxford English Dictionary even shows us a quotation referring to pastry flourishes. It is also used to refer to a bookbinder’s tool, carrying such an ornament for stamping on a case cover.

The most common fleuron is perhaps this one, derived from an ivy leaf, and apparently sometimes called a hedera, the botanical name of ivy. These flourishes were often brought together to form borders and rules, though modern taste has moved away from the clutter this often produces.

Here are a few examples from my Linotype Collection book published by the Mergenthaler Type Corporation. On the right hand page you can see how these ornaments could be combined in bravura arrangements. In the hot metal world these fleurons would be available in various sizes — digital or film setting enables you of course to output various sizes from a single “master”, which is why the Linotype book only shows one size. They refer to them as flowers, the least flowery of the possible appellations.

Stanley Morison, a strong advocate for clutter-free typography, was involved in the setting up of the typographical journal The Fleuron; indeed it was he who proposed the name in a letter to Oliver Simon in October 1922. “I should like you to consider re-naming our Typographical Annual. The name I recommend is The Fleuron. This would be an advantage I suggest. In the first place the title ‘Typography’ is very stiff and not absolutely free in the public mind from technical connections. The Fleuron possesses just that note of historical & romantic feeling which we need to express.” Volumes 5, 6 and 7 were edited by Morison (Simon edited the first four) and Morison’s volumes were typeset and printed in Cambridge at the University Press, where as Typographical Advisor to the Press he could supervise the entire process of production. Only these seven volumes were published. The publication was losing money, as anything as lavish almost must, and the demands on Morison’s time became too much to be manageable. Plus, to a large extent, the intention of the journal, to raise standards of British printing, had been achieved after the seven years of The Fleuron‘s life.

The combination of type ornament elements on the title page is obviously an appropriate use of fleurons.

According to The Bookman’s Glossary (R. R. Bowker, 4th edition 1961) Lettre de forme, lettre de somme, lettre de bâtarde are “the three general classifications of Gothic type forms as found in the 15th century. The first is the Pointed and most formal; the second is the Round and less formal; the third is a Cursive form. They correspond to similar classifications of lettering used in the manuscripts that preceded printing.” Nomenclature is of course confusing with something as old as this: lots of people referred to the same thing by different names, and some minority usages still survive in unexpected corners. Just choosing to privilege the French terminology represents a choice of road taken.

In the examples below “Pointed” is referred to as Textura.

From The Dawn of Western Printing at Incunabula.

Here, from Incunabula, is a full discussion of the three forms. It tells us that Gutenberg’s type was based on the script textura quadrata, commonly used for bibles and derived from the “protogothic” script which developed in northern France around the eleventh century. A more rounded script, called rotunda, evolved in Bologna in the 12th century. Lettre bâtarde derived from less formal writing and tended to be used for documents in the vernacular, not in Latin.

Of course it was all more complicated than this, but we can think of scribes using half uncial, chancery hand, and insular scripts, which began to seem unreadable to many as different regions of Europe favored one version over others. Different scripts would be used for different types of book. Alcuin of York (c.730-804) succeeded in persuading scribes to use the more elegant Carolingian minuscule for most purposes, thus allowing international comprehensibility. While the Pointed, Textura style was the basis for the type used by Gutenberg in his Bible, the Carolingian minuscule, a few years later formed the basis for the types which we recognize today as Roman which were introduced by printers in Italy.

Simplified relationship between various scripts, showing the development of Uncial from Roman and the Greek Uncial. From Wikipedia.


Any account of these matters must inevitable over-simplify. Researchers look back on what was the individual practice of thousands of scribes and printers, and impose a pattern on their myriad activities. Such a pattern does provide a sort of history, but doesn’t of course imply any intentionality on the part of the actors. It’s not like Gutenberg said to himself, let’s copy the pointed script. He just knew that this was how you wrote bibles, so, obviously, that’s how you’d print them too. No doubt Aldus Manutius didn’t sit there musing “Why do we have to copy this barbaric Gothic nonsense? Let’s go back to Alcuin and start anew”. He followed the style of earlier classical and secular works, formalizing things as printed types will inevitably do.

My post on Black letter may have some relevance in this context.

This is so obvious to anyone in the publishing business that we are liable to forget that the word stet may be opaque to some other. It’s Latin for “let it stand” and is used whenever, in checking proof or marking up a manuscript, you change your mind after deleting something before second thoughts tell you it should really remain. Write Stet in the margin and the typesetter knows that they should ignore the deletion and let it stand as previously written. To avoid any ambiguity about exactly what the stet refers to we usually put a series of dots under the words involved. We also tend to put marginal proof marks in a little circle to divide them off from any neighboring clutter.

Illustration from The Chicago Manual of Style. Stet is 7 lines down on right.

Quite what the stet means in this photo of a tattoo from BuzzFeed is a bit opaque to me: sort of like Noli me tangere perhaps. Maybe the nails suggest it’s a message about a decision to stick in a poker hand.

A couple of years ago I posted about rivers, those trails of white which can occasionally be found meandering down a page of type. There must be a name for the opposite — a meandering trail of dark wandering down a page, but I’ve either never heard it or have forgotten the word.

This dark river results from a chance proximity of double “g”s or other ink-heavy letter combinations occurring above and below one another. Get two or three of these together, and the eye picks up the blob and shapes it into a river. A funny effect is how, once you’ve noticed it the effect insists on being seen further up and down the page. In this page, from the excellent Anniversaries I by Uwe Johnson (New York Review Books, 2018*) you first pick it up in the center of the page just below half way down, where we find these words in the middle of the line one above the other, slanting a bit leftwards: right, going, approving, front, office, allowed, expert, through, Berlin. Notice also how entirely innocent words (innocent of any dark type clustering) get dragged in by their position in the line. For instance “going” and “expert”. Our eye just includes them in the row because they are there. And then it begins to extend the line upwards and downwards, implicating “disparagingly”, “locally”, and “though” in the line-up. We are pattern-seeking creatures, and this is what we do.

This doesn’t do any harm, and isn’t a “bad” thing. Probably it’s only maniacs like me who latch onto such quirks of typesetting. It’s an artifact of computer-assisted typesetting — well actually of any typesetting. A Monotype compositor in the old days could have broken up such a river by tightening the word spacing here and loosening it there so that these doubled characters were no longer aligned in such a regular way. Of course many comps wouldn’t have bothered: in such things consisted the difference between a good type shop and an OK one. Desperate hot metal compositors seeking to balance their lines have been known to introduce or delete a little word here or there for the sake of aesthetic balance. Occasionally the author might notice.


* I’m about to start on Volume 2, having just reached 19 April 1968.

This blog format uses a line space to divide paragraphs from one another. In conventional text setting there’s no space between paragraphs and the first word of a paragraph starts indented a little to the right of the lines above and below, often 1-em. A hanging indent is the opposite: the second and subsequent lines start indented to the right of the first one, which sort of makes it look like the copy is hanging from the hook of the first line.

As this slightly crooked photograph of a page from the index from Hugh Williamson’s Methods of Book Design reminds us the hanging indent is often found in plays, short notes, lists and bibliographies. And of course indexes.

Erica Van Horn, self-described as half of Coracle Press, tweets this picture of a sign in County Tipperary with the comment “Old sand-cast aluminum letters, hence the spacing.” Equal space between all the letters works out more or less OK here.

So maybe we need to force all public signage back to sand casting, because the letterspacing of the first line is really all one could ask: even and steady — maybe we could do with a bit more between the I and the R at the end. The second line suffers from a bit too much space either side of the A but is otherwise pretty good.

This picture of a French sign for a place south of Grenoble, not sand cast I think, comes from the same source, and is pretty well letterspaced. I believe it shows the other half of Coracle Press, a poet himself, perhaps waiting for lightning to strike.

British road signs tend to be Cap & lower case, as are US ones. Readability is what road signs should be all about, and Cap & lc tends to achieve that end best. Letterspacing lower case type leads to the exact opposite, which is why in German it is used to draw attention to a word, to emphasize it by making you take a little longer reading it.* Road signs need to be read quickly and accurately. No fancy typography here please. With caps, letterspacing does enhance readability.

This sign at the entrance to the George Washington Bridge has obviously been hit by a bus at some point. It shows rather luxurious spacing for G and W but because they are different words this looks fine. The SOUTH at the top is a bit of a mess. Rather than Cap & small cap it’s a Cap S from one size of type, and the rest of the word in a smaller type size, which is reflected in the thickness of the letters as well as their height. There’s precious little spacing going on in that word. If you’re going to put letterspacing in the Cap & lower case Bridge, why skimp on it between i and d? Urban myth holds that road signs are made by prisoners, so perhaps one should not expect too much typographical nicety.

Conclusion: if you are going to letterspace your Caps, ideally use a variable letterspace between different combinations of letters, depending on the amount of apparent space they each bring because of their shape. But variable letterspacing takes time, judgement and thus money, so a constant letterspacing is likely to be much more common. With a constant letterspace, a wider space is likely to cover up more problems than a tighter one will.


* This is because Fraktur/Black letter had no italic form, so that option wasn’t available. Of course printing from movable types originated in Germany, and to begin with no type faces had an italic form until 1501 when the first one was cut in Italy.

LATER: Here’s an interesting street sign from Queen’s New York, set up to commemorate the invention of Scrabble, which took place on this block in 1938. Image thanks to Atlas Obscura.

Dave Addey, interviewed recently on The Kindle Chronicles, is the author of Typeset in the Future: Typography and Design in Science Fiction Movies, (published by Abrams in December 2018) a copy of which he proudly displays to us in this picture from the podcast webpage.

Maybe the subliminal message is meant to be that the future is all about careless typography!* The slight haziness of this photo helps us see how different the first three letters of TYPESET appear from the next three, which huddle together while their predecessors relax in reasonable space provided solely by their shape. Clearly no designer wasted any time thinking about their comfort. In the next line, the built-in spaces at the bottoms of F and T make URE look cramped. Adding a bit of letterspacing would cure these problems and the rest of them and improve things immeasurably. You’d hope someone who is interested enough in typography to write a book about it would notice these things, though I guess the cover was designed by the same person who did the “modern” interior. (You can “Look Inside” the book at the Amazon link above.)

The typeface which drove the creation of this book is Eurostile Bold Extended which Mr Addey noticed was the face used in 2001: A Space Odyssey, and then picked up in many other sci-fi films. It obviously has a futuristic look about it, but I suspect its use in sci-fi movies is more because of imitation of a successful predecessor than it is about any inherent sci-fi-ness of the font. On the other hand I do have to confess to a tendency to want to use Garamond or Bembo for a book about Renaissance culture. Sans serif is “modern”; space travel is modern: QED? Letterspacing may have to be seen as old fashioned: I hope not.


* Confirmed perhaps by Marvel Studios’ logo?

You’ve probably noticed that sort of antiquing effect of using a V instead of a U in typesetting. It’s sort of analogous to those “Ye Olde Tea-Shoppe” signs.

Here’s a pretty extravagant example from an early Culver City studio sign. The underlined O is another antiquing device harking back to medieval scribal practice. Did early Hollywood feel the need for a tradition-boost? I guess so. It was a new business, seeking legitimacy. MGM used the same V in their old sign too — see Letterspacing 2.

The reason it has this antiquing effect is that it actually is antique. It wasn’t until 1629 (or maybe it was 1619) that capital U became an accepted letter when Lazare Zetzner started using it in his print shop in Strasbourg. Zetner also introduced cap J, but that’s another story. See Job case for confirmatory evidence and evidence of the dogged adherence by the print industry to the way-we’ve-always-done-things philosophy. (Louis Elzévier of Leiden is credited with the introduction of the distinction between lower case i and j and u and v in 1518.) They needed introducing because Latin had had no place in its alphabet for j or u, so neither did the early printers.* Scribes had actually evolved a system of distinguishing between u and v, treating them differently at the start of a word and in the middle. V stayed the form used at the beginning by printers — thus the long-term absence of such a sort in printing houses.


* As they had no u early printers obviously didn’t have a double-u either, w, which they would make up by putting two v-s together — v-s in their role as u-s of course, which is why we, unlike the more logical French, thus name the character.
















The Boston printer of this 1693 book obviously had Ws in his Roman and Italic fonts, but not in his black letter typeface where two elaborate V-s are deployed in the word Witches — showing that readers were still somewhat used to the convention. Actually I suppose they might really be Us, mightn’t they? This showing of the typeface Old English suggests that it may in fact have been the old English V which was added later, with the U shown here jumping from sounding as both V and U to just U.

See also Why n, m not v, w? which has a movie poster featuring VV in place of W.

Jeremy asked in a comment on my previous letterspacing post that I do more letterspacing analyses. So here’s goes.




He suggested the Hollywood sign, but I think that’s a bit unfair. Just getting the damn thing up there was hard enough: demanding good lettersapcing from the crew would be just too much. Besides the “letterspacing” changes depending on where you’re standing when you look at it. See how the H and the D behave in these pictures. But Hollywood, in its metonymic sense, is open to letterspacing criticism: for an industry which spends so much money on the look of their films it seems odd to me that they seem to care little about the appearance of lines of caps. Take for example the poster for the archetypical “That’s Entertainment”.

Quite apart from the awfulness of the typeface used in this poster — the best one can say is that maybe someone thought it was entertaining — there’s a feeling of their trying to cram the title into that ugly box while keeping it as big as possible. Come on MGM, make the box bigger, and let the title breathe. It’s not like they always lacked typographical taste. Their old sign was nicely letterspaced, and even had that affected old-style feature of the V in place of the vulgar U.


The fact is that certain letters carry their own built-in spacing. The “A”s in the movie poster need to be further from the letters preceding them than from the ones after: that bend to the left appears to push the character away from what’s following. The half-closed-eyes test will show you that the color of the line clumps into black areas and white areas. The aim should be to make for as even a color as possible. Spacing helps this. Of course, with a quirky face like this, letterspacing the line may draw undue attention to the letterforms themselves. Let’s try an experiment.

Here’s a slightly letterspaced version of the word ENTERTAINMENT. I think it looks a lot better, and might be even better if I’d done a more thorough job with my trusty Xacto knife. Somewhat lazily I left the two NT combinations alone, and allowed that to govern the amount of letterspacing I introduced. I think the thing would benefit from a little more letterspace overall. If I’d added a bit of space between the N and the T — this looks especially necessary at the beginning of the word — that would have helped everywhere else. A little extra spacing between the E and the R and on both sides of the M would help too. The spaces before and after the N ought to be the same. Still, all in all, preferable to the MGM version — to me anyway.

The title cards used in the movie use a different typeface.

Can you imagine any less Fred-Astaire-appropriate face than this clunky grotesque? Well, yes, maybe the one used on the poster. But again the letters are smashed together as much as possible: it’s almost like the designers were on bonus for fitting the most type into the smallest space. All of Liza Minnelli needs letterspacing, Most obviously MI but also, slightly surprisingly the LLI at the end. It’s almost as if the designer realized that the shape of the Ls provided their own spacing, so tried to balance the tight spacing elsewhere by reducing the space following L. The letters touch now! Type-blind in Hollywood.