15 November 2008

The books they tried to ban

Test your knowledge of censors and would-be censors in this quiz from the Guardian.

The cover illustration is from the first printing of Candide in the Penguin Classics series (1947). This was volume 4 in the series. The text may be a very early design by Jan Tschichold, who took up his role at Penguin in March 1947, but the setting, in Monotype Bembo 270, does not follow his famous composition style – dashes are unspaced em-rules, there are extra spaces after sentence full stops, and the long-tailed R is used.

The cover is the original pre-Tschichold design by John Overton; the roundel is by William Grimmond.

I’ve now managed to look closely at the first few Penguin Classics, and the Overton/Tschichold question is rather more complex than implied by the above, or by the simple statement in Baines, ‘only the first seven titles appeared in this design, before it was re-styled by Jan Tschichold in 1947–8’. Not surprisingly, the transition from one design to another in a series in production was not clear-cut. There are early PCs with Overton covers/Overton text; Overton covers/Tschchold texts and vice versa. Some books feature pre-war bowing Penguins, some a Penguin standing on an open book (Baines, p. 251); at least one with a Tschichold Penguin on the half-title, but no device on the title-page. Another has an Overton ‘jacket’ wrapped around what looks like a Tschichold cover.

Baines, P. (2005) Penguin by design. London: Allen Lane (pp. 46, 64–7)

Labels: , , , , , ,

20 September 2008

What photocomposition meant for type

Here are some images from my presentation 'Absolutely no type’ at this year’s ATypI conference.

Books produced by early phototypesetting systems publicized the fact that they were produced with ‘absolutely no type’. What letterforms were chosen for these new systems? How did they relate to existing type designs? What opportunities were taken (or missed) in the creation of new founts? How did the new typefaces for new machines affect the designers and typesetters who used them? By looking at the earliest phototypeset books, manufacturers’ and printers’ type specimens, and printers’ archives 1950-1970 we can find out more about the time when the certainties of metal typography began to dissolve into the new world of film.

All phototypesetting devices broke the link that existed in metal type between character width and escapement, that is the horizontal space in which a character sits. The latter could now be varied independently of character width, allowing any amount of under- or over-spacing of letters. This point is seized upon in this specimen for the Bawtree machine of the 1920s.

The Intertype Fotosetter promoted the new freedom of type design. Its hot-metal faces were constrained in two ways: characters could not kern (that is the top stroke of f could not hang over the following character), and character widths for roman and italic had to be equal, to allow for duplex matrices (which carried both fonts). Garamond seems to have been the first typeface adapted for the new machine, and the revised designs show how both constraints have been thrown aside.

Economy and efficiency were always te selling points for the new machines. The weight of the pieces of film used for a job was compared with the weight of the lead type that would previously have been necessary:

Labels: , ,

28 August 2008

New types for old

This beautiful cover image is from a booklet describing the Rotofoto process, a photomechanical composition system developed in the late 1930s by George Westover, who had worked for Monotype.* Rotofoto, Uhertype (a Hungarian–German system), and the American Intertype Fotosetter are interesting because they show hot-metal type designs being adapted for photocomposition, and setting a high standard right at the start of commercially viable photocomposition.

The Uhertype, whose types are comprehensively discussed by Christopher Burke in his book Active literature: Jan Tschichold and the New Typography, had a comprehensive programme of type design, including versions of Monotype’s Gill Sans and Deberny & Peignot’s ‘French Roman’. The Fotosetter’s first typeface seems to have been Garamond, chosen no doubt because it showed off the phototypesetter’s ability to handle kerning.

The Rotofoto, reflecting its roots within the Monotype Corporation, offered Times New Roman and Monotype Old Style series 2. It’s not clear whether these were redrawn to any degree, or simply photographed from pulls of Monotype-set metal type. The Monotype connection was necessary: the keyboard for the Rotofoto was a Monotype one, and the unit widths of Rotofoto designs would have had to match those of the parent Monotype font.

I’ll be talking more about these and other early phototypesetting machines and the types they used at the ATypI conference in St Petersburg in September.

* See Boag, Andrew, ‘Monotype and phototypesetting’, Journal of the Printing historical Society, new series, 2, p. 58

Labels: , , , ,

26 August 2008

Laurence Urdang

The lexicographer who was a pioneer of computerized dictionary typesetting, Laurence Urdang, died recently. Here is his obituary from the New York Times. (You’ll need to register.)

The following is from my article in Typography Papers 4:

The production of the Random House Dictionary in 1964 was a landmark in the computerization of dictionaries. The managing editor, Laurence Urdang, was the moving force in the early computerization of dictionaries, and immediately envisioned a complete process in which text was entered, stored, sorted and compared, and finally transferred to a typesetting machine. The Random House Dictionary text was keyboarded after writing and each entry was divided and entered in fields assigned to different levels of information (for example ­headword, pronunciation, definitions, etc.). This made it possible to ­prepare information for each level and in each of 150 subject fields, ‘ensuring better uniformity of treatment and far greater consistency among related pieces of information than had been achieved on other dictionaries.’ (Urdang, 1984).

Though Urdang was successful in sorting and establishing the continuity of information throughout the dictionary, he was not able to set up a usable interface between the database and photo­typesetting equipment of the time. Two machines, the Photon and the Videocomp (the US version of the Hell Digiset), were technically capable of being driven by magnetic tape, but the expected slow speed of composition caused by the frequent font changes in dictionary text, and the Videocomp’s inability to produce a true italic, ruled them out. Eventually print-outs from the database were used as copy for hot-metal Monotype composition.

For more information, see: Urdang, Laurence (1984). ‘A lexicographer’s adventures in computing’, in Dictionaries: journal of the Dictionary Society of North America, no. 6 (1984), pp. 150–65

Labels: , ,

11 August 2008

Making notes user-friendly

Readers reasonably report difficulties when flicking back and forth from text to endnotes in editions of classic novels. What can designers do to help them?

There’s obviously a need for discretion in cueing notes in a text intended for continuous reading. One really wants the reader to be able to decide whether to follow up any cue to a note, or simply let the text wash over them. So, first question, what cueing marks to use? An incrementing/sequential system, or a single mark for all notes? Superior numbers or a symbol system?

Oxford English Novels, a hardback (later paperback) series of the 50s and 60s, used notes numbered by page. This means that almost all notes are cued by single-digit numbers, reducing the disruption in the appearance of the line. InDesign can handle by-page numbering. The notes at the end were identified in the following way:

Page 4. (1) It droppeth like the gentle rain: Shakespeare, The Merchant of Venice, iv. i. 181.

A problem here is that the start of each note is identical (‘Page‘), and the note number has to be parenthesized to distinguish it from the page number. The lemma (the words quoted from the actual text that are being glossed) is therefore some way into the note. The reader’s operation in seeing a note is as follows:

1. notice the cue
2. notice the words immediately before the cue
3. notice the current page number
4. turn to the back of the book
5. scan through the notes to find ‘Page X’
6. scan to see the relevant note number – if there is only one note to a page, then the note number is omitted as unnecessary
7. confirm that the lemma matches the words you are expecting
8. read the note

Probably the most difficult part of this operation is 5 – the identical starts to notes and the consistent running headline (simply ‘Explanatory Notes’) don’t help.

When OENs were re-purposed as [Oxford] World’s Classics in the 70s, the system was perpetuated. You might still find an OWC with this system. Newly-set OWCs used a system with much more help provided for the reader in the design of the notes pages, but with a less helpful cueing system in the text. Instead of a sequence of numbers, a single cue mark was used, the asterisk. Because asterisks vary in design from font to font, including some which don’t look like asterisks (Bembo and Plantin, for example, have 5-pointed starts instead of asterisks, and Ehrhardt, the default typeface for OWCs, has a very hairy asterisk that fills in at small sizes) it was decided to standardize of Baskerville asterisks, whatever the text face. A model for this is the annotation in many Norton texts, where a degree sign (º) is used.

The presentation of the notes was re-thought with a stub column for the page number (which is inserted only when the page number changes, and with the lemmas, still italicized, brought to the start of the note itself in the main column.

91crinkum-crankum: a winding way.
abating: a reduction in price.

This system puts a bit more work on to the reader in the book text, because the step of noting the likely lemma (the context of the note) is now the critical step, but provides more help in the actual look-up:

1. notice the cue
2. notice the words immediately before the cue
3. notice the current page number
4. turn to the back of the book
5. scan through running headlines to find ‘Notes to Page X’
6. scan down the stub column to find ‘Page X’
7. scan down the main column to find the lemma that matches the words you are expecting
8. read the note

The use of the stub column, and the use of a vertical space between each note (the latter admittedly used in the OEN system) are the essential components. Essentially the notes are presented as a continuous table, whereas in the OENs they were a simple list. But the use of the page information in the running headline is a crucial piece of redundancy (redundancy meaning the duplication or re-presentation of information in a way that helps the reader) allowing for stage 5 in the reader’s process. A further consideration is that the page extents in these running headlines (‘123–134’) should probably be set in full rather than in a space-saving convention (‘123–34’).

As an aside, superior numbers are now available in correctly designed form in OpenType fonts (PostScript fonts used to be very variable in their support) so it’s easier to specify numbers that align correctly and blend in colour with the text. There isn’t any excuse for using normal figures scaled and aligned as superiors 123 – which were often too light, too narrow, and too high, unless the designer had carefully specified the parameters to be used.

I'll add some illustrations shortly.

Labels: , , , , , ,

08 August 2008

Some readers like footnotes (2)

There is certainly a trade-off between convenience of access and simplicity of appearance in deciding whether to place notes on the page or at the end of the book – especially if the book started life as a non-academic text and explanatory notes have been added later. In the early days of phototypesetting, when pages were made up by hand using (literally) cut and paste, footnotes became an additional cost, and were avoided where possible by many publishers. Large-scale, automated page make-up systems such as Miles and Penta made the position of notes irrelevant, as the pagination process could deal with foot-of-page, end-of-chapter, or end-of-book notes equally well; but these systems could not compete on cost with simpler Mac-based applications such as Quark XPress, which had nothing like the functionality. Footnotes in XPress used to be a real pain.

InDesign’s relatively competent handling of footnotes means that there is no reason for foot-of-page notes to be avoided any more, but I wonder if there is still a folk-memory that footnotes are ‘difficult’? I was pleased to see that some readers are still in favour of footnotes, as evidenced by this blog – but note the rueful ‘I guess footnotes have been done away with in this day and age’, as if publishers have persuaded readers that ‘there’s no call for them now, sir.’

Labels: , , , , ,

28 July 2008

The men in white coats

In films and tv dramas of the 1950s and 1960s, the scientist (usually male) was instantly recognizable by his white coat. As the printing industry began to move to photo-typesetting and computer-aided composition in the same period, the image of the printer in a white coat was part of the message to customers that here was a new, less inky age, with highly trained technicians producing perfect results from finely calibrated equipment. The printer was seen with the new tools of his trade: a light-table, film, and a scalpel. All that was missing was the stethoscope. Even when the machinery was not so different (Monotype used the same basic keyboard for its Monophoto machines as for its hot-metal system) the operator was duly dressed up in a white coat for the publicity photo, eschewing the traditional printer's brown garb.

This attempt to rebrand the printer went alongside calls for designers to re-think themselves: ‘Print designers … should be trained not in art schools, but in schools of engineering print design.’ Thus John Duncan in 1964. ‘There is still too much woolly thinking in out training programmes for designers. Too often the basic disciplines of draughtsmanship and the cardinal responsibility of communicating an idea or message are overlooked or neglected and issues are clouded by the striving for the vague goals of so-called originality and aestheticism,’ wrote Lawrence Wallis in 1965. ‘This is an age of specification writing,’ continued Duncan, accurately describing the role of the designer for computer-controlled composition as someone who had thought out all the issues beforehand, so the the machines could run at maximum speed, with little need for time-wasting corrections. ‘What other industry,’ wondered Wallis, ‘would put up with the specification being altered half way through the production of a job, or even worse with no specification at all, or with an inaccurate one.’

Alongside the elevation of the specification was the development of work-flow diagrams: the one below even schematizes the operations with labels such as ‘intellectual–manual’ and ‘automatic–electronic’.

What was this a reaction against? Perhaps the best example I can find from the early post-war years of a designer being determinedly anti-scientific (if you take scientific as meaning a method involving measurable, repeatable things) is Jan van Krimpen, whom Robin Kinross called 'the principal bearer of tradition in Dutch typography’. Here is is talking of how he would ‘specify’ amendments to a typeface design:

‘I meant to offer … to have them put in order, under my own eyes, by my own punch-cutter. … When we talk of punches cut by hand we use terms like “a hair or even less” or “the tiniest little trifle” &c. I am sure they could not convey or mean anything to the people of the [Monotype works who] talk of “tens” and “thous[ands of an inch]”.’

Sources: Penrose Annual, 1964; The Monotype Recorder, 43, 2, Summer 1965; OUP archives

Labels: , , ,