Course outline


References Doc

Enquiry Doc

Thursday, January 31st 2019

The workshop is essentially about questioning type. how do we develop it, how is it displayed, what does it do. The work eventually goes into Typographic Singularity (Hopefully at Elephant West).

The brief is Make a piece of typographic work that aadds some dimension, such as

  • Time (kinetic type)
  • Space (dimensional typography)
  • Data (generative stuff)
  • Interactivity

The work needs to be based on a text, which can be:

  • a location or
  • a factual statement or
  • a poetic statement or
  • an opinion of yours.

The outcomes can be speculative (yuch).

Week one is about subversion of tools and processes. If your tool is InDesign, question its assumptions (why is the page limited? Why does it give me default font choices) and subvert them.

The history of type (slightly abridged)

  • Lettering ≠ Typography
  • Type is about systems, repetition, process
  • The first kind of writing is Cuneiform
  • Interestingly, Cuneiform can be applied to different spoken languages in the same way the Roman alphabet can.
  • Early written languages are essentially tools for bureaucracy: Most of teh clay tablets we have say stuff like Farmer so-and-so has 12 goats, and owes 3 sacks of grain in taxes
  • Cuneiform is a 3d-language: the depth of the cuts carries information. This is why we 3d-scan them instead of photographing them
  • Next: The Romans
  • You can roughly draw this line to describe the development of the modern alphabet from the roman: Square capitals → Rustic Capticals → Unical → Carolingan Miniscule → Modern Writing
  • For about 400 years (from Rome to Gutenberg), writing was only done by trained monks.

36-line Bible Gutenberg’s 36-line Bible (1458-1460). Commons

  • Letterpress is great for Roman type, but other languages (Arabic, Asian languages) often have to be compressed, simplified to work in letterpress.
  • OpenType gives us way more power: We can essentially program all kinds of behaviour directly into our typefaces (such as contextual alternates).
  • Unicode is big enough to hold enormous charactersets (like you might need for Chinese), way more than a printer’s typecase or Linotype keyboard.
  • Typefaces can advance social goals: Aravit cobines Arabic and Hebrew script in such a way that speakers of both languages can read it.
  • Monospace typefaces exist to make typewriters work.
  • Duospace
  • in the 1970s, photosetting allows people to do all kinds of thinhgs that were impossible in letterpress: Stretch, compress, rotate, scale type freely. Fonts for photosetting had to have reverse ink-traps so they wouldn’t look rounded off (because light would bleed around the edges).
  • Demos is a typeface that’s inspired by the smoothed-over look you get from photosetting (also the kind of thing that would be impossible to cut into a punch).
  • Compare also Retina and Charter (the phonebooks with the mad ink traps) and Miniscule (a typeface that’s designed to be readable at 2pt)
  • A whole aesthetic comes from dot matric printers and low-res LED-displays. LEDs of course are also very good at making type move.

Poem Field No. 1 (1967) by Stan Vanderbeek

  • Machine readable typefaces: OCR-A and OCR-B
  • Wim Crouwel (1967): New Alphabet
  • Tomato: Sony Corporate Identity
  • With LCD screens hinting starts to become a thing. Wonder how much that changes how we design, think about, read type. Verdana was a huge design effort, largely because of all the manual hinting.
  • Then in the 90s we start to get what we’d probably call Grunge.
  • Emigre, Fuse, Raygun Magazine
  • Brody: FF Blur (which is in the MoMA)
  • Also around the early 2000s: Experiments in interactive type
  • FF Beowolf by Blokland and Rossum is probably the first typeface that has behaviour programmed into it: Each time you type a letter, the vector points are moved around by a randomised amount (within certain limits — these are the cuts of the typeface). This is essentially asking the question: Does every letter in a typeface always have too look the same (as it has done the entire history of printing)?

Febuary 28, 2019

Nicky Hamlyn on text in film.

  • Duchamp (1910). In those days you couldn’t buy a camera: You had to have one made. Rotating poems that repeat themselves, also wordplay and moving around of letters. Anemic Cinema (1926). See also Vertigo records in the 70s.
  • Also: L.H.O.O.Q. (1919)
  • Word Movie (1966) by Paul Sharits. Like a lot of his early work, this is done frame-by-frame.
  • Don’t look back documentary on Bob Dylan.
  • Subterranean Homesick Blues. Dropping cue cards.
  • Michael Snow So Is This (1982). This film is two hours long: does that seem like a frightening prospect? (The actual thing is 48 minutes long, of course this kind of falls apart when you watch it on Youtube). Apple ripping it off. Film about
  • Title sequence from Jean-Luc Goddard: Pierro Le Fou. Technicolor was shot on three rolls of 35mm black and white film with coloured filters, which would later be added together in printing. The red, blue and white is a Jean Luc-Goddard thing. Letters come in in alphabetic order.
  • Associations (1975) by John Smith. Also: Steve hates fish (2015), which is the French > English translation thing.
  • Kurt Kren often uses simple technical ideas to produce interesting effects. 42/83 No Film (1983): A collapsing of the title and the film. The thing announces it’s a film, yet it can only exist as a film: negative image is a fundamentally photographic process.
  • Lost Highway opening sequence

March 7, 2019

Idea that a pretty basic mathematical process (Levenshtein) is mashing these different authors together. Original author, secondary author, the machine, the person typing are all interacting with the same text. The fundamental operation here is the Levenshtein distance:

Next week: Continue the exploration (we’re making the books). Two weeks later (on the 28th) is another session that’s about what’s actually going to be in the show.

I think me trying to find interesting text/letterform combinations is a bit contrived. The more compelling thing is to produce every possible combination of books (minus the originals) and finding out where interesting coincidences happen. In other words:


The number of books I need to make based on this principle is:

So for 12 sources (which seems like a lot):

I guess the good thing is that I can just start making books (once I’ve written code to do that automatically) and keep adding sources until I run out of time.