Does anyone know who invented ink traps, or what year they were first used

chuck chang

In typefaces, i know the best known example of ink traps is Matthew Carter’s Bell Centennial. But It was released in 1978, i believe that he wasn't the first to use ink traps. I found traces of ink traps in this article (as shown in the figure) which was published in 1957
https://issuu.com/birkhauser.ch/docs/adrian-frutiger-typefaces/10

chuck chang

I started to know ink trap fonts from this article and became very interested in their emergence and evolution  https://tosche.net/blog/ink-traps-and-pals

Michel Boyer

In Sofie Beier's Ph.D. thesis Typeface Legibility: towards defining familiarity we can read on page 59

One way of preventing the ink from dissolving the letterforms in the smaller sizes is by opening up the junctions in inktraps. An early example of this, where the outer side of the stroke is cut off at a straight angle, is demonstrated in the work of Johann Michael Fleischman (fig.27), a German punchcutter employed by Enschedé from 1743-68 [...]

The legend of Fig 27 reads:

Example of 8 point type by Johann Fleischman, demonstrating the use of ink-trap in the 1800 century (H. Carter, 1937, p.5).

and the reference is:

Carter, H. (1937) ‘Optical scale in typefounding’, Typography, 4, 2-6

Cory Maylett

I could easily be wrong, but I suspect there was no single inventor of ink traps. Cutting little notches into the type blocks would seem to have been an obvious way to mitigate a common problem that almost every early printer experienced, right from the beginning of moveable type.

Ruixi Zhang

I am confused about Figure 27 in Sofie’s thesis. Were those two letters supposed to be from an actual scan (enlarged, of course) of a printed page? If so, the ink would have dissolved at the straight junctions to create curves, wouldn’t it? Figures 28 and 29 on that same page seemed to show pre-printed letterforms—Bell Centennial loses its ink traps and becomes almost monoline when printed*. A fair comparison should be to show the metal types (flipped, of course) in Figure 27, shouldn’t it? (Side note: I don’t think “1800 century” means what the author intended to mean.)

* At its intended size, using “crappy” printing method.

Michel Boyer

To my eye, Fig. 27 looks very much like what can be found in this bad (but interesting) scan of Harry Carter's paper.

chuck chang

I have the same doubt, the use of ink trap font printing on paper should not be able to see the traces of the trap, this is the affirmation of its functionality, Maybe someone can find more pictures like this to illustrate?  I have very few pictures at the moment ，i found it from this video

chuck chang

This image may be wrong, but I wish I could find more pictures of such pre-print ink trap fonts

Craig Eliason

Mostly you'd need to be looking at punches. Or some examples of master drawings for matrix manufacture. I was a little surprised to not see more inktrap examples in this Letterform Archive collection of Linotype drawings, though it does include these two from the late 1960s: 

Craig Eliason

I'd be curious about the history of the terminology here too: For some period, or in some circles, "inktrapping" referred to the problem of counters or crotches closing up, rather than to the design feature intended to prevent that. 
A. Lawson's Anatomy of a Typeface (1990) uses the word that way, for example.

Ruixi Zhang

Craig Eliason said:

I was a little surprised to not see more inktrap examples in this Letterform Archive collection of Linotype drawings […]

In Merlin Bold fractions, 7 pt., 1969, the numerator “3” (solid) and the denominator “2” (dashed) both featured inktraps, too.

Nick Shinn

The purpose of the “traps” is arguably not for ink.

Bear in mind that trapping is the process by which areas of ink that abut in a layout are minutely oversized, in order for them to overlap in printing, thereby not leaving a fine-line gap of white paper showing, should registration of the separate plates not be absolutely perfect. And this was a general principle, not just for type-against-colour (when not surprinted), but colour-against-colour, and halftone-against-colour, 

This trapping was implemented as part of photographic pre-press production (filmstripping), by over-exposure, which did not involve ink.

In the production of photo-typositor fonts, the purpose of the traps and ticks was to pre-empt the softening (decay) of the type shape during font production and phototypesetting. So in historical investigation, the phototype filmstrips/reels/Diatronic grid-fonts are worth a look, to check for the effect. It is not mentioned in Frank Romano’s History, however, so was perhaps a bit of an under-the-radar trade secret.

When I produced the artwork for my first published typeface, Gryphon, in 1980, for Headliners International, the traps and ticks/flicks were required. Subsequently, for Shinn Sans in 1985, Typsettra also required the treatment, so I assume it was de rigueur by the 1980s. 

Mark Simonson

To put it another way, the traps and ticks in phototype fonts were especially important because of the analog nature of phototype. Phototype fonts had to be duplicated for distribution and you would loose some sharpness each generation away from the original art, including the final type galley and any copies made of it (photostats, negatives, plates).

It isn't really needed in digital type, except to counteract the effects of ink spread and similar issues in the final application of the type.

Nick Shinn

Thanks Mark, that’s what I was groping towards!
When one adds up the number of photo reproductions that occurred from phototype font original artwork to offset printing plate, it was something like six or seven—plenty of opportunity for sharpness to wither.