Williams Caslon Small and The New Yorker

Williams Caslon Small is my new text font for small sizes that I did for the partial redesign (as of Sept 23) of The New Yorker. Its proportions are based on Caslon's original Long Primer size, and the shapes are based on my Williams Caslon Text (for 10-12 point), already published by Font Bureau.

Attached is a comparison of the old and new. The old was Sabon I believe at 8/8, negatively tracked, which was pretty taxing to read. Mine is set at 8/9, also negatively tracked, by 20 units. I think mine ideally shouldn't be negatively tracked more than 10 even in such a narrow column, but it still works, and is much more comfortable a read. You have to actually see this in print for a real comparison, but I'm gratified that so far it's had a very positive reception.

TNY creative director Wyatt Mitchell explains the ideas behind the redesign in this video.

The redrawing of the iconic New Yorker titling face Irvin, with additional raised characters and ligatures, was done by Ben Kiel of House Industries. Christian Schwartz also apparently did some modifications of his Neutraface (also published by House Industries) for the redesign.

GOAT-new-old.jpg403.69 KB
eliason's picture

A definite improvement!

John Nolan's picture

Nice. (And what an improvement!)

dhannah1000's picture

Great Job! Keep It Up TNY!

k.l.'s picture

Very nice! (If the typographer had letterspaced Neutraface a bit, it'd have looked more generous. Red lines at least.)

Frode Bo Helland's picture

The ligatures and alternates in Irvin is a huge step backwards, IMO. The text face changes is where the real thing is happening. Kudos!

Nick Shinn's picture

At first impression, Caslon is a strange choice—doesn’t fit the Art Deco era vibe that Irvin and Neutraface carry.

Apart from misplaced historical allusion, it’s rather high-minded and didactic (to quote from the review of The Butler shown above) and lacking in glamour, especially in Bill’s readability-first treatment.
Something less of a workhorse, with a bit more style, would suit the New Yorker more, IMO.

However, I do like what I’ve seen of the Irvin shenanigans, and perhaps the plainness of the body text can complement the visual interest there—I’ll keep a look out for it in print to see if the real thing impresses me more than what I’ve seen online.

John Hudson's picture

From a readability perspective, this surely seems an improvement. But stylistically I think the redesign is a bit of a dog's breakfast. Let me be clear that I've never liked The New Yorker, and the precious Art Deco-isms are one of the things that turn me off. But if they're going to stick with those artifacts, then they should have conviction. Now the display typography and the text seem to belong to different publications.

quadibloc's picture

I definitely think that Williams Caslon Small is a beautiful typeface, and the change is definitely an improvement from a readability standpoint.

However, Nick Shinn is also correct in noting that the New Yorker aims at a sort of Art Deco look. Caslon still fits that better than a lot of other typefaces would, so it's not a total miss, but there indeed likely could have been better choices.

Or could there? When I try to think of an Art Deco face, what comes to mind are faces like Kabel, which are sans-serif, and not ideally suited for body copy. Still, Sabon did seem to fit the "look"; could its readability have been improved, or could it have been replaced by a different, slightly bolder, Garamond (as well as adding one point of leading) with a similar result from a readability standpoint?

William Berkson's picture

Thanks, all, for the kind words.

A little context for my Anglo-Canadian friends, Nick Shinn and John Hudson.

First of all, my scan is only of one corner of the page. Much of the page is text, so the decorative elements of the new version of Irvin and the drawings function as more of a punctuation on the page, and dominate it less than in my scan.

Second, some historical context. When The New Yorker was founded in 1925 Rea Irvin, their first art director, drew the whimsical all caps titling face they use (based on contemporary lettering, as the video indicates), and chose the then new (1923) Linotype Caslon Old Face for everything else. Ever since then, The New Yorker has stayed loyal to Caslon, and their unique display face, which they now call Irvin.

The look Irvin created paired the whimsy of his display face with the sober but amiable Caslon. The two typefaces are indeed quite a contrast. But it is a contrast that has visually branded the magazine for 88 years, and is loved by its readers, who now number over a million every week. It may be a look that has lasted the longest of any magazine still in print. The freehand whimsy of Irvin also is a visual bridge to the many freehand illustrations in every issue, and its famous one-panel cartoons, always captioned in Caslon Italic. The two contrasting feelings of Irvin and Caslon somehow manage to visually work with both serious reporting and weekly short stories from the most renowned US writers, as well as humorous pieces and the cartoons.

So the current redesign wanting to stick with Caslon is certainly neither odd nor unhistorical. In fact, their having used Sabon for the listings for a time was, in my view, the odd and unhistorical choice.

As to the redesign, I had nothing to do with it except completing my Caslon Small for the listings. If anyone asked me, I'm sure I would offer tweaks, including what Karsten suggests, spacing out the Neutraface Caps. And though I am a great admirer of Neutraface and of Christian Schwartz's work generally, I'd rather see it be all Caslon and Irvin, as in the original. (I recently looked at editions from 1929-30, which I found quite wonderful.)

However, I still think the current redesign works quite well. It also serves their new design purpose. Since everyone can check listings on their smart phone, they wanted to emphasize their beautifully written cultural critiques and commentary, and the new more continuously readable look serves that.

Finally, a word on Art Deco. To me, the interwar period in the 20th century is America's best, as far as decorative arts, along with the neo-classical period of American Federalism. It is also very characteristic of the best of New York City style. Maybe it is a weakness that I like it so much, but if so it is a weakness I share with a lot of my countrymen.

Nick Shinn's picture

@Bill: The New Yorker has stayed loyal to Caslon…

The how come it’s what appears to be a Garamond (Sabon?) in the “old” image above?

@Bill: …chose the then new (1923) Linotype Caslon Old Face…

Thanks for this information.
That Caslon revival certainly relates to the historicist era of American design, which is when the magazine originated—and the Regency dude. There was too an overlap with Art Deco. So the choice of Caslon is indeed appropriate to the magazine’s historical branding. However, I do feel that the Linotype Caslon had a bit more style than any phototype/digital versions since. The present setting is cramped and grey in tone, whereas the Old Face had space and sparkle:

Disclaimer: Like Bill, I have a great love and respect for American graphics of the early 20th century, but I am a bit of a purist on historical matters, and prefer to see old typefaces set in the manner for which they were designed—in the case of Caslon, that would be with a little bit of letterpress bounce, no kerning, and most importantly, with no extra character spacing in justified settings.

dberlow's picture

Really Nick. That's ugly. I'd burn it.

William Berkson's picture

David, this old New Yorker page looks unbelievably better in reality than in the scan. I'd explain why, but you probably wouldn't believe me, because I wouldn't believe me.

Some things, like the very wide word spacing, lack of kerning, and very loose and wide italic are a result largely of Linotype limitations, and I certainly don't believe there is any virtue in being 'pure' about it, as Nick recommends.

quadibloc's picture

The scan is rather low contrast, just for starters. And having seen it, I remember what the New Yorker used to look like. So I believe you.

jasonc's picture

Wonderful work, William. Congrats.

Nick Shinn's picture

Bill, the lack of character spacing in that old specimen is not due to technological limitation, but typographer preference. I don’t know about Linotype, but Tolbert Lanston originally planned to have character spacing in the Monotype machine’s justification and figured out how to do it; however, in tests people didn’t like the way it looked, preferring word-space adjustment only.

Quark XPress is largely responsible for the look of justified text today, by making character spacing part of the default H&J setting.

@Bill: I certainly don't believe there is any virtue in being 'pure' about it, as Nick recommends.

I didn’t recommend, I preferred.
Being a purist in matters of style concerns taste, not “virtue”:

Manny, yes, our tastes differ. Really.

William Berkson's picture

Thanks, Jason!

The wide word spacing in Nick's scan comes, as I understand it, from a combination of the way the Linotype machine operated, and the pressures on compositors. As the wikipedia article explains, the Linotype machine put a wedge between words, when compositor hit the space bar. Then when a line was complete, the machine would in push the wedges to expand the spaces to fill out the line and justify it. A lazy compositor could do less hyphenation, and just let the machine justify the line by shoving the word spaces out wider. This is the kind of thing that Geoffrey Dowding excoriated in his 1954 book Finer Points in the Spacing and arrangement of type. In the subsequent phototype era, the pendulum swung the other way.

The wide letter spacing in Linotype Caslon Old Face was a deliberate choice by the designers at the Linotype company, and was probably determined by Chauncy Griffith, who was overseeing the production of fonts at the time. The letter spacing may be a little wider than the original Caslon Long Primer typeface, but the original letter spacing is also very wide compared to what is common today – at least judging by the sample in Caslon's 1766 specimen book. Most likely it was Griffith's attempt to be 'pure,' in the sense of being true to the look of the original, as much as he could, given the limitations of the Linotype machine.

dberlow's picture

Bill, thanks, I know what the New Yorker looked like in hot metal. As one went up to books back then, I liked it more. As one went down to newspapers, I liked it less. It was a purely economic issue, never really in the hands of typographers.

Now, I like it that they all have to same wide choices, and that none of them have the taste for what could,/em> be accomplished today to make it look like it used to, but won't. Making decent looking type rattle along the baseline on the screen is pretty hard.

Nick Shinn's picture

No, you can’t have a taste for something which doesn’t yet exist, but there is always an interest in non-slick type; for display typography at the moment.

Rough typography—sometimes an attempt to humanize corporate marketing, other times the expression of a high-touch sensibility to humanize the high-tech lifestyle—is not to be confused with imitating the past.

As screen resolution increases, rattled text type, or some such thing, will become more feasible.

johndberry's picture

Nick –

Quark XPress is largely responsible for the look of justified text today, by making character spacing part of the default H&J setting.

That’s true. When I researched and wrote two technical white papers in 1991 comparing the composition capabilities of QuarkXPress and PageMaker (I did this for the Aldus Corporation, maker of PageMaker, but it was a technical comparison, not a promotional piece), looking under the hood of both programs’ composition engines, I discovered that, no matter how you set the defaults, QuarkXPress would invoke letter-spacing too early in the decision tree of its H&J. In PageMaker, if you set all the letter-spacing values (Minimum, Optimum, Maximum) to zero, the only time you’d see any letter-spacing to justify was when a line contained only a single word. (At that point, something has to give.) QuarkXPress’s H&J routines would invoke letter-spacing even when you tried to forbid it.

There was a time when the New Yorker first started using QuarkXPress to set its type when you would see horrendous letter-spacing throughout the narrow typeset columns. They fixed the defaults within a few issues, but the legacy of that letter-spacing still lives in the New Yorker today, two decades later.

Syndicate content Syndicate content