It's more beautiful than Legato, for sure.

Text is not about beautiful letters.

Bravo David.

hhp

Of course the obvious problem with the comparison is that one is a serif and one is a sans.

Is that a problem? The question is about whether the innovative detachment of the two sides of the stroke is beneficial compared to more traditional serifed types (in Turnip's case) or more traditional sans (in Legato's case). It's the same question. Though sans are traditionally less influenced by writing to begin with.

interesting question

I think this is just a failure of terminology rehashed over and over again, that everything without serifs is one big general class of type.

For what it's worth, here is what was going through my head while drawing Turnip:

I think of Bookman and the late-19th/early-20th century "oldstyle antiques" as faux oldstyles: the serifs/terminals are kind of calligraphic-ish, the structure not so much. The inner and outer shapes do virtually the same thing; the diagonal stress is gone, and with it, the dynamism that you get from the calligraphy-inspired thick/thin relationship.

Seeing how Bookman feels static feels in text, I figured that a sense of movement within the letter shapes, especially horizontal movement, has value in a block of text. Not sure if it's about making a text more readable, but maybe more compelling? I wanted that kind of movement in Turnip, like what happens naturally to shapes rooted in calligraphy, but in a more blatant, non-calligraphic sort of way. So I tried pitting inside against outside in hope that the tension would make each character look like it was unable to sit still.

I wrote a little more about the drawing on my personal site, if it's relevant: http://djr.com/typefaces/turnip/#story

The screen rendering is great too.

BTW, with the Regular and Book being so close in weight, did you consider making them uniwidth?

hhp

David, what do you think about the idea that the letters are to some extent constructed of oval outer shapes and more squarish counter shapes? I mean, did it just kind of happen organically during construction, or was it part of what you were trying to do from the outset?

Ryan — Did you read the backstory that DJR linked to? I think that basically answers your question:

I accentuated the tension between round, doughy outer forms and crisp, angular inner counters.

. . . if you interpret “round, doughy" as more or less oval and “crisp, angular” as squarish, that is.

Not mentioned in that piece however is the ghost of Dwiggins, swirling all around us...
http://www.flickr.com/photos/ohbendy/5943111278/

hhp

those 2 arent the same thing. Tension can exist in many other formats other than circle outside square inside. But if someone sets out with the idea beforehand, 'and now I will draw something with roundish outer shapes and squarish counters.'...

What is interesting to me about this design, with regard to Peter's original question, is the way in which the results of David's approach differ optically from those of Evert's in Legato. This is most obvious at larger sizes, where edge perception overrides shape perception, but it leads me to think that there is probably a mid-to-large size range at which Legato is more readable than Pumpkin. The difference I perceive is that at larger sizes the counters of Pumpkin start to pop out, especially in the heavier weights, whereas in Legato the relationship of inner and outer shapes aims for a less dynamic tension and maintains a strong sense of black shape structure even as size and weight increase. As David says, the effect of his approach is to 'make each character look like it was unable to sit still'. Is that a good goal in terms of readability? I don't know. It is roughly the opposite of my own approach, which is to create a sense of stability, but then I am often dealing with writing systems that have built-in instability of various kinds.

I think comparison to other new-calligraphic types, such as, Legato, and dare I say it--Fenland, are just. That is not meant to demean Turnip in any way whatsoever. But there seems to be a new school of calligraphic-geometric being formed. It's the new black.

John, you were making pumpkin pie right before that post, weren't you... ;-)

Seriously: Great observations - and David's "unable to sit still" does make me... fidgety! But at least it's a worthy challenge to the status quo.

Ryan, the new school (still a rag-tag bunch of guerrillas) is actually specifically non-calligraphic. It's not the new black yet, but things are looking much better than they did just a few years ago.

hhp

I have recently encountered one classic example of square inside round... AdLib.

Er, yeah. Turnip, not pumpkin. Could have been worse. Could have been rutabaga.

did you consider making them uniwidth?

I suppose I was thinking about Book and Regular as being less like true grades (that solve a technical problem) and more a stylistic choice. But that's post-rationalization, really...it's a good question, and if a user ever needed compatible width versions, I would be happy to create them.

Ryan: But if someone sets out with the idea beforehand, 'and now I will draw something with roundish outer shapes and squarish counters.'...

Inner/outer play wasn't a real factor in the original drawings...it crept in as I was looking for ways to make the design pop a little more in text.

Hrant: Not mentioned in that piece however is the ghost of Dwiggins, swirling all around us...

Definitely...he was all about this sort of un-calligraphic dynamism.

John: Is that a good goal in terms of readability? I don't know. It is roughly the opposite of my own approach, which is to create a sense of stability, but then I am often dealing with writing systems that have built-in instability of various kinds.

That's a good question. :-)

I guess that's sort of the thesis of this typeface, that there's something interesting in running contrary to common traits of text face design: balance, stability, evenness.

The strange thing about Dwiggins though is that -as far as I know- he never explicitly verbalized his strong un-calligraphic leanings - in fact he often extolled the pen! Or maybe his thoughts have been "filtered for publication"... But probably it's a case of a designer with incredible inventive intuition - so strong that his formal intellect could not keep up to explain what was going on. In fact some of Dwiggins's formal statements can only be described as ludicrous, such as the view that extenders are only for show, with no positive role to play in readability!

running contrary to common traits of text face design: balance, stability, evenness.

Which must make sense at least on some level, since information can only arise from contrast.

Do you remember that slide from my Mexico City talk (that you requested) where I showed two opposite "poles"? One thing I was trying to say there was exactly that the textbook Modernist desires for things like balance, stability and evenness belong towards the display end of the axis, not the text end (although neither extreme of the axis can exist in a pure state).

hhp

Hrant: Which must make sense at least on some level, since information can only arise from contrast.

Contrast, in the sense of differentiation, isn't at odds with balance, stability or, even, evenness in design. You don't need a building to be unbalanced and off-kilter in order to tell where the door is and how many windows there are.

With regard to differentiation: in my 'Same Difference' lecture at St Brides a few years ago, I looked at kinds of sameness that occur in writing systems and the job of differentiation that the type designer must do, and at the kinds of difference that occur and the job of 'making same' that the type designer must do. On the whole, I reckon we are given more difference than sameness to work with, so the bulk of our work is in creating kinds of sameness that do not decrease the necessary sameness but that produce visual harmony and stable relationships between signs.

The strange thing about Dwiggins though is that -as far as I know- he never explicitly verbalized his strong un-calligraphic leanings - in fact he often extolled the pen!

WAD essentially thought of what he was doing as “calligraphic.” Doesn’t necessarily fit with most of today’s practitioner’s definitions. But that’s how he thought about it.

In fact some of Dwiggins's formal statements can only be described as ludicrous, such as the view that extenders are only for show, with no positive role to play in readability!

I don’t think that Dwiggins was the only designer of his time who might have thought like this. Victor Hammer, for instance, believed that ascenders and descenders slowed down reading. He preferred Latin to any of the modern European languages he was familiar with, as it features much less of them in text. Not that Dwiggins and Hammer worked under similar circumstances, but … there you go.

Kent, my contention is that Dwiggins's thinking was out of touch with his doing.

Dan, even though Hammer's belief is more extreme than Dwiggings's, to me it's not as shocking because we don't hold up Hammer's work as aiding readability while we do do that for the M-formula. I think Hammer created amazing display type, while the most amazing thing about Dwiggins was his un-calligraphic output.

hhp

I think Hammer created amazing display type.

In fact, all of Hammer’s typefaces are text typefaces, as far as he was concerned (except perhaps Andromaque). He certainly used them for book printing … quite extensively so. Hammer’s work is certainly calligraphic though … he was in the Edward-Johnston-line. Our association with his work as being in the “display” arena is probably because of his stylistic unorthodoxy, and also because of the terribly poor photo and digital versions of American Uncial. This, though, has nothing to do with David’s awesome typeface! Sorry for the tangent.

I brought Hammer up because I think that our current ideas around reading might only have come about 70 some-odd years ago. Before these ideas got to designers and foundries, other things might have been going on. I suspect that there were plenty more type designers who thought like Dwiggins or Hammer might have.

I know Hammer believed his fonts were for text, but most people now don't (and I believe our aggregate views do become more refined over time - although dips are unavoidable).

We do know that designers follow trends, and I think this applies not just to style but to thought as well. Morison convinced Dwiggins and JvK to go slanted-Roman*, but then they changed their minds. And as you say I'm sure the anti-extender stance of Dwiggins and Hammer were not isolated views. But is it true that extenders have only gotten respect in the past few decades? I think there were periods in the past where people realized they're important functionally.

* Although not the right way.

hhp

the new school (still a rag-tag bunch of guerrillas) is actually specifically non-calligraphic.

But it sooo is though. I remember Nick Shinn talking about how Fenlad couldn't be constructed with a nib pen by only one hand, but needed both left and right or something like that. And I know the Tankard specifically states he was trying to think in a way that could never been done with a nib pen, or something like that, but to me he failed utterly in that. You can't tell me those curves (curves, not shapes) dont come directly from classical penmanship. For instance, fill in the counter out of the |o|, and look only at the outside shape, that doesn't look calligraphic? Or something like that.

You can't tell me those curves (curves, not shapes) dont come directly from classical penmanship.

Uh, that's what I've been doing, sorry. Or are you talking about the outlines being chirographic? Virtually nobody does that. Read this though: http://typophile.com/node/31095

You can create any shape with a pen, if you contort it enough. But does it make sense to go to the grocery store crawling on your back, naked?

You do have to take into account third-party interpretation though, and (as with Dwiggins) sometimes even the designer can't fully wrap his head around the results. Like how some people -who should know better- insist on seeing chirography in Legato. To me that's just ridiculous apologism.

hhp

Hrant: Do you remember that slide from my Mexico City talk (that you requested) where I showed two opposite "poles"?

I do! One of the dichotomies that most spoke to me was "expression" (for display) vs "service" (for text). Related to the original post on readability, I struggled with how Turnip should fall on this spectrum...namely how prevalent the inner/outer play should be.

John's right that the inner/outer play can take over at mid-to-large sizes, but I was mainly interested in text size (print and screen), where the effect is more subtle but there's still a little kick to it (M-formula sort of stuff, I suppose).

I can't speak for the goals of the other designers mentioned, and I can't say that this "kick" improves readability across the board. But for Turnip, I had a certain kind of use in mind. When you write with dialect, for example, it's not about directly conveying information but about cultivating a reading experience. Similarly, I consider the "expressive" aspects of Turnip to be part of "services" rendered.

Kent, my contention is that Dwiggins's thinking was out of touch with his doing.

Yes, I’m well aware of this. And my contention is that your thinking about Dwiggins’s doing is out of touch with his thinking about his doing. You have conflated your “chirography” with his “calligraphy.” Dwiggins was not Johnstonian.

We all see the past through our own contemporary lenses. What inspires you about Dwiggins’s experimental work may or may not be what inspired Dwiggins. Which is fine: we take our inspirations where we find them. But our interpretations do not in turn define the objects of our inspiration nor the motivations of their originators.

because we don't hold up Hammer's work as aiding readability while we do do that for the M-formula.

Again, to be clear, you may hold up [your interpretation of] the M-formula as “aiding readability;” but if you study the original formulation carefully, you will find that that was never Dwiggins’s objective, certainly not in any sense that you like to employ that phrase.

You may see potential in the results of his explorations that inform your own particular thinking and that you wish to build upon, which is certainly a valid prerogative; just be mindful that your aims are not necessarily the same as Dwiggins’s pursuits.

If his results coincide with your thinking about Readability, it may be just that: coincidence.

I must be expressing myself poorly, because I actually agree with almost everything you just wrote. I'm not nearly as informed about Dwiggins as you are, but I do understand that his M-formula idea does not directly support my ideas on readability (and I'm certainly not trying to co-opt Dwiggins's prestige).

I guess what I'm saying is that Dwiggins -even though I think he was a genius, and I never use that word lightly- didn't realize all the ways that his ideas could be useful (which is normal for anybody); and one of those ways strikes some contemporary designers (and not just me BTW) as more significant than anything he seems to have explicitly thought about: the breaking of chirography to benefit functionality. Was that dumb luck, or killer instincts? We can't know - but faith makes me prefer the latter. And if he didn't even realize he was being un-chirographic*, oh well - at least he gave us a wonderful gift.

* Or un-calligraphic, which in spirit at least is quite close.

hhp

I'm not a font designer, and so have a different starting point here. I looked at the Font Bureau's specimen (printed out) and I think the Regular size, shown with bold and italic, looks like a very readable text font for extended text. If you agree with this, please tell me why it is so. To my eye, at this 9 point size, the square insides are not noticeable. It just looks like a nice low contrast font. So is this a nice font because of the square insides, even though I can't see them? Or is it because of the big (wide) counters? Or some other reason. This is the way my brain works. Look at some type that works and figure out why it does. So... why?

Off topic #1: I have ranted in the past that italics need to have the same darkness as the roman. http://typophile.com/node/77718 And Turnip's italics look good to me.

Off topic #2: Type specimens should show the text fonts at the size they will probably be used! Maybe the designer thinks they look best for long text at size 9 with pretty tight leading, but I am skeptical that a novel or textbook would have them this small. I think the designer should be forthright and say that, for instance, the font looks best at 10.5/13.1 and the specimen should be set this way. I shouldn't have to look at small text and try to guess how it will look at a usable size. And I would prefer unjustified text so I don't have to guess at how it looks straight out of the can.

If you like Turnip, you might want to take a look at Coline: http://www.emilie-rigaud.com/index.php?/project/coline/

Edit: What do you think about the capital letters like T and F not letting the lowercase letters slide in close? I wish the specimen had shown a few of these. My guess is that it looks just fine, but I wish I could see it.

even though I can't see them?

There are different kinds of seeing. Rest assured that your inner mind can see things that are "subvisible"; and everything matters.

italics need to have the same darkness as the roman.

It depends. Not all Italics can/should have a strong slant, and/but they cannot fail to work for emphasis in a body of Roman (even when just an "I" is emphasized, which is one reason so-called "upright Italics" can't really work) so sometimes you might elect to lighten the color. Or make it look smaller on the body (rarely done). Or both. Or any combination of the various factors above...

I would prefer unjustified text

Indeed - there's no gauging the spacing otherwise.

Coline BTW is quite interesting - I hadn't noticed it before.

--

BTW, something that's bugging me:
In the screen version of Turnip they reduced the characteristic tension between the insides and the outsides. Does this mean the screen impedes un-chirographic design? Or maybe it's that we're not treating the screen right? I'm not sure.

hhp

Who is "they"?

I must be expressing myself poorly, because I actually agree with almost everything you just wrote.

Okay, yes Hrant — from your restatement, it sounds like we’re on the same page. My apologies if I misread or over-reacted earlier.

I shouldn't have to look at small text and try to guess how it will look at a usable size.

J. Tillman — Yes, this is an unfortunate side-effect of having certain standardized showings.

And I would prefer unjustified text

Indeed - there's no gauging the spacing otherwise.

I’ll let you in on a little secret: FB text specimens are carefully composed so that the justified text is very close to the natural word spacing of the font. Shown here is the justified Turnip setting (left), compared with the same in a flush-left setting (right):

Who is "they"?

Uhhh... yeah. :-)

FB text specimens are carefully composed so that the justified text is very close to the natural word spacing of the font.

Good to know!
(Sounds like how I write my emails - and used to write my Typophile posts.)

hhp

DB: I think "they" is me. :-)

Hrant: The RE styles are quite different, but I don't remember consciously reducing the inner/outer tension. What you are seeing may be a factor of weight: in Turnip (print and screen), the inner/outer tension gets louder as the design gets heavier and there is more allowable variation in thickness. TurnipRE-Regular is lighter than Turnip-Regular and even Turnip-Book.

It also may be affected by how you are seeing the RE fonts rendered. A preview of this and other Reading Edge fonts across various renderings is here: http://www.fontbureau.com/ReadingEdge/

J Tillman: I agree that the Font Bureau PDF specimen shows the font on the smaller and tighter end of Turnip's range. If you are interested, I would be happy to send you supplemental materials. Since it appears I can't attach PDFs to this thread, please get in touch: http://djr.com/about/contact

I'm not the expert at making these specimens, but I do know that care is taken to show the font without a lot of primping, so you are seeing default spacing (or something very close to it), despite the justified setting. [Edit: I started responding, and then got breakfast and missed Kent's post. See his explanation above.]

I'm basing my comment on the notes on the "n"s here:
http://imprint.printmag.com/wp-content/uploads/turnip_comparison.02.png
It does say "accentuate[d]" and "preserve", but since the original (or at least the one that generates pertinent discussions like this thread) is the print version I think instead those really mean "preserve" and "dilute" respectively... Like how the "Classic" model is old and slow, the "Faster" and "Better" settings could be called "Crappier" and "Slower", and "widescreen" often ends up meaning "shortscreen"... :-/

About that "they" again: I'm glad to hear you did the mods yourself* but you're not the [only] one who nailed down the parameters for good onscreen rendering. In fact in a way even Microsoft and Apple could be seen as part of that "they".

* Not because I wouldn't trust third-parties (especially at FB that's a feature, not a bug) but because it means you know how to do it.

Breakfast? Well OK, if you must. But remember, "Lunch is for wimps." ;-)

hhp

I'm basing my comment on the notes on the "n"s here:

Ahhh...yeah my shorthand here is confusing. What I should have said is this: "Describing the design in general, Turnip's inner-outer tension is accentuated." ...and then: "Here, in the Reading Edge version, the accentuated inner-outer tension is preserved (i.e. not diluted), even as other things have changed."

I think the inner/outer tension feels about right when viewed in the context of other weights (with adjustments made to the RE font's scale and spacing, to compare).

About that "they" again: I'm glad to hear you did the mods yourself* but you're not the [only] one who nailed down the parameters for good onscreen rendering. In fact in a way even Microsoft and Apple could be seen as part of that "they".

I get what you're saying. I do think those parameters are still being nailed...it's not like there's a consensus or list of rules, though certainly there are attributes in common.

At least with Turnip RE, this was less a list of modifications and more of a reimagining of the design for a different medium. Being familiar with Verdana and the earlier members of the Reading Edge series, I felt free to deviate from the standard design where I wanted to (like those changes listed in that comparison, and others), or to make no changes when changes weren't necessary (like thick-thin contrast, which you'd expect to get much lower in a serif for screen, but wasn't necessary in Turnip).

J Tillman: What do you think about the capital letters like T and F not letting the lowercase letters slide in close?

While I had InDesign open, I made this image of F and T words compared to other cap-lowercase combos, for you to see.

In the screen version of Turnip they reduced the characteristic tension between the insides and the outsides.

Isn’t it a bit misleading to be looking at outlines anyway? I mean, how much “tension between insides and outsides” can you display in the space of 7px high when the stem is essentially a single px weight?

But if you can't display it, why reduce it in the outlines meant for the screen? Or even: might you not exaggerate an important characteristic of your design to make sure it shows up on the screen?

(David, thanks for the elaboration.)

hhp

But if you can't display it, why reduce it in the outlines meant for the screen?

First of all, DJR claims that it hasn’t been reduced in the outlines for Turnip RE.

So that aside, in general one might reduce idiosyncrasies in outlines for screen (i.e., from those meant for hi-res, albeit small size, in print) in order to better manage the practical rasterization at various small sizes.

might you not exaggerate an important characteristic of your design to make sure it shows up on the screen?

My point is that certain characteristics can’t show up on the screen — such as we have now, anyway — not for the RE targets.

So, to go back to one of your earlier questions —

Does this mean the screen impedes un-chirographic design?

— I suspect the answer may be [currently] Yes.

Which would be perplexing, considering it's made up of dots and not strokes...

Maybe instead it's that the screen's coarseness limits the inclusion of useful details.

hhp

Squared counters can reduce noise on screen. Quite a bit, in fact. Here’s an /o/ and a /g/ I once drew as a screen experiment, albeit without rendering samples. Squared white + diamond shaped black.

really, only the bottom loop of the g looks very square to me. I mean i know its there elsewhere, but its not exaggerated enough for people to take notice. IMO.

Talking about Frode Franks post

Squared white + diamond shaped black

Giambattista Bodoni (and Morris Fuller Benton) say

its not exaggerated enough for people to take notice.

Which, in a text face, is exactly a feature and not a bug.

hhp

@enne_son: Not had time to read the entire thread, but it might help others participate in the discussion if you defined terms and phrases like “systematic violation of (expansionist) chirographic logic” and “deliberated dissociation.” Perhaps it’s just me, but I have always been a fan of Plain Language. Especially in academic writing.

Christopher, I'm assuming that most readers of these forums are familiar with Gerrit Noordzij’s term expansion. Noordzij uses this term to characterize the contrast scheme typical of the pointed nib pen, and some readers will know that the contrast in Bookman is of this sort.

Those who've read the links provided will probably connect readily with the ‘dissociation’ comment.

Perhaps I could have elaborated more in my opening post, but the thread has taken off nicely, including posts by the creator of the font.

Peter

I don’t think the issue of the chirographic relationship between inner and outer outlines (i.e. how difficult it would be to write a font glyph with a pointed nib pen) is particularly relevant to serifed, contrasting stroke typefaces.

You might just as well discuss their “punchographic” quality.
Fred Smeijers did something like that in Counterpunch.

The Stroke is a useful paradigm for analyzing type structure with words, and type designers may certainly use it as part of the conceptual scaffolding of a face, but the relationship between stroke and counter in serifed, contrasting stroke type plays a minor role for the reader, entangled in a complex relationship between glyph parts.

Most letters (18 out of 26 in the English lower case) don’t have enclosed counters, and serifs and terminals are more instrumental in determining negative space than the stroke.

While it is easy to see, and postulate, a relationship between the two elements of stem shape and negative shape in a sans face, it’s almost impossible to see this relationship in a contrasty serifed face, because stroke shape is confounded with serif shape, terminal shape and stroke contrast, and obscured by optical cheats, such as the heavy bottom of Times’ /e, that further divorce the typographic stroke from one that might have been made by a simple broad-nibbed pen held at a consistent angle.

Nick, I have to say that makes no sense to me, on a number of fronts, such as: serifs can be chirographic too (in fact that's their origin!); a counter doesn't have to be fully enclosed to be relevant as a counter; Times is much more chirographic than Turnip; you don't have to hold the "pen" at a consistent angle to make chirography...

hhp