What does e.g. "<typeface name> 11 on 14" mean?

mars0i's picture

Sorry to ask this--it's surely a FAQ, but the nature of the question makes it impossible to find existing answers with normal search functions!

More than once I've seen something like the following in a copyright page or colophon. This example is from a copyright page:

"Text set in Janson 11 on 14."

What does the "11 on 14" mean? I believe I recall seeing "n point on m point" or something similar, so I now believe that the numbers are point sizes.

(Wild guess: The first number is the font size, and the second specifies the amount of leading. Maybe "11 on 14" means that there are 3 points of extra space?)

eliason's picture

Your wild guess is right.

David Vereschagin's picture

Also written shorthand as 11/14.


Joshua Langman's picture

Short for 11 point type on 14 point body. From the days of metal type, you could cast a character at a certain size on a body (base) of a different size. Then came to be used even when referring to leading added between lines cast on normal bodies.

mars0i's picture

Excellent. Thanks guys. This has been bugging me for a few weeks. I particularly appreciate Joshua's elaboration. Easier to understand if you know the origin. I hadn't heard of larger body technique.

Té Rowan's picture

As an aside, the shorthand notation exists in CSS where you can say:

P { font: 11pt/14pt "Droid Serif", serif; }

Joshua Langman's picture

I worked recently in a Monotype shop where they were doing a project that required 1 pt of leading between lines. In metal type, there's no such thing as 1 pt strips of lead, so they cast 10 pt type on a 9 pt body, and used 2 pt strips of lead. Yes, that means the characters slightly hung off the edge of the body.

kentlew's picture

> In metal type, there's no such thing as 1 pt strips of lead,

Really? Do you mean, perhaps, that they didn’t have any in your shop, or that they couldn’t be used with Monotype (for some reason I don’t know about)?

Here’s the relevant page from the 1923 ATF catalog offering 1 point leads in assorted lengths (apologies for the quick and dirty phone snapshot):

> Yes, that means the characters slightly hung off the edge of the body.

Incidentally: not possible with Linotype, only Monotype (and foundry, I suppose, but I would think it rare to be cast with vertical kerns).

hrant's picture

> there's no such thing as 1 pt strips of lead

Yeah, but in other materials (like brass) there is.


Joshua Langman's picture

In your picture, you'll notice it advertises "brass leads," an interesting oxymoron. The shop I was in cast their own leads — out of lead, not brass — and it was explained to me that you couldn't cast the lead that thin.

eliason's picture

Another reason for different metals is to make it easy for the compositor at a glance to recognize which one to reach for.

kentlew's picture

> The shop I was in cast their own leads

Ah, okay. That explains it. Makes sense now.

Yeah, “lead” as a term for spacing material is only partly related to lead, the metal.

Some were made of type metal (although a softer alloy, I think, so you could more easily trim to fit — but I could be wrong), and type metal, in turn, is only partly composed of lead. But 1pt leads were typically made from brass.

I don’t know that it’s so much an oxymoron, as a bit of a misnomer.

Or maybe some of both.

5star's picture

Alright, I'm confused. I snipped a diagram from wiki metal type and added in red what I think is what you're saying...

...is this right? Could that diagram mean 11pt on 14pt, with the 14pt meaning the body or shank and the 11pt the face?


eliason's picture

The diagram is wrong: the extra space would be in the "vertical" dimension, not the "horizontal."

hrant's picture

Neil, the height has nothing to do with it.
Leading is by definition added between
lines of type to make them further apart*.
There's such a thing as "internal leading"
(which I call "talus") but that's really part
of the design, not something a typesetter
handles directly.

* In the digital world it
could be negative too.


Joshua Langman's picture

Although in metal type, a lot of things that are now the designer's decision, like the size of the word space, the amount of "talus," where the character sits in the em square, etc., were determined by the person casting the type or by the machines, not by the designer.

I do remember that the Monotype shop could cast, or get from somewhere, 1.5 pt leads, which they colored purple to differentiate them. Probably the reason for not using brass leads is that if everything's made from type metal, you can just dump the whole mess in the cauldron when you're done printing — no sorting required.

hrant's picture

> were determined by the person casting the type

Indeed, good point. What you describe above is
such a case - where essentially the talus is negative.

Like if you look at Gill's final drawings for example,
they don't even have overshoots for the round parts!
The Monotype works boys did all kinds of refinements
behind the scenes.


5star's picture

Oh ok, gottcha. So the font is 11pt and the leading is 14pt.

hrant's picture

Close. :-)
The leading is 3pt. The linespacing is 14pt.


Joshua Langman's picture

InDesign disagrees. But you knew that.

hrant's picture

I guess I didn't!


mars0i's picture

I didn't know that metal type was still in use. Books I'd read gave the impression that it was a thing of the past. Is metal typesetting common, or is it just used in small specialty presses, or ...?

JamesM's picture

> InDesign disagrees

Yep. The words may have different definitions, but in common usage "leading" is often used instead of "line spacing".

I heard a dictionary editor say that when 2 words (or phrases) have similar meanings, the shorter one often becomes dominant, and this might be another example of that.

Té Rowan's picture

That dict editor had more ’n a skerrick of sense, klar.

Note: It may be that the iD programmers chose 'Leading' over 'Line spacing' so they could push the selectors a bit closer together in hope tthat the user could grasp both in one glance. Feh. Reckon this would have worked better:

Size -- [11pt X] / [14pt X] -- Line spacing

Joshua Langman's picture

Yes, "leading" in layout programs means something different than leading in metal type.

Té — I had to hunt around the program to get a screenshot with the fields labeled. In the normal menu bar, it just stacks the two of them with no labels, but you are literally typing one number "on" (top of) another.

mars0i — It's still very much in use (or so I'd like to believe), but you're right that it's only by small specialty presses. It now belongs, along with the letterpress printing processes, more to the realm of fine art than the world of mass production. The shop I was at, though, did get some big projects, like the official new edition of the Book of Mormon, which had to be typeset EXACTLY like the first edition, including the following word division on the title page:


Try to get away with that today.

Florian Hardwig's picture

Leading vs. line spacing, illustrated by Nick Sherman:

William Berkson's picture

Thanks for posting the link to Nick Sherman's illustration. Nick hopes to correct the usage, but I think that train has already left the station. JamesM, the theory of your dictionary expert is interesting, and sounds right for this case.

hrant's picture

> when 2 words (or phrases) have similar meanings,
> the shorter one often becomes dominant

Indeed. I also see this when languages are mixed. For example
in Armenian the word for "thank you" is six syllables! So we
say "merci"... :-/ And I've caught myself saying the Armenian
"mintchayt" (which means "while in reality") while speaking
English; my sister uses "Ya'ni" (Arabic for "What I mean is",
or "I mean, come on, really!") in the same way.

Something else: a single syllable is ofter too short,
especially in proper names, which acquire a diminutive
suffix (as an endearment). When speaking Armenian for
example "Tom" often becomes "Tomig".


John Hudson's picture

All this is complicated further if one sticks one's head inside the OS/2 table and notes that the term linespacing is there used for what is in effect built-in leading.

InDesign's use of 'leading' is specifically a baseline-to-baseline measure; that is, from the baseline of the current line of text to the baseline above.

mars0i's picture

I don't think I could write the way that I do without a computer, and I like the control I get with LaTeX, so I guess I'm a fan of computer-controlled typesetting. I have not paid attention to how the books I read are printed. But I still like hearing that some continue to be typeset in metal. Thanks Joshua.

Té Rowan's picture

I think that Murphy guy was right: If it can be got wrong, sooner or later it will be.

washishu's picture

When I served my apprenticeship in a small jobbing print shop in the 60s, we had some—not many—1pt leads. They were very fragile and I didn't like them. I'm not sure where they came from as we didn't have our own mechanical composition and any large job that required such was sent out (although if we were not busy, the boss would sometimes have me handset a large job to keep me earning my wages; I recall once setting a car-purchase contract for our biggest customer. It was two full pages of foolscap, almost full measure, set in 8pt solid Granby, justified. It was a bitch and there were a number of pied lines due to the small size in a very long measure).

hrant's picture

Neil, looking for something else I found an old diagram
that might help you (although it's important to keep in
mind that terminology does shift around over time).

Type-sort Anatomy Diagram from H Carter / LeGros & Grant


5star's picture

Thanks Hrant!

So in Fig. 4. - Plan of type, 5.'Line-to-Back ,is today's line spacing? Or would it be 8. of the same Fig. 4., the Body?

I also noticed comparing the illustration I snipped from wiki to the one cite from Carter & Grant's that the Face's location is not the same. Which adds nicely to my confusion. And since the letter is located up from the 'bottom' of the stem shouldn't the Body be illustrated as having three segments - Upper Body, Face, and Lower Body? Or, Upper Shoulder, Face, and Lower Shoulder?

hrant's picture

No, "linespacing" is either the leading added between
the lines (so totally absent from those illustrations) or
the leading plus the height of the line of type* (so still
essentially absent from those illustrations). In any
case it pertains to a setting, not the font itself.

* So #8.

We don't use "line-to-back" any more. :-) If I had to talk
about that, I'd call it "ascender height plus upper talus"
or "em height above baseline"...

That illustration you included (in your first post of the 5th)
is indeed confusing because I don't think any font has [had]
caps that touch the top; even if the ascenders aren't longer
than the caps you almost always see some talus, especially at
the top (since descenders are rare sometimes they're allowed
to even hang off the bottom, to save vertical space).


Bert Vanderveen's picture

Leading, line-height, whatever… The super-rational Germans of Berthold initiated the term ‘transport' in early instances of their photocomposition equipment & that was a completely transparent term for what it is — defining the offset/distance to the next line.

As we are all children of technology, why not settle on ’transport'?

(Do not forget that transport can also have a negative value!)

5star's picture

Leading used with wood type.

hrant's picture

An elaboration:
> I don't think any font has [had] caps that touch the top

In a caps-only (metal) font this can happen.
But in that case there would definitely be
no space left below either - so that capital
"H" would span the entire height (or maybe
just leave a little bit of room for the tail of
the "Q" and/or "J").


mars0i's picture

I'm a bit confused about terminology. Is the point size = distance from lowest descender to highest ascender + talus? Or distance from baseline to capline + talus? Or?

A related question: If something is typeset as 11 point type without anything extra (11 on 11, so to speak) is there a standard amount of space that appears in addition to the distance from the lowest descender to tallest ascender? Or is there no standard--every type designer gets to choose what they think looks good? I'm assuming that that characters on different lines should never be able to touch, unless there is negative leading.


hrant's picture

> Is the point size = distance from lowest
> descender to highest ascender + talus?

Where the talus is usually partly at the top and
partly at the bottom (usually more at the top).

BTW, few people use "talus". Some
people use the term "internal leading".

> is there a standard amount of space that appears in
> addition to the distance from the lowest descender
> to tallest ascender?

There's no standard, but most type designers do seem
to try to play within the same talus ballpark, which is
on the order of 5% of the Em. But some designers do
overshoot the em!

One thing to keep in mind is that the talus
should be proportional to the x-height.


mars0i's picture

OK, so my understanding, then, is that without extra leading, the distance from baseline to baseline will be Em-height + talus, for a total of roughly 1.05% of Em-height, but probably more for a tall x-height, or less for a small x-height.

And then, suppose, for example, I look at a serif typeface in which horizontal serifs have flat outer edges--not cupped, for example--and supposing that the top joins of the M is finished with these serifs, rather than being pointed, for example. Then if an M appears over another on the next line, the distance between the upper edge of the top serifs on the lower M, and the bottom of the lower serifs on the upper M--that's the talus. However, ascenders might extend into the space defined by these two M's, and bottoms of curved letters might extend down into the space, as might the points on the V and M, for example, the bottom of the J, etc. And maybe if the typeface has very high ascenders and very low descenders, the designer would make the talus larger, I suppose.

Sorry for putting things in this awkward way. I'm trying to get a precise understanding, but I'm trying to work around the possibility that terms are defined in ways that I didn't expect for a given typeface. My understanding is that there's some play in the terminology. I've learned enough recently to be unsure about everything, but not enough to be sure about anything, I guess.

Do I have it right?


Joshua Langman's picture

"11 on 11" is called "11 solid."

hrant's picture

> without extra leading, the distance from baseline to baseline will be Em-height + talus

No, the talus is built-in to the Em-height.
So the distance from baseline to baseline = the Em.

The talus is basically the sum of the space
above the ascender height and below the
descender depth, within the Em.


kentlew's picture

Hrant sort of acknowledged this, but I feel compelled to point it out again: the term “talus” is a complete Hrant neologism.

Don’t get too hung up on it. No professional type designer that I know uses the term. Most won’t even have heard of it (and of those who may have encountered it on this forum, I can’t think of any who have chosen to adopt it). I don’t even know anyone who bothers to refer to “internal leading.”

In metal, the measure of the body of a piece of type was it’s point size (and vice versa). When set without any intervening leading — or, as it was called, set “solid” — the distance between baselines was the same as the point size.

[This is assuming, for the sake of keeping things simple, that this is a setting of the same font on all lines.]

For practical reasons, metal type faces did not usually have extenders that reached to the very top or bottom edges of the piece of metal. So there would usually be a little extra space, such that descenders on one line and ascenders on the line below didn’t crash. (Hrant has chosen to coin a special term for this extra space.)

For the Linotype machine, this extra space had a prescribed minimum. I don’t know if foundries or Monotype had their own standards for this.

In digital type, there is a virtual corollary to the body, but there is no physical constraint or requirement to keep parts within the bounds of the body. Different type designers take different approaches.

Also, since we're no longer stacking little metal blocks upon one another, the virtual body has come to have less and less technical significance. In Postscript Type 1 fonts, it was still pretty explicit (in the Ascent and Descent values). But with TrueType and OpenType, there are multiple values describing various vertical metrics, all used by different platforms and applications to determine vertical relationships.

So, these days, setting text without any explicit linespacing measurement is not necessarily the same thing as setting metal type solid and you may or may not get a value equivalent to the point size, depending upon the platform, software, and preference settings.

Of course, if you explicitly specify 11 pt linespacing (or leading — however the app chooses to call it), then your application should definitely set it with a baseline-to-baseline measure of 11 points.

hrant's picture

1) I didn't make it up. I wish I remembered where I got it, but I'm
pretty sure it was from a "professional type designer" :-/ sort of book.
It might have been DeVinne.

2) I use it because it's needed.* I've heard others use "internal
leading" but to me that's lame. I haven't heard of anything else.
Why is it needed? Well, it's a relevant design parameter, since it
affects apparent size and line-to-line not-crashing. When I make
a typeface, very early on I need to figure out the vertical metrics,
and the talus is part of that. How do others do it?

* Not to grasp the idea, but to communicate it.

As always, I'm open to a better term.
But I don't understand the resistance to having a term.


Té Rowan's picture

The term 'talus' make sense, sorta, it being Latin for 'ankle'.

kentlew's picture

Hrant — Sorry, you know I have a knee-jerk reaction to your jargon sometimes. I do remember now that you appropriated it from someone else. And DeVinne sort of rings a bell. So I’ll blame him for the failed neologism. ;-)

My point remains: anyone who attempts to use it in conversation with professional type designers will most likely find himself faced with blank stares.

I don’t dispute the value of the overall concept within a consideration of general parameters. I’ve just always engaged it from the other way around, such that there’s little point to referring to that space directly.

The concept that I’ve always used to approach this aspect of vertical metrics, and the one which corresponds to my historical meanderings, is “size on the body.” That’s how I’ve encountered it referred to historically. That’s how I think of it.

So, my approach to a new design idea is to think about the interrelationship of x-height, cap-height, ascender-height, descender-height, and overall size on the body (i.e., the sum of the ascender + descender and it’s relation to the em).

It seems really simple and straightforward to me — how do I want to fit my design in the em? How much clearance do I want between ascenders & descenders when set solid? Which is usually also a function of what I consider to be the target size of my design.

Since I usually also consider how I want to split this between the head space and the foot space (also terms I’ve seen used occasionally, btw), a single new term derived from Latin for ‘ankle’ just seems silly to me.

hrant's picture

Note that I always "asterisk" my first use
of "talus" in a given discussion. I do hope
though that one day I won't have to, just
like "bouma". :-)

> my approach to a new design idea is to think about the
> interrelationship of x-height, cap-height, ascender-height,
> descender-height, and overall size on the body

That can work, but don't you think it's missing
the concept of how to manage line collision?

Especially in terms of splitting the "blank" space
into top and bottom; for example it's nice to be
able to quickly opine that the top needs a bit
more space:

> the head space and the foot space

OK, now give me a good term for
their sum and "talus" is a goner. :-)


kentlew's picture


hrant's picture

Hey, not bad.


mars0i's picture

Thanks folks! I'm learning a lot.

(I got confused and thought that Em-height was the height of the M. I realize now that I learned at one time that it was the width of the M, and just checked Wikipedia for a quick update and learned that that is an obsolete though original meaning. Between that and the preceding discussion, I think it's clear now.)

Syndicate content Syndicate content