"Boston" Numerals?

hrant's picture

In De Vinne's "The Practice of Typography: Plain Printing Types", there's a sample of a face that's intriguing for its use of what are often called "hybrid" numerals:

That's actually a collage from a more extensive setting. The font is the #19 of the Phelps, Dalton & Co., and was probably made at the end of the 19th century. This makes it the earliest "true" hubrid numeral scheme I've encountered - in fact by far: the most recent ones I knew of previously were in Fenway (Carter) and some of de Groot's work. The Austin numerals (like the kind Carter has recreated in Miller) have some hybrid traits, but I find them too irregular (although in a nice way) to really qualify as a scheme.

Anyway, this might be an opportunity to find a better name for "hybrid numerals", and I propose calling them "Boston numerals", not just because of this Phelps design, but also because Carter is in Boston. Luc[as] loses out though. :-(

What do you guys think?

(I hope Rodolfo and Kent are listening...)

hhp

kakaze's picture

What are hybrid numerals?

hrant's picture

Basically old-style numerals that have bodies notably larger than the lc x-height, to the point where they sometimes seem like lining numerals shifted up/down - sometimes they even look like they're floating around a centerline.

Some people (like me) think that hybrid numerals are the optimal style for text work.

--

BTW, I'm starting not to like the "Boston numerals" label... :-/ I did some reading, and it's looking like Alexander Phemister might have been the designer/punchcutter at Phelps, Dalton & Co. (AKA Dickinson Foundry) at that time, and apparently he was a very bright fellow, so...

Plus I still have a nagging feeling that Austin might deserve the credit after all. Dunno, yet.

hhp

dyana's picture

Besides, Carter & Cone is in Cambridge.

John Hudson's picture

New England figures.

gerald_giampa's picture

Nantucket, clearly Nantucket. Heavily Influenced by Newfoundland Screech.

Yeah gotta love those figures but I am far from recommending them as a candidate to completely "purge oldstyle figures in text".

Cast brass early Boston hardware store figures probably predate Hrant's showings. Which is neither here nor there. John Downer would find nothing unusual about that.

Don't speak loudly or Mike Parker may show up. Boston hardware store figs are those which Starling Burgess based for one of his typefaces.

Goudy had used these style figures in Forum Title. The figure 4 in Goudy's give more probability for a balance setting. Notice dates in Hrant's showing.

http://www.lanstontype.com/ForumTitleRoman.html

I see nothing, and I mean nothing, is sacred!

Now "oldie style figs", has joined "upper case" as mud flat terra nova for the "New Typographical Bourgeoisie."

Not to be confused with Bourgeois, approximately 9 pt. midway between Long Primer and Brevier which most of them, probably were not aware

So I ask them what they think of the suggestion below? Pretty strange. I kind of like strange, that's why I hang around here.

http://www.lanstontype.com/Bodoni-Roman.html

hrant's picture

> New England figures.

That seems much better than "Boston".
But I'm realizing that if there's going to be any geographic reference, it really needs to be Scotland: both Austin and Phemister were Scottish (although the latter's working life was in New England). The former certainly deserves credit for breaking from the oldstyle/lining polarity*, and the latter might have in fact been the first to discipline the deviation into a clear new style (I might find out when I check out a 1888 Dickinson specimen book at UCLA). Plus in this case I think it's not right to favor Carter anyway, since it seems de Groot did it first (among living designers).

* Back around 1800. In fact, he (or maybe Bell) also deserves credit for the 3/4 lining numerals, which I also like. There seems to have been a lot of numeral innovation in the Highlands, for some reason.

So what I'm leaning towards is calling them Phemister, since Scottish is too broad: I'd like to be able to refer to "Bell numerals" (3/4 lining) and "Austin numerals" as well (like in Miller). But that's assuming Austin didn't make the first true hybrids as well! More research required...

--

> Notice dates in Hrant's showing.

You mean that they predate Goudy?
The absoute latest that font could be is 1902, but it's probably 2-3 decades earler.

The Forum numerals are interesting, and seem very similar to the ones in ATF Garamond (which did however come a bit later), in that they're a very extreme case of hybrids, almost lining. BTW, I seem to remember Kent pointing out a precedent to the ATF Garamond numerals from around 1900, but I'm not sure.

> http://www.lanstontype.com/Bodoni-Roman.html

Those I call "French old-style", since they were pioneered by the Didots (thanks to Rodolfo for pointing that out, a while back now). And I actually think that vertical arrangement is better than our conventional one, which is too low on the body overall.

hhp

gerald_giampa's picture

http://www.lanstontype.com/Bodoni-Roman.html
French Oldstyle Figures, indeed they are.

Notice dates in Hrant's showing.

I was just pointing out that Goudy used a slightly hanging "figure 4". For those new to this thread, these are not French Oldstyle figures. See the very interesting top posting by Hrant. Below is Goudy's version of alternate figures. These follow somewhat Hrant's posting. These are not French Oldstyle Figures.
http://www.lanstontype.com/ForumTitleRoman.html

I suggest Goudy's addition may have been helpful to balance the 1848. See 1848 without such balance, illustrated by Hrant's showing.

I am a little confused though. First, Hrant you say that your showing is the earliest you have found, but now indicate there were earlier experiments across the great pond. Are the earlier ones consistent with the #19 of the Phelps, Dalton & Co?

That question aside, I am not sure what they should be named. They are not French Oldstyle, they are too lining to be oldstyle and too oldstyle to be lining. Perhaps old lining figs would do. Tempting as it may be, naming such things after either location, foundry or designer can be a misleading.

Unless one can clearly establish the source not only to be fact, but to be first in fact.

kakaze's picture

"Basically old-style numerals that have bodies notably larger than the lc x-height, to the point where they sometimes seem like lining numerals shifted up/down - sometimes they even look like they're floating around a centerline. "


Ah, okay! I was just recently using Poetica for a wedding invitation (cliche I know), and I noticed it's numbers moved up and down like OSF numbers, but not enough for my liking. I wound up using the lining numerals.

hrant's picture

{OK, OK...}

> I was just pointing out that Goudy used a slightly hanging "figure 4".

Not just that - all the oldstyle up/downs are there too.

> these are not French Oldstyle figures.

Well, of course not.

> I suggest Goudy's addition may have been helpful to balance the 1848.

1) What's 1848? That's just the year Samuel Dickinson died.
2) What's "balance"? I'm looking for the person who first made hybrid numerals; Goudy isn't it.

> but now indicate there were earlier experiments across the great pond.

Since Austin had a track record for numeral innovation (3/4 lining, as well as what might be called "erratic"), and since Phemister worked for the same foundry (back in Scotland) immediately following him, it's possible that Phemister got more than just a hint from Austin. More research is required.

> old lining figs

That's an interesting angle... but they're not lining! :-/

> Unless one can clearly establish the source not only to be fact, but to be first in fact.

But you can never prove something does not exist. So it's a matter of sufficient certainty (and staying open to future correction - like what happened with Garamond/Jannon). If I find proof that Phemister made that #19 (noting that he was both punchcutter and designer* for most of his career, certainly after 1872 when he got promoted to full partner at Phelps), and I don't see anything like that in Austin's oeuvre, then "Phemister numerals" it is, since his "lineage" of numeral innovation is strong.

* Very rare during his era.

hhp

eomine's picture

> What's 1848?

Hmm, see fifth line in your sample: '16th December, 1848'.

rcapeto's picture

Hi Hrant, I'm hearing now. : )

"Boston numerals"? Interesting suggestion - but it seems that you're
now on to different things? (I'd be afraid in any case that people
translated it here as "n

gerald_giampa's picture

"Scotch on the rocks" Go figures!

gerald_giampa's picture

That's an interesting angle... but they're not lining! :-/

What about "not lining figures"?

But you can never prove something does not exist.

I agree, you can't. To make things worse, sometimes they too often do exist. I suggest naming as if to indicated "discovery" or suggest "certainty" would be folly.

cerulean's picture

They aren't lining? If they aren't lining, then why oh why must the huge gaping space around the numeral one be preserved? It's what I most dislike about lining numerals in text, and everybody seems to have gotten used to it except me. If you saw a word such as, say, "ions" or "late" set like that, you'd immediately say "fix that crapulous kerning, it practically looks like two words." What I would like to see more of, on the flip side of these sort-of-lining oldstyle numerals, are non-lining equal-height numerals, with visual kerning on par with letters.

cerulean's picture

Is my face red. Thank you for setting me straight on the terminology, but you can see why I was confused; in most conversations "lining" would seem to be shorthand for lining tabular, and "oldstyle" would seem to be shorthand for oldstyle proportional. I was under the impression that no one had ever yet considered that lining and tabular could be separated, hence my pet peeve. My hand has been out of typography for a long time, and I'm mostly exposed to the "defaults" of the masses, in which lining tabular figures are often all you get.

hrant's picture

> Call them Scotch figures once and for all

I'm taking this pretty seriously now...

> Don't let this happen to you.

You should see how I drive.

BTW, there is also some evidence that the Ancient Egyptians (the ones before the ones we think we know* about - the ones that made the Sphinx, not the much later Pyramids) visited the "New World".

* But: http://news.bbc.co.uk/2/hi/americas/3254852.stm

Speaking of "transitional", last night I found myself revisiting a passage in Updike that Kent had once pointed out. In it (vII, p230-1) Updike says that one Hunter introduced lining figures in 1785, but then shows what he calls "transitional" figures that look like hybrids, except you can't be sure because there's no lc to compare to. The thing is, his "transitional figures" look very much like the Miller (Scotch) ones, except those weren't made before the Hunter innovation (as far as I can tell), so how can they "transition"?

--

Kevin, actually some people use "lining" to mean fixed-width (probably because such figures do enable vertical alignment across subsequent lines). BTW, it is interesting that the numerals in that #19 seem to be fixed-width. I don't remember seeing that in [early] hybrids - it seems to be an influence of lining figures (although oldstyles are actually easier to make fixed-width).

hhp

eomine's picture

Trying to find out who first designed this kind of figures sounds great. But, I don't see why this new terminology should necessarily be tied to its history. Let's see, who first designed lining figures?

If these 'hybrid' figures need a new name, then it should be something related to its 'physical' aspect:

> Basically old-style numerals that have bodies
> notably larger than the lc x-height, to the
> point where they sometimes seem like lining
> numerals shifted up/down - sometimes they even
> look like they're floating around a centerline.

:-)

John Hudson's picture

'Scotch figures' may be confusing to many people, since there is already a Scotch classification of type, many examples of which do not include this kind of figure.

I think having what amount to nicknames for different figure styles is fine, but we should first have an accurate way to describe figures. The most important distinction to this discussion is that between lining and ranging figures. The term 'lining figures' is used so commonly for cap-height figures that we tend to forget that it is an abbreviation for 'cap-height lining figures', i.e. of figures that are a common height, aligned to the caps. It is perfectly possible to have lining figures that align to other heights, e.g. to smallcap height, even to x-height and, in the case of Bell's famous figures, 'three-quarter cap height lining figures' (commonly called simply 'three-quarter figures' -- again, an abbreviation). Just as the term 'lining' does not by itself imply anything about the height of the figures -- other than that it is common --, so the term 'ranging' doesn't necessarily imply oldstyle figures, i.e. figures that range above and below a median defined by the x-height. It is perfectly feasible to have ranging figures that range above and below a median defined by the smallcap height, or even the cap height, although the latter would look very odd. From this it should follow pretty obviously that what we're talking about in this thread, precisely, are three-quarter cap-height ranging figures. Anything else is simply a nickname (of which there might be many: Scotch, Phemister, Boston, etc.) or an abbreviation (three-quarter ranging figs).


NB: using 'lining' to mean fixed-width is simply wrong. Lining and ranging describe vertical characteristics of figures; proportional and tabular describe horizontal characteristics. The truth of this is obvious as soon as one sits down to design typefaces with multiple figure styles: cap-height lining proportional, cap-height lining tabular, x-height ranging (oldstyle) proportional, x-height ranging(oldstyle) tabular, smallcap-height lining proportional, smallcap-height lining tabular, etc.

gerald_giampa's picture

three-quarter ranging figs :-)

hrant's picture

> it should be something related to its 'physical' aspect

True.

--

John, although I don't believe "ranging" is a good qualifier (too rare and ambiguous), I like the directness of your approach. Even though "3/4 oldstyle numerals" is cumbersome, it's certainly better than "hybrid", and without [near-]total historical clarity, just slapping somebody's name on it is too cavalier. The only problem with it that I see is that it implies there are full-height oldstyle numerals... But I guess there might even be such a thing (like maybe Quadraat Headliner could use them).

BTW, this is the second time you've made me change/adopt terms. The first two were "chirography" and "wedge serif". I only have two on you ("bouma" and "notan") but I'm not sure how committed to those you are. :-)

hhp

John Hudson's picture

The only problem with it that I see is that it implies there are full-height oldstyle numerals...

As I wrote: 'three-quarter ranging figures' should be understood as an abbreviation of 'three-quarter cap-height ranging figures'. If you want to be really precise, use the latter. Of course, making this distinction does oblige us to stop calling Bell's figures 'three-quarter figures' and start calling them 'three-quarter cap-height lining figures'.


I like 'notan' -- although I suspect we disagree about its role in a subtle but important way --, but you can keep bouma.

kentlew's picture

Sorry I've been too damn busy with projects and deadlines to join in. I was fine with "hybrid". I don't mind "Scotch" as a nickname. But I think John's terminology is the most descriptive and precise, if that's your intention. I like "three-quarter ranging".

For the record, the lining figures in Whitman are proportional, not tabular, since they were intended to work primarily in all-caps headings and such, not so much financials. It wouldn't take much to adjust them to tabular width and I'm prepared to do that if FB ever gets a request.

-- K.

kentlew's picture

Oh, I see that Stephen edited the Whitman reference out of his post. So my last statement might seem non sequitur. Oh well, I'll let it stand.

Back to the grindstone . . .

rcapeto's picture

'three-quarter ranging figures' should be understood as
an abbreviation of 'three-quarter cap-height ranging figures'.


The best thing about Vox

hrant's picture

Well, the "cap-height" part we can dump - it's totally moot.
And in practice just "3/4" will work most of the time, because the lining and oldstyle versions of 3/4 numerals (please don't use "figures", it's unfocused) share more than they don't: a sensitivity to functional scale. Somebody who makes 3/4 oldstyle nums will probably make his lining style 3/4 too.

hhp

John Hudson's picture

Rodolfo, most people call me 'John'. Some people call me 'Mr Hudson'. Those who wish to distinguish me from my brother call me 'Mr John Hudson'. Those who need to distinguish me from some other John Hudson might append '...Tom Hudson's son' or '...the type designer' or '...the Anglo-Canadian type designer'. I have a friend who has known me since we were thirteen, and she's the only person who calls me 'Johnny' with impunity.

The terminology we use to refer to things is determined by our need of precision, which typically is determined by whom we are talking to (although in my case it is also influenced by a certain, ahem, 'retentiveness'). Sure, I know what you mean by the nickname 'Didone', but will a non-typographer?

rcapeto's picture

Mr Hudson,

that

hrant's picture

> serve us a short, practical, easy term for typographic online chat

That's a very good point. So I'll keep thinking...

hhp

John Hudson's picture

On the basis that three-quarters of anything implies a quarter of the same. And on the basis that the numerals in question are Ranging. I propose the name 'Quarange Numerals'.

The slogan shall be 'Quarange: it rhymes with orange'.

aquatoad's picture

Free Range Numerals.
Cockadoodletwo.

Stephen Coles's picture

Just an aside to say that I am a big fan of this topic, Hrant.
It shows how far Typophile has come that it's a legitimate
forum for proposing new naming standards. I don't mean
that in a sarcastic way. I'm very sincere about how mature
and useful the discussion at this site can be. Type on!

Stephen Coles's picture

I believe you're confusing the term "lining" with "tabular", Kevin.

Heights
Lining = Numerals are all the same height, lining with caps
Oldstyle = Numerals are of various heights for better
flow with mixed-case text

Widths
Tabular = Numerals spaced with the same width for easy
setting of columnar data on multiple lines
Proportional = Numeral spacing of varying widths,
in the same way as the letter glyphs.

A good text family should include at least 2 types of
numerals, Oldstyle Proportional and Lining Tabular, but the
really useful families include 4 variations, one font with
each width and height combination. See Neutraface.

Birdseeding's picture

I think this thread deserves revival after seeing this somewhat clumsy term. "3/4 oldstyle figures" didn't really stick like "wedge serif" and "chirography" did, dit it?

hrant's picture

Hey, it's been over 8 years?!
Numerals in fonts have come along way, baby.

And in that time I guess good old "hybrid" has proven
its staying power. As lousy as it is it's still better than
"short-ranging", which is totally confusing.

hhp

Nick Shinn's picture

These figures are conventionally old-style.
Sure, they have a large “zero-height”, but that doesn’t make them a lining-oldstyle hybrid.
Many “normal” oldstyle figures have zero-heights that are larger than x-height.
And many “normal” oldstyle figures have zero and one that are lining-style (rather than a one that looks like a small-cap “I” and a zero that is a monoline circle), but we don't call those hybrids, which IMO would be a more accurate use of the term.

hrant's picture

No, conventional old-style is where the bodies of the numerals
are x-height. It's only quite recently that designers have taken
up the good habit of breaking that established practice. If you
look at fonts from around the time this thread* started you'd
be hard-pressed to find hybrid numerals.

* It might even be true that hybrid numerals
started spreading faster due to the numerous
Typophile discussions on the topic.

As for the way the zero and one look, that's moot here: we're
talking about height, and the way a couple of numerals look
is not enough to label the whole set of them as a new idea.

hhp

hrant's picture

BTW, a textbook case of hybridity being non-kosher in
the very recent past was what happened to Georgia, which
initially had superb hybrid numerals* that were eventually
toned down to near-conventional proportions. I think the
chances of that sort of thing happening now are much lower.

* http://typophile.com/node/63979#comment-377194

hhp

Nick Shinn's picture

@Hrant: conventional old-style is where the bodies of the numerals are x-height.

You are mistaken.
Consider that most conventional of faces with OSF as default, Georgia, which you just referred to.
Or the ClearType faces.

hrant's picture

As I said earlier in the thread, Carter was one of the very
early proponents of hybrid numerals; Georgia is famous,
but it's only one font*. As for the ClearType designs, they
are exactly part of the recent trend that I describe.

* And, to repeat, the hybridity of
its numerals was tellingly neutered.

Try to compile a list of fonts with hybrid numerals that came
out before Typophile caught on. You will find almost nothing.

hhp

Té Rowan's picture

Lining, stepping and ranging? Sounds better to me than lining, hybrid and ranging. When someone mentions hybrid, I think of roses.

Nick Shinn's picture

Try to compile a list of fonts with hybrid numerals that came
out before Typophile caught on. You will find almost nothing.

Older faces with old-style figures that aren't “x-height”: Bodoni Antiqua, Plantin, Perpetua.
Preface (one of mine).

riccard0's picture

When someone mentions hybrid, I think of roses.

No, no, hybrids are the environment-conscious ones (an twos, and threes... ;-)

hrant's picture

Bodoni Antiqua (URW++) does have gently hybrid numerals.
BTW, do you know how closely it's based on Bodoni's work?
It would be cool if old Giambattista actually make hybrids.

The digital version of Plantin seems to have full lining nums,
and I don't have any of my metal typeface specimen books
handy. Do you have a link/sample?

Perpetua: The Italic does indeed have nice hybrids, but the
Roman's nums are actually shorter than the x-height! Which is
quite strange. In any case Gill was indeed a numeral innovator;
the tall "2" in Perpetua's Roman and Joanna is something we
should all be doing frankly. Although the French were ahead of
everybody in terms of OS nums (until they took a break from text
fonts for a while* and when they resumed they picked up some
nasty Anglo habits).

* "Thanks" to Charles Peignot.

Preface: those nums are barely a hair taller than the x-height.
I've seen that done before, but I've never understood it. Being
a typeface designer I would notice it and believe it's intentional,
but how could it appear as anything but a "mistake" to laymen?
Isn't it a sort of golden rule of visual design to either deviate
clearly or not at all? (Which is not to say that rules shouldn't
be broken - to me it does need to make sense though.)

So, how long did it take you to find ~1.5 examples of
early hybrid numerals? :-) BTW, the #19 of the Phelps,
Dalton & Co. handily predates those.

hhp

Si_Daniels's picture

>When someone mentions hybrid, I think of roses.

Are you taking the prius!? :-)

Té Rowan's picture

Nope. I'm not toy'ting wit' ya. Got an eco-friendly vehicle already anyway: A bike.

Birdseeding's picture

This digital Plantin version does have hybrid numerals: http://myfonts.us/td-iueTpR

quadibloc's picture

Maybe the face has changed, but all I can see on MyFonts for Francisco Serial are full oldstyle figures.

Syndicate content Syndicate content