How to do lowercase numerals in LaTeX?
Gernot Hassenpflug
Not sure if terminology is correct, as Google, the UKTeXFAQ and CTAN
did not come up with any hits: some websites and books on typography
point out that using non-aligned or lower-case numerals in text looks
better than the baseline-aligned text which is always used in
mathematical environments.
So, for example, addresses and phone numbers in the text could be
written in these lower-case numerals. However, I can't find any
reference to such numerals in standard LaTeX. Is the use of such
numerals frowned upon? What would one need to search for to find a
suitable package on CTAN? I'm guessing European typesetting makes use
of such numerals even if the US does not.
Regards,
Gernot
--
Grrr!! ...Pick a reason...
Robin Fairbairns
> Not sure if terminology is correct, as Google, the UKTeXFAQ and CTAN
>did not come up with any hits: some websites and books on typography
>point out that using non-aligned or lower-case numerals in text looks
>better than the baseline-aligned text which is always used in
>mathematical environments.
sorry about the uk faq drop-off. i've got this question as an action
item, but keep not finding the energy to write an answer, which
requires me to do some research first. i'm getting old...
(faq contributions are always welcome...faq-devel@tex.ac.uk -- no, i
don't mean money, i mean text!)
--
Robin Fairbairns, Cambridge
William F. Adams
These're called medieval or oldstyle figures and usage of them is
mostly font-specific.
The _The Comprehensive LaTeX Symbol List_ lists these and macros for
them, and some font packages (e.g., mathpazo) make them available:
\usepackage[osf]{mathpazo}
William
Tariq
Could you please be more specific and let us know what exactly the
command is to use the non-aligning digits. The same problem has vexed
me for the longest time and I have been very frustrated at not being
able to solve it satisfactorily. Thanks.
Tariq
Bob Tennent
Try \oldstylenums{0123456789}. If such "old-style numerals" are to be
the default (for numerals in text), use the eco package. If you're
not using Computer Modern, use the textcomp package.
Bob T.
William F. Adams
As I noted, it varies by font. In particular, you need a font w/ the
oldstyle figures encoded and named so that LaTeX understands it as
such unless you're going to use OpenType using XeTeX (or luatex once
it's done).
Here's an example which should get you started:
\documentclass{minimal}
\usepackage[T1]{fontenc}
\usepackage{lmodern}
\usepackage{textcomp}
\begin{document}
Testing, 0, 1, 2, 3, testing, 4, 5, 6, 7, 8, 9
Testing,
\textzerooldstyle,
\textoneoldstyle,
\textthreeoldstyle,
testing,
\textfouroldstyle,
\textfiveoldstyle,
\textsixoldstyle,
\textsevenoldstyle,
\texteightoldstyle,
\textnineoldstyle,
\end{document}
HTH!
William
anon k
> Not sure if terminology is correct, as Google, the UKTeXFAQ and CTAN
> did not come up with any hits: some websites and books on typography
> point out that using non-aligned or lower-case numerals in text looks
> better than the baseline-aligned text which is always used in
> mathematical environments.
They're all baseline-aligned, but the 'old-style figures' that you want
have ascenders and descenders. The full-height sort are called 'lining
figures'.
There are others too, called 'tabular figures,' that all have the same
widths so that you can line them up easily in tables. These tend to be
in very modern typefaces.
It isn't so much a matter of 'looking better' as it is of being more
legible.
One of the big ideals of the fine press tradition is to not look like
anything at all, i.e. to be so good that the reader doesn't consciously
notice anything. We want to give the reader a certain subconscious
'feel' as he reads, but to keep his conscious mind on the sentences, not
the book. I had a student once tell me that he thought LaTeX to be an
unfair advantage in writing term papers because it made it so easy to
give a strong impression of well finished work no matter how roughshod
the research and writing were.
But the Kelmscott Press is a good counterexample to show that not
everyone follows that ideal. Maybe it was discreet in its day, and
connected well with the revival of certain medieval manuscript
conventions, but its output looks conspicuously stylized now.
> Is the use of such numerals frowned upon?
It's frowned upon to use old-style figures in all-caps text, and vice versa.
It's frowned upon to use either set of figures in a typeface that
natively has only the other kind: usually the type designer makes this
decision when designing the shapes of the numerals.
But there are always people who decide to do the opposite, sometimes to
excellent effect. It is better to think of what you want the written
piece to achieve, than to follow conventions without justifying them.
Some conventions are historical hangovers from technological limitations
in printing, papermaking or binding that have since been overcome.
> I'm guessing European typesetting makes use
> of such numerals even if the US does not.
They're used in Europe and in the US. The biggest determinants are
which typeface is in use, and whether the numerals are in body text or
in display. Maybe some manuals on book design, rather than on
typography, would be useful to you?
John Harper
anon k <nos...@nul.nul> wrote:
>
>... the 'old-style figures' ... have ascenders and descenders.
>One of the big ideals of the fine press tradition is to not look like
>anything at all, i.e. to be so good that the reader doesn't consciously
>
>But the Kelmscott Press is a good counterexample to show that not
>everyone follows that ideal.
I lost my respect for Kelmscott when I looked up the Little Treatise
on the Astrolabe in their edition of Chaucer and found that they had
omitted the diagrams. It's impossible to follow without them!
On the two kinds of numerals, I found that the following works, and
it still works with mathptmx replaced by mathpazo, or with the
\usepackage line removed completely to give Computer Modern.
\documentclass{article}
\usepackage{mathptmx}
\begin{document}
1234567890 \oldstylenums{1234567890}
\end{document}
Sixty years ago L.J.Comrie wrote in the preface to Chambers's Six-
Figure Mathematical Tables "It is only in the U.S.A. that the modern
face figure persists to any extent. A very convincing testimony to the
superior legibility of old style figures came from the compositors and
proof readers of these tables, who emphatically agreed that the figures
here used were less fatiguing and so less liable to misreading than
modern face figures."
-- John Harper, School of Mathematics, Statistics and Computer Science,
Victoria University, PO Box 600, Wellington 6140, New Zealand
e-mail john....@vuw.ac.nz phone (+64)(4)463 5341 fax (+64)(4)463 5045
Gernot Hassenpflug
"old-style" numerals (and I am glad to see I am not the only one who
had this 'problem'), a massive and heart-felt thank you. To Robin
Fairbairns, I'll try to send a contribution to you this weekend from
the advice given by the other posters.
Ever since I discovered that there is a physiological reason/mechanism
underlying every movement in martial arts, either with respect to
oneself or w.r.t. the other person(s), I have become interested in the
physiologocal reasons for typography. I'll have a look into book
design too, thanks for the hint.
--
BOFH excuse #24:
network packets travelling uphill (use a carrier pigeon)
jorge sallum
Have a look in the Companion about it. The option is "hanging", but I
don't remenber the syntax now.
J.
anon k
> In article <4DSgi.4631$vi5....@newssvr17.news.prodigy.net>,
> anon k <nos...@nul.nul> wrote:
>> ... the 'old-style figures' ... have ascenders and descenders.
>> The full-height sort are called 'lining figures'.
> ...
>> One of the big ideals of the fine press tradition is to not look like
>> anything at all, i.e. to be so good that the reader doesn't consciously
>> notice anything. ...
>>
>> But the Kelmscott Press is a good counterexample to show that not
>> everyone follows that ideal.
>
> I lost my respect for Kelmscott when I looked up the Little Treatise
> on the Astrolabe in their edition of Chaucer and found that they had
> omitted the diagrams. It's impossible to follow without them!
I'm not sure that it normally had diagrams. But I guess that they could
have been particularly nice with Morris at the helm to optimize the layout.
> Sixty years ago L.J.Comrie wrote in the preface to Chambers's Six-
> Figure Mathematical Tables "It is only in the U.S.A. that the modern
> face figure persists to any extent. A very convincing testimony to the
> superior legibility of old style figures came from the compositors and
> proof readers of these tables, who emphatically agreed that the figures
> here used were less fatiguing and so less liable to misreading than
> modern face figures."
In tables and in body text, I generally agree. But in all-caps
headings, old-style figures can be very distracting. Or
attention-grabbing, if attention is what's wanted, as is often the case
in poster design.
I also agree from the hand-composition standpoint. Old-style figures
are much faster to sort out when you're getting tired and your stick of
type has just pied onto the stone...
Georges
For plain TeX applications, I have a macro file modeled after the one
in Knuth's TeX book, which establishes the oldstyle numerals as
virtual fonts included in cmolddig by Rowland McDonnell. Once this is
loaded, all numerals are oldstyle. If one wants the lining digits,
there is a \liningdigits command.
Georges
John Harper
anon k <nos...@nul.nul> wrote:
>John Harper wrote:
>> In article <4DSgi.4631$vi5....@newssvr17.news.prodigy.net>,
>> anon k <nos...@nul.nul> wrote:
>>> ... the 'old-style figures' ... have ascenders and descenders.
>>> The full-height sort are called 'lining figures'.
>> ...
>>> One of the big ideals of the fine press tradition is to not look like
>>> anything at all, i.e. to be so good that the reader doesn't consciously
>>> notice anything. ...
>>>
>>> But the Kelmscott Press is a good counterexample to show that not
>>> everyone follows that ideal.
>>
>> I lost my respect for Kelmscott when I looked up the Little Treatise
>> on the Astrolabe in their edition of Chaucer and found that they had
>> omitted the diagrams. It's impossible to follow without them!
>
>I'm not sure that it normally had diagrams. But I guess that they could
>have been particularly nice with Morris at the helm to optimize the layout.
Chaucer ended 57 of his paragraphs with "For more declaracioun, lo here
thi figure."! His figures (i.e. diagrams) have been published in
Gunter, R.T. (ed.) "Early Science in Oxford" vol.5 (1929?), Oxford.
Some more modern editions also include them.
>> Sixty years ago L.J.Comrie wrote in the preface to Chambers's Six-
>> Figure Mathematical Tables "It is only in the U.S.A. that the modern
>> face figure persists to any extent. A very convincing testimony to the
>> superior legibility of old style figures came from the compositors and
>> proof readers of these tables, who emphatically agreed that the figures
>> here used were less fatiguing and so less liable to misreading than
>> modern face figures."
>
>In tables and in body text, I generally agree. But in all-caps
>headings, old-style figures can be very distracting.
So old and new style ought to be both available in all type faces, and
the appropriate one used for figures (i.e. digits), just as upper and
lower case are (or ought to be:-) for letters.
Gernot Hassenpflug
> In article <GU8hi.23083$C96....@newssvr23.news.prodigy.net>,
> anon k <nos...@nul.nul> wrote:
>>John Harper wrote:
>>> In article <4DSgi.4631$vi5....@newssvr17.news.prodigy.net>,
>>> anon k <nos...@nul.nul> wrote:
>>>> ... the 'old-style figures' ... have ascenders and descenders.
>>>> The full-height sort are called 'lining figures'.
>>> ...
>>>> One of the big ideals of the fine press tradition is to not look like
>>>> anything at all, i.e. to be so good that the reader doesn't consciously
>>>> notice anything. ...
>>>>
> /../
>>> Sixty years ago L.J.Comrie wrote in the preface to Chambers's Six-
>>> Figure Mathematical Tables "It is only in the U.S.A. that the modern
>>> face figure persists to any extent. A very convincing testimony to the
>>> superior legibility of old style figures came from the compositors and
>>> proof readers of these tables, who emphatically agreed that the figures
>>> here used were less fatiguing and so less liable to misreading than
>>> modern face figures."
>>
>>In tables and in body text, I generally agree. But in all-caps
>>headings, old-style figures can be very distracting.
>
> So old and new style ought to be both available in all type faces, and
> the appropriate one used for figures (i.e. digits), just as upper and
> lower case are (or ought to be:-) for letters.
That is more or less the impression I got from reading what I managed
to get my hands on: so I was a bit surprised that this issue seems so
font-dependent. Still, it is solvable, thanks to the many useful
replies.
Robin Fairbairns
>har...@mcs.vuw.ac.nz (John Harper) writes:
>>[...]
>> So old and new style ought to be both available in all type faces, and
>> the appropriate one used for figures (i.e. digits), just as upper and
>> lower case are (or ought to be:-) for letters.
>
>That is more or less the impression I got from reading what I managed
>to get my hands on: so I was a bit surprised that this issue seems so
>font-dependent. Still, it is solvable, thanks to the many useful
>replies.
why should you be surprised? nothing's perfect, and there are many
would say that perfection is illusory, anyway.
--
Robin Fairbairns, Cambridge
Gernot Hassenpflug
Well, because after reading about the topic, I thought that Donald
Knuth (and Leslie Lamport) among others would have mentioned it a bit
more. I'm not criticizing mind you, just expressing my personal vibes
here :-)
--
BOFH excuse #410:
Electrical conduits in machine room are melting.
anon k
> In article <GU8hi.23083$C96....@newssvr23.news.prodigy.net>,
> anon k <nos...@nul.nul> wrote:
>> John Harper wrote:
>>> In article <4DSgi.4631$vi5....@newssvr17.news.prodigy.net>,
>>> anon k <nos...@nul.nul> wrote:
>>>> ... the 'old-style figures' ... have ascenders and descenders.
>>>> The full-height sort are called 'lining figures'.
>>> ...
>>>> One of the big ideals of the fine press tradition is to not look like
>>>> anything at all, i.e. to be so good that the reader doesn't consciously
>>>> notice anything. ...
>>>>
>>>> But the Kelmscott Press is a good counterexample to show that not
>>>> everyone follows that ideal.
>>> I lost my respect for Kelmscott when I looked up the Little Treatise
>>> on the Astrolabe in their edition of Chaucer and found that they had
>>> omitted the diagrams. It's impossible to follow without them!
>> I'm not sure that it normally had diagrams. But I guess that they could
>> have been particularly nice with Morris at the helm to optimize the layout.
>
> Chaucer ended 57 of his paragraphs with "For more declaracioun, lo here
> thi figure."! His figures (i.e. diagrams) have been published in
> Gunter, R.T. (ed.) "Early Science in Oxford" vol.5 (1929?), Oxford.
> Some more modern editions also include them.
Thanks, I must have a look for these.
>>> Sixty years ago L.J.Comrie wrote in the preface to Chambers's Six-
>>> Figure Mathematical Tables "It is only in the U.S.A. that the modern
>>> face figure persists to any extent. A very convincing testimony to the
>>> superior legibility of old style figures came from the compositors and
>>> proof readers of these tables, who emphatically agreed that the figures
>>> here used were less fatiguing and so less liable to misreading than
>>> modern face figures."
>> In tables and in body text, I generally agree. But in all-caps
>> headings, old-style figures can be very distracting.
>
> So old and new style ought to be both available in all type faces, and
> the appropriate one used for figures (i.e. digits), just as upper and
> lower case are (or ought to be:-) for letters.
I would think so too, but not all type designers agree. And when I
consider examples, I start to sympathize. Helvetica with old-style
numerals just seems a bit odd to me.
Karl Ove Hufthammer
>> So old and new style ought to be both available in all type faces, and
>> the appropriate one used for figures (i.e. digits), just as upper and
>> lower case are (or ought to be:-) for letters.
Not only that, but both tabular and proportional versions should be
available too, IMO (making four different variants).
> I would think so too, but not all type designers agree. And when I
> consider examples, I start to sympathize. Helvetica with old-style
> numerals just seems a bit odd to me.
Even Helvetica's 'default' numerals look odd when used in a nontabular
context, especially with large font sizes. See this document for an
example:
http://www.uib.no/People/st11188/latex/helvetica-tabular.pdf
Helvetica (or really HV-Math) is used in the chapter headings. Look how
*absolutely terribly ugly* the spacing is for chapters 10 to 19 (especially
chapter 11) and chapter 21 and 31. The reason is that the width of a '1'
is the same as the width of a '2' (so the numbers would line up properly
if used in tables).
My solution was to hack the display of chapter numbers to decrease the
spacing. My code for this is not very elegant, but the result is, if not
nice, at least *nicer*:
http://www.uib.no/People/st11188/latex/helvetica-proportional.pdf
All I want for Christmas is a Helvetica with proportional numbers.
--
Karl Ove Hufthammer
William F. Adams
wrote:
Well, there's Berthold Akzidenz Grotesk --- they did a light weight w/
oldstyle figures for setting their business cards. I find it quite
handy.
William
anon k
We are now getting onto the difference between a typeface and its
implemented fonts. Proportional numerals definitely exist for Helvetica
in metal type!