The Number sign—now and then—across languages.

While wondering how to draw the number sign for my digital version of De Vinne, I stumbled into these very interesting considerations by Jonathan Hoefler:

«Nº was the number sign before # became a number sign, and it refreshingly serves this one and only purpose. Compare the #, which when preceding a number is read as “number” (“#1 in my class”), but when following a number means “pound” or “pounds”²2 If you’re curious what the # symbol has to do with the abbreviation lbs., here’s one possible missing link (“70# uncoated paper”), leading to printshop pile-ups like “#10 envelope, 24# bond.” To programmers, a # can mean either “ignore what follows” (as in a Python comment) or “use what follows” (when referencing a page fragment, or a Unicode value in html.) To a proofreader, a # means “insert space,” so in the middle of a numbered list, the notation “line #” does not mean “line number,” but rather “add a line space.” Because of #’s resemblance to the musical symbol for “sharp” (♯), it’s a frequent stand-in for the word “sharp,” and often the correct way of rendering a trademarked term such as The C# Programming Language. The # is rapidly assuming musical duties as well, especially in online databases, leading to catalog collisions like “Prelude & Fugue #13 in F#.” How fortunate a designer would be to have a numero symbol, with which to write “Prelude & Fugue Nº 13 in F#,” or “Nº 10 Envelope, 24# bond.”»

David Jonathan Ross, in Roslindale, uses the common form (just stress reversed) but I was wondering how much, and to which degree, in the late 1890s/early 1900s, the current /# was in use, as opposed to /
Nº.
In fact, in Italy we have almost no use for /# as opposed to /Nº. When we list numbers we always use either N. or No.

Your thoughts based on experience?
Tagged:
«1

Comments

  • # is a fairly old mark for numbering, supposedly used for (or originated as?) with collation / bundle / pack units marking by merchants and trade workers. I remember having seen a specimen in an older type book with such a reference, I haven’t any sample at hand, though.
    The use of # for “pounds” is just a poor-mans workaround i.m.h.o., since other, real and well-known abbr. signs for Pound are in existence for many Hundred years.
  • In most cases I made both signs. # on its default place and No. on uni2116.
    You can use the locl feature to change # by uni2116 for Italien language - but I'm not sure its appropriate.

  • andreas said:
    In most cases I made both signs. # on its default place and No. on uni2116.
    You can use the locl feature to change # by uni2116 for Italien language - but I'm not sure its appropriate.

    Thanks. Oh, actually the italian is not a problem, it’s usually typed as “N." or "No." :-)
    I was more wondering if it would be nice to design the /# with horizontal instead of vertical stress, as it is in Roslindale, or in Carter's Miller. And I was wondering if there were historical lead examples to look at.
  • # is a fairly old mark for numbering, supposedly used for (or originated as?) with collation / bundle / pack units marking by merchants and trade workers. I remember having seen a specimen in an older type book with such a reference, I haven’t any sample at hand, though.
    The use of # for “pounds” is just a poor-mans workaround i.m.h.o., since other, real and well-known abbr. signs for Pound are in existence for many Hundred years.
    In an “elzevir" or "scotch modern" kind of face, Andreas, would you put the stress.
    Here’s how it’s designed in Miller (also vertically slanted) but not sure it’s appropriate for the “robust and eccentric” De Vinne. :-)
  • I prefer often to have it straight upright in a Regular font, vs. slanted in an Italic. The horizontal bars are basically just horizontal. A horizontal stress is sort of standard, I’d say.
    This is how it looks like in some of my fonts:


  • I prefer often to have it straight upright in a Regular font, vs. slanted in an Italic. The horizontal bars are basically just horizontal. A horizontal stress is sort of standard, I’d say.
    This is how it looks like in some of my fonts:
    Thanks! I have to admit that making it upright, especially when more stressed, gives me a strange impression. For example, in the fourth example, first column, of your fonts, is nice, but in the first it gives me a sense of excessive stiffness. It may be that we’re more used to inclined forms, though. The last one, bottom right (in the foliated typeface) is very nice. Well, in fact the typeface looks very nice in general! :D
  • Just some relevant notes:
    • Shady Characters wrote about the octothorpe: https://shadycharacters.co.uk/2011/05/the-octothorpe-part-1-of-2/
    • The only reason ‘№’ is a character on its own is the lack of an ‘N’ on Russian typewriters.
    Thanks! Yes, I have read some article, which probably were informed by the one you linked.
    I know “№” is forcedly added because of that, but actually I like the idea of having it, maybe accessible through a stylistic alternate, for “numero” in Italy would not be written using /#.
  • John Savard
    John Savard Posts: 1,124
    edited January 2021
    The use of # for “pounds” is just a poor-mans workaround i.m.h.o., since other, real and well-known abbr. signs for Pound are in existence for many Hundred years.

    And I just checked; ℔ is U+2114.

    How fortunate a designer would be to have a numero symbol, with which to write “Prelude & Fugue Nº 13 in F#,” or “Nº 10 Envelope, 24# bond.”»
    Or "№10 Envelope, 24℔ bond."

    • The only reason ‘№’ is a character on its own is the lack of an ‘N’ on Russian typewriters.

    Oh, all right; "Nº 10 Envelope, 24℔ bond." then.
  • The inclination of the glyph (and also the width) is indeed a subtle issue. And also the question of wether and how much the stems should descend from the baseline. As I said, I lean towards an upright design in some Regular styles, but that is not a fixed credo. I always view and check this char. together with the figures, as I do with monetaria and a few other ch.s.
  • notdef
    notdef Posts: 168
    edited January 2021
    Thanks Josh!
    Nº was the number sign before # became a number sign

    The Nº combination originates with scribal abbreviation. The first letter of the word is followed by the last letter superscripted (Latin: “numero”), which allows for quick and compact composition of common words. 

    Now, my comment about № only being a character due to Russian typewriters is not the entire story. It has taken on a life of its own, and is often stylised differently from the regular N, more like an upright cursive swashy shape. Today it is used more akin to the &: giving display text an ornamental historical flavour.

  • Claudio Piccinini
    Claudio Piccinini Posts: 681
    edited January 2021
    The inclination of the glyph (and also the width) is indeed a subtle issue. And also the question of wether and how much the stems should descend from the baseline. As I said, I lean towards an upright design in some Regular styles, but that is not a fixed credo. I always view and check this char. together with the figures, as I do with monetaria and a few other ch.s.
    I always find your thoughtful considerations very useful. Many thanks. In the end, I decided to go with a quite standard solution, following Ross' example in Roslindale and Miller: since De Vinne’s numerals (at least in the early version, which I am following faithfully) are slightly less tall than the uppercase (as it is in Miller), I did it this way:

    And yes, I always follow the same rationale: when I design a symbol, a sign or a typographic element, I calibrate it according to the height and proportions of the glyphs which is meant to accompany, in this case the numerals.
  • 3. It is interesting that Hoefler, as quoted above, does not mention "hashtags." I do not know when those remarks of his were written — very possibly before the advent of social media — but this use for the symbol is now ubiquitous. The glyph has of course been known variously in common speech as a number sign, a pound sign, or a hash sign. The last of these names, particularly prevalent in computer programming, became the preferred term in a social media context. A hashtag is thus a tag (descriptor) preceded by a hash [sign], but the word hashtag is now often used incorrectly to refer to the sign itself. See also https://www.etymonline.com/search?q=hash.
    “Hash sign” sounds about right. :-)  Although it would be fascinating to see and document where the stylization from the "Lbs" symbol occurred.
    "Hashtag" is clearly referred to the tags, in italian we call it "cancelletto" (which is a sort of term of endearment for “gate", or “small gate"). Which is the main, historical meaning of "hash" in english? (I mean, before its use in notations), if it has one?

  • Thanks Josh!
    Nº was the number sign before # became a number sign

    The Nº combination originates with scribal abbreviation. The first letter of the word is followed by the last letter superscripted (Latin: “numero”), which allows for quick and compact composition of common words. 

    Now, my comment about № only being a character due to Russian typewriters is not the entire story. It has taken on a life of its own, and is often stylised differently from the regular N, more like an upright cursive swashy shape. Today it is used more akin to the &: giving display text an ornamental historical flavour.

    I love the form. But I have to say it’s not necessarily “decorative” per se, mostly the cursive form helps to make it stand out, and not just in large type sizes.
  • In my experience, the written form when it is meant as "number" is often small and high, like an asterisk. But typographically its density asks for a larger rendering, and I think that works better for its now-common "hashtag" meaning anyway.
  • Nick Shinn
    Nick Shinn Posts: 2,198
    @Claudio Piccinini
    when I design a symbol, a sign or a typographic element, I calibrate it according to the height and proportions of the glyphs which is meant to accompany, in this case the numerals.
    Me too. And doing that for figures usually requires no adjustment for #hashtag usage.
    The trick is to get it to work with both lining and oldstyle figures.
  • Nick Shinn
    Nick Shinn Posts: 2,198
    edited January 2021
    It would be cool if the Cyrillic-concept № were also in the standard Western encoding, it’s a fun glyph to design—not too difficult but some room for interpretation. 
  • In my experience, the written form when it is meant as "number" is often small and high, like an asterisk. But typographically its density asks for a larger rendering, and I think that works better for its now-common "hashtag" meaning anyway.
    Independently of the hashtag, as an american comic books collector (where it is obviously used a lot) I always saw the written form (in notes, from sellers, etc.) written at the same height of the numerals, or so.
    But when I refer to my Fantastic Four collection I write, say #140, when I refer to an italian publication I write n.140 or N.140. :)
  • It would be cool if the Cyrillic-concept № were also in the standard Western encoding, it’s a fun glyph to design—not too difficult but some room for interpretation. 
    In some typefaces, not many I think, it is chosen as the standard form. But I don’t recall many right now. :-)
  • The trick is to get it to work with both lining and oldstyle figures.

    /numbersign.lf/

  • John Savard
    John Savard Posts: 1,124
    edited January 2021
    In FORTRAN, if one wishes to write a large number as a numerical constant, scientific notation is represented as 4.0E+7 or 4.0D+7, the latter case applying if the constant is to be a double-precision value. Algol (at least in its publication language, but also in some implementations) included a special character, a reduced-size number "10" included in Unicode as U+23E8 for this function, apparently only because of the Algol implementation on the Soviet Buran computer. Even APL did not have a special symbol for this function, still using the letter E (despite distinguishing between the minus sign, and the unary minus within constants...and this despite the fact that APL handles operators in both unary and dyadic forms naturally).
    In an attempt of mine to define a computer language, I used the symbol # for an operator, where a#b is a * (10^b), which could also be used in constant subexpressions as the way to write a constant in this form. I felt it was reasonable to use a "number" sign to help in writing numbers.

  • Joshua Langman
    Joshua Langman Posts: 75
    edited January 2021
    Claudio, see the etymology website that I linked to above. See also the second entry here. One of the senses in the M-W definition for "hash" is "pound sign," and the first sense given under "pound sign" is the currency symbol for the British pound. Like the symbol #, this too evolved as a stylized abbreviation for libra. Thus, # (as a unit of weight) and £ (as a unit of currency) derive from the same word. Latin libra, in turn, meant "scale, balance" — thus the astrological sign.
  • Claudio, see the etymology website that I linked to above. See also the second entry here. One of the senses in the M-W definition for "hash" is "pound sign," and the first sense given under "pound sign" is the currency symbol for the British pound. Like the symbol #, this too evolved as a stylized abbreviation for libra. Thus, # (as a unit of weight) and £ (as a unit of currency) derive from the same word. Latin libra, in turn, meant "scale, balance" — thus the astrological sign.
    Thanks! But the main meaning of the word is "minced meat"?
  • Mark Simonson
    Mark Simonson Posts: 1,726
    edited January 2021
    The connection in meaning is to do with chopping in different directions, like the lines in a #.
  • The connection in meaning is probably to do with chopping in different directions, like the lines in a #.
    That’s what I suspected… great! :-)
    P.S. I’ll get back to you soon, and thanks much for all of your help!
  • It may also be related to the word "hatch," as in "cross-hatching."
  • It may also be related to the word "hatch," as in "cross-hatching."
    You’re right, “hatch marks", I see, in mathematical notations:
    https://en.wikipedia.org/wiki/Hatch_mark
    In italian, it should be simply “tacche".
  • As it turns out, the # symbol has another name—octothorpe. I found a rather extensive article on the subject here…


    The symbol's source is U2114—℔.


  • Oops. Missed Claudio Piccini's post above with the same link.