Slashed zero with downstroke?

Johannes Neumeier

I've been looking into legibility issues and some such characters which quite easily can be made explicitly unmistakable. When looking at slashed zero variants there is a chance of replacing one confusable character with another by adding the slash or dot (found this to be an informative paper).

Is there a reason why slashed zeros are not more commonly drawn with a downstroke (like in this font, for example) and instead the variant with the slash is used that resembles above all else the slashed O and empty set / average sign? It seems to me ironic that a feature intended to reduce letter identification errors might actually increase them, so I was wondering if there is an equally problematic catch with the downstroke I am not aware of.

John Hudson

The modern slashed zero developed in an Anglo-American programming context in which confusability with Ø wasn't an issue. I do like that variant with NW–SE slash though.

I say modern slashed zero, because there are manuscript slashed zero examples dating from the first century of introduction of the zero into Europe. See page 5:
http://tiro.com/John/SameDifference2-DECK.pdf

Hrant Հրանդ Փափազեան Papazian

Since a slashed zero is a drastic way to avoid any confusion, I also greatly prefer the backslash (not "downstroke", please :-) form. Those Anglo-Americans should have been able to foresee that. Here is a nice list compiled by Studio Het Mes of fonts that feature a backslash zero.

BTW that Bigelow piece is great but it's missing arguably the best general form of OS zero: weighted only on one side (with my preference for the left). See the one in Whittingham.

Christian Thalmann

I don't like the idea of slashing the zero along the «wrong» diagonal; in a humanist typeface, it would have to look either wrong or very heavy. I prefer the dotted zero in my typefaces, e.g. in Cormorant:

not "downstroke", please

If anything, it's a bend sinister.  ;o)

Hrant Հրանդ Փափազեան Papazian

Heraldic terms always welcome!  :-)

Concerning your point above, I would say maybe humanism doesn't have to be so shackled by one particular –not to mention obsolete– marking device...

Nick Shinn

The dotted zero can easily be mistaken for 8 in many typefaces.

Hrant Հրանդ Փափազեան Papazian

Or a Greek Θ.

Thierry Blancpain

Differentiating similar-looking characters is only important when there’s a realistic chance those characters might appear next to each other in a situation where the meaning isn’t already very clear by its surrounding. Not too many people are going to read "Illinois" in Helvetica and think the state is actually called "Liiinois" or "Iiiinois".

I personally have never seen an predefined code to be copied or entered by someone* that might either contain Greek Θ or Ø, and thus the slashed 0 is usually perfectly fine. It differentiates it from the uppercase Latin O and that’s all that is needed in that situation. This might be a little less perfectly clear in Greece / countries using the slashed O in their language, but as mentioned by others, there are further considerations like pen logic that play a role, too.

*which is imho when this kind of legibility is really important.

Hrant Հրանդ Փափազեան Papazian

@Thierry Blancpain
An "unrealistic chance" can creep up on you... Although it's rarely worth sacrificing too much style for a remote possibility of confusion, the question here is about how to best alter a regular zero for extreme cases. Why would a potentially confusable forward slash (or dot) be preferable to a non-confusable backslash? Pen logic? How many people complain about the diagonal of the conventional contrasty "Z"? Rather than being overly sensitive towards an imaginary marking device I'd rather be sensitive towards actually existing Greeks, Swedes, etc.

BTW "all that's needed" generally leads to things like Arial...

Ray Larabie

I'm working on a technical style typeface where the O and zero are the same and I used a backslash. But it's not just the orientation of the slash that differs from the Ø. The Ø's slash goes all the way through the circle while the zero slash only goes through the counter.

Johannes Neumeier

That Whittingham example is interesting - I was wondering if the unmodulated zero (does it have an actual name?) would come up in this discussion, but this lopsided solution is new to me. That unmodulated zero though does speak to the necessity of such a distinction, even outside of programmers' terminals.

I would agree that in contrast rich humanist faces the weight of the stroke in the reversed orientation might become too heavy - then again you could ask if a special version of zero does not do that either way?

Good to see, however, that no obvious reason has been listed not to use this version of a slash zero in other cases, and that, in fact, it's not as uncommon as I suspected. Thanks for your input!

Hrant Հրանդ Փափազեան Papazian

The classical contrastless zero is in fact an extreme "style-violating" case of differentiation (and I'm no fan of it). But I suspect it was more in service of people who had to set/distribute metal type by hand, as opposed to readers. To me when designers today use that form it smells like the rotting clothes of long-gone French royalty.

Johannes Neumeier

Hrant H. Papazian said:

The classical contrastless zero is in fact an extreme "style-violating" case of differentiation (and I'm no fan of it). But I suspect it was more in service of people who had to set/distribute metal type by hand, as opposed to readers.

Interesting indeed. I see the logic behind your guess, but wonder if there is any source that could shed light on this, one way or another? It ventures somewhat from the topic, but I'd be very curious to learn more of the inception of that type of zero. If anybody has any pointers for what or where to search, I'd be much obliged

Kent Lew

Didn’t the representation of our modern concept of zero as numeral/figure basically start out as a contrastless symbol from the beginning?

I would think the relevant question is when in history did contrast get added to harmonize it with the rest of the written/typographic figures?

(Which then necessitated strategies for differentiation from Oo in certain contexts.)

Hrant Հրանդ Փափազեան Papazian

Did the people who invented the zero symbol even use contrasty marking tools?

In any case it would be interesting to learn the history, but I can't discern how it might help actually design it for the living.

John Hudson

Re. historical forms of zero in European palaeography, see page 5 of my Serebro Nabora presentation: http://tiro.com/John/SameDifference2-DECK.pdf

The illustration used there is from G.G. Neill Wright's The Writing of Arabic Numerals (1952).

Nick Shinn

No doubt the contrastless zero is style-violating, but it is kind of painted into that corner by another kind of style violation—namely, the vertical stress of 6, 8 and 9, which is not in keeping with the angled stress in the bulk of the broad-nib informed lower case.

This means that sequences such as 606, 908, etc., and of course 00, would have have formed nasty clots, in the normal angled stress—and 1100 been extremely uneven in colour, especially considering the monowidth (“tabular”) constraint—had the zero been rendered like the letter “o”.

Another style violation—reverse contrast (e.g. in Stempel Garamond) also addresses this issue, but has never caught on.

And the contrastless zero is not totally odd, as the oldstyle four also has thin strokes going in different directions.

When one considers the way the thick and thin strokes are distributed in the classic oldstyle figures, and they way that numbers comfortably differentiate (in monowidth) in mixed case setting as being slightly “tracked out”, the design of the oldstyle figures, including the circular, mono-thickness zero, is really a very clever system, and much better than the didone.

Hrant Հրանդ Փափազեան Papazian

> sequences such as 606, 908, etc., and of course 00, would have have formed nasty clots

You mean like "book"?  :->

Instead of making a really bad zero, let's fix deeper. After all "6" and "9" are even more confusable than with zero.

Nick Shinn

No, not like “book”, much worse!

Compare <9-letter_o-6> with <9-zero-6>.
The fat part of the long strokes of 9 and 6 coincide with where the join is, close to the extremum.


And on a purely semantic level, the negative centre of contrastless zero, its round nothingness, is promoted above the value of its positive component, which merely defines a boundary with minimal presence, and is not thus the trace of pen strokes.

Hrant Հրանդ Փափազեան Papazian

Sure, it's worse (although your "o" having vertical stress unlike the "b" and "d" is a cheat) but also less damaging; my point there was to rethink more.

I'm a huge fan of differentiation, I just think there are more clever –and less distracting– ways, like giving only the zero horizontal contrast.

Christian Thalmann

Historical or pen-logical considerations aside, am I the only one who to find a monolinear /zero just jarringly out of place in a contrasted font? I'd sooner have a reversed-contrast /zero or even one of those stressed only on one side.

Thomas Phinney

I agree that the monoline zero is jarring in a font with significant contrast. Although I prefer conventional stress over any of the other solutions, I will certainly concede that there are times when a more-easily-distinguished zero is important.

In the end... having an alternate for the zero so one can have it either way (monoline or conventional stress), that would be great!

Nick Shinn

I don’t think it’s an issue at body text size, Christian.
In display work, oldstyle is rarely a good option.

Adam Jagosz

I can live with a no-contrast zero (I'm fond of it because when I first saw it in some book it looked pretty old style – like a relic from the times when they "didn't know yet how it's done"), but it must be rather perfectly circular! @Nick Shinn The picture you uploaded looks like the zero is an /o taken from Arial! It looks like it only differs from /o in the outer path and not the inner.
Now for something completely different, is it normal for a slashed zero to have the slash strike through the glyph like in /oslash? I'm used to slashed zeroes where the slash doesn't cross the body (most monospaced fonts). To me a slash extended beyond the oval looks like an error, a grave one. It however happens in some fonts, now I observed it in Linux Libertine.

Edit:
I looked into the document John Hudson provided, and page 5 suggests a strike through zero can be justified historically... I'm still not fond of it.

Christian Thalmann

Nick Shinn said:

In display work, oldstyle is rarely a good option.

Ah? Why is that?

Nick Shinn

Firstly, in all cap setting it’s much too obvious, to the point of appearing to be either a mistake or very precious. But it can be sweet with small caps, pointedly quaint.

Secondly, in mixed case setting so much depends on the particular number. 1960 is always a good year, but 100% doesn’t carry its weight.

And of course, the issue of how much zero looks like the letter “o” is magnified.

I also don’t like “f” ligatures in display settings; nuances designed to help difficulties disappear in the flow of small text have the opposite effect in display, with their mechanism laid bare. 

Johannes Neumeier

Christian Thalmann said:

Nick Shinn said:

In display work, oldstyle is rarely a good option.

Ah? Why is that?

Well, it goes against the whole rationale behind the inception of lining figures - oldstyle figures looked too small and unbalanced when lettering posters in vulgar all caps. That's not to say you can't, of course.

I agree with Adam, the slash extending past the zero's round outer shape looks like a mistake. The historical handwriting examples showing slashes through the zero that extend beyond, and that's speculation on my part, might just be the expression of swift pen flow. Confusability with ø was likely less an issue than it is now in our multilingual information society, and typefaces that cater to its demands.

Since I've been paying more attention to this, I've also found examples of slashed zeros with a horizontal stroke (or even just a centered dot). In my opinion that, too, replaces one confusability for another, by making the slashed zero close to theta.