From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Typography (Rated B-class, High-importance)
WikiProject icon This article is within the scope of WikiProject Typography, a collaborative effort to improve the coverage of articles related to Typography on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
B-Class article B  This article has been rated as B-Class on the quality scale.
 High  This article has been rated as High-importance on the importance scale.

Modern vs Transitional/Old Style[edit]

The writer claims that Old Style fonts are more readable without citing any research, studies, or any other evidence to support this. It's also claimed that Modern serif fonts have greater contrast between thick & thin lines and have very heavy vertical lines, but Computer Modern has much thinner vertical lines than Times New Roman. -- (talk) 04:54, 26 May 2009 (UTC)

(1) There is no "the writer"—this is a wiki. But I agree that evidence would be good, and it might be hard to find evidence that shows old style typefaces are more readable than, say, transitional. There is however evidence that shows that modern typefaces perform more poorly, in a variety of ways under a variety of conditions. (2) Because Computer Modern has optical size variations, some of the smaller text sizes don't exhibit as strong a contrast as the larger sizes. But if one takes the 12 point version of Computer Modern and compares it to the standard Windows version of Times New Roman, the former has MUCH higher contrast. While bolder verticals are typical of modern typefaces, Computer Modern is not a particularly good example of the genre. That doesn't mean it isn't a modern, or that the description is wrong, any more than the ostrich being an atypical bird means it's not a bird. Thomas Phinney (talk) 07:44, 28 May 2009 (UTC)

What's the evidence for this line: "Europeans often consider sans-serif type easier to read.". The few books I find in my local library in sans-serif fonts are usually from US publishers. Some modern magazines in the UK use sans-serif fonts (to look trendy), but serif is considered easier over in the UK too. Surely it's to do with the serifs providing extra visual clues to the shape of the letters. -- Tarquin

As longe there is no evidence delivered, i suggest removing this statement. Fantasy 10:08 Mar 17, 2003 (UTC)
I have somewhere lodged at the back of my brain for over a decade (so please don't ask a source) that serifs guide the eye while reading. To this day I get a massive headache (and fail) when trying to speed read sans-serif, so there must be some evidence to support this. Kim Bruning 19:05, 24 Oct 2004 (UTC)

An image would be good; I'm not sure whether I know what a serif is yet, but I think so. -- Sam

I tossed one up. Feel free to improve... --Brion 07:59 Oct 21, 2002 (UTC)

I've found a glossary of typography terms at . I think that all of them should be on Wikipedia. Nikola 10:29, 21 Sep 2003 (UTC)

That link is broken; here's the corrected link to the Adobe glossary. --Bentorr 23:24, Dec. 9, 2005 (UTC)

Another remark from the back of my brain was that when carving letters, people would first put in the end points of the letters, to prevent them from accidentally chiseling past and ruining the letter. I learnt dtp over a decade ago and looked all this stuff up way back when, so this definately would need to be hunted down with proper references. Kim Bruning 19:05, 24 Oct 2004 (UTC)

Old Style fonts like Garamond and Palatino families are so beautiful. I hate the fact that stuff like Times New Roman is more common nowadays. Elegant Garmond is probably my favorite font. ThVa (talk) 11:48, 29 July 2008 (UTC)

"Elegant Garamond" is Bitstream's name for their knockoff of Granjon, a Linotype typeface. Oddly, though named after someone else, "Granjon" is based on an actual design by Claude Garamond. Thomas Phinney (talk) 16:58, 30 July 2008 (UTC)
The wiki article on Granjon points out that the italic version was not designed by Garamond.ThVa (talk) 09:30, 12 October 2008 (UTC)

Anyone else notice[edit]

Typically serif fonts are used for body text because the serifs tend to guide the eye along the line, while sans serif fonts are used for headings and for small sections of text, because they typically look 'cleaner' to the eye.

Wikipedia uses sans-serif as body text. Ah, anarchy will surely ensue.--ZayZayEM 00:29, 16 Mar 2005 (UTC)

Fortunately it hasn't. I think that's mainly because you can set a wikipedia style that does support serif fonts in your user settings.phew Kim Bruning 00:59, 16 Mar 2005 (UTC)

I think the true situation is that while serif fonts are more legible than sans when printed, sans are actually easier to read than serif on the screen, because the much lower resolution of the screen does not reproduce the serifs very cleanly. And as far as Wikipedia is concerned, the default settings seem to use a sans font on the screen and a serif font for printing, which is just as it ought to be! rossb 22:05, 14 Apr 2005 (UTC)

Sans Serif as a separate article[edit]

Why does a search for "sans serif" redirect to this page? Sans serif fonts have an entirely different classification system than serif fonts, so I think it should be a separate article. Vontafeijos 18:58, 3 November 2005 (UTC)

The reverted paragraph[edit]

First I make clear that I am not the person who wrote that paragraph.

However, what the paragraph was saying is quite clear: It was claiming that the lowercase letters which span between the baseline and the ascender line used to have what it calls a "nib" (a serif on one side of the stem) at x-height. If you know what the "long s" looks like, you should have no problem understanding that paragraph. But maybe it needs a reference or something.—Gniw (Wing) 05:53, 3 December 2005 (UTC)

I never heard of it, but it seems plausible; if there is a reference it definitely should be in the article. Nikola 22:49, 3 December 2005 (UTC)
My problem was with the structure of the paragraph itself, not the content. "The current schedule of serifs has lost one" doesn't make sense. "Nib" is obscure. It's also unclear as to what "it presumably served to carry the eye along as well" means. If all of this is indeed correct, then it definitely needs to be rephrased so that those unfamiliar with typography can understand it. Vontafeijos 05:11, 4 December 2005 (UTC)
I guess I did too many “reading comprehension” exercises in primary school; I never realized that these didn’t make sense and I just kept reading along. Here are (or were) my understanding of what these should mean:
  • “The current schedule of serifs has lost one” — If we make a list (“schedule”) of places we can place serifs on letterforms, and compare it with the corresponding list in medieval times, the current list is one short (i.e., there was somewhere serifs could be placed during medieval times, but that is no longer somewhere we can nowadays find any serifs; or, we have “lost” one place in a “list of standard places to put serifs”)
  • “Nib” — I just guessed the meaning from the context; i.e., the various references to the “long s” quite unambiguously give the meaning “a small serif on one side of the stem”
  • “it presumably served to carry the eye along as well” — a reference to the traditional understanding of the function of serifs (“to guide the eye”), and a statement to the effect of “this extra serif presumably served the same function as other serifs”
I don’t really know if the claim is correct, but I believe this is what the original contributor intended to mean. Per Nikola, this does sound plausible (judging from the structure of the “long s”), but we probably need a reference or something. (I don’t have a reference.)—Gniw (Wing) 05:48, 4 December 2005 (UTC)
I still maintain that the paragraph I reverted is too tough to understand; I read it several times and didn't fully understand what it was talking about. I agree that the content is plausible. But the kind of interpretation that you've done on the paragraph in the preceding entry should not be necessary. The article should be clear and easy to understand for anyone who visits Wikipedia. If some version of the paragraph is to be restored, I suggest that it be completely restructured, expanded, and perhaps it should include an illustration to clarify the meaning. Vontafeijos 17:01, 4 December 2005 (UTC)

A short report that p.66 of Type & typography (ISBN 0-8230-5528-0), one of the books I got from the store yesterday, contains a photograph (second photo from the top on the left side) with such serifs. In this photo, only lowercase “l”’s have them; the lowercase “d” does not have them. No date is provided for the specimen; it is only identified as “Didones (Vox, BSI)”. —Gniw (Wing) 05:41, 12 December 2005 (UTC)

Only a couple of typefaces ever made had that x-height serif on the lower-case el (notably the Romain du Roi). It was never common enough to even be called rare - it is almost unheard of. The x-height stroke on the long-s is indeed more of a nib or incomplete cross-stroke in design, rather than a serif. (The f has a full cross-stroke.) —Thomas Phinney 05:04, 31 July 2006 (UTC)

Serif != Roman[edit]

I deleted the article's equation of "serif" with "roman" on the ground that it's not true. Gniw restored the reference, saying that it is true. I believe this is an error. I base my view on the Chicago Manual of Style 15th ed., which uses the word "roman" to mean plain text, in contrast to underlined or italicized text. In fact, the Chicago Manual uses a sans-serif typeface for its examples, including examples of "roman" type. E.g., CMS 8.184, 8.187, 15.45, 17.59, 17.68-69, 17.156. The Legal Bluebook 18th ed. does the same---sans-serif examples of "ordinary roman" type, in contrast to italics or small-caps (including the words "ordinary roman" as an example). BB 2.1. The New York Times Manual of Style and Usage, under the entry for "roman," defines that as "upright typefaces." And the Adobe glossary of typographic terms, referred to earlier in this discussion, agrees---"roman" means upright, as opposed to italic, type. (For specialists, also note that the Federal Rules of Appellate Procedure, as well as many state rules of legal procedure, also clearly use "roman" to mean non-formatted text, and have requirements about serifs in a different part of the rule. See the Seventh Circuit Court of Appeals's guide to typography, contrast para. 2 (serifs) with para. 3 (roman).)

The consensus of these sources, then, is that "roman" means unformatted text; the definition has nothing to do with serifs, and roman type can have serifs or not. As far as references on style or typography, these are (in my opinion) pretty authoritative; I am not aware of any authority that says otherwise, though of course I would like to see Gniw's sources. But in the meantime, I think it's appropriate to take out the reference to "roman."

-- Bentorr 23:20 Dec 9, 2005 (UTC)

We have a generation gap here: "Roman" definitely has (or at least had) two distinct meanings, one meaning "upright", and the other meaning "serif". It was confusing (like the two distinct meanings of "gothic") but this ambiguity was common knowledge as recently as even 10 years ago.

I'm sure I can find some references, since I learnt all my typography from reading lots of books (more than 10 years ago). However, I'm supposed to be taking a Wikileave now, but I'll try to dig up sources if you really don't believe me.—Gniw (Wing) 00:18, 10 December 2005 (UTC)

The American Heritage Dictionary of the English Language, Fourth Edition states that "Roman" can be "of or being a typestyle characterized by upright letters having serifs and vertical lines thicker than horizontal lines." As you can see, this dictionary believes that roman type must be upright and must have serifs. However, I have found some discrepancies with this definition. In advanced design software (such as Adobe InDesign), typefaces can be set as "Regular," "Roman," "Italic," "Bold," "Medium," "Condensed," "Light," etc., depending on what is available for that particular font. "Roman" is not an option available only to serif fonts; it is available to Times New Roman, but it is also available for the normal version of Lucida Sans and Impact.

I propose, then, to keep the reference in, but not to make it definitive-sounding. It does seem that "roman" refers to straight up-and-down text rather than serifs more often than not.

--Vontafeijos 00:47, 10 December 2005 (UTC)

I am opposed to not making it definitive-sounding. Either drop it from the article (which I think is wrong), or just include it. If I were to choose between the two, I would say "serif type" is the less definitive term, not "roman type".

For some unknown reason, I seem to have lost most of my books on typography during my move a few years ago. I managed to find something I didn't feel to be among the best references, but should serve to illustrate my point:

Rögener, Stefan et al. Branding with Type: How type sells. Mountain View: Adobe Press, 1995. ISBN 1-56830-248-7.
  • Typefaces are divided into two main categories: roman (seriffed) and sans serif. (The term “roman” is also used to describe letters that are upright, i.e., roman and italic.)
    [ p.23, ¶2, ll. 3–7; Emphasis mine ]

(Back in 1995—exactly 10 years ago—the term “serif type” either did not exist or was so fringe that Adobe didn’t even bother to include it as an alternative. Or maybe not so fringe, since U&lc mentioned “serif” type in 1996.)

The problem with referring to the Chicago Manual of Style (a very fine reference, and I learnt lots of punctuation rules there, very useful for typography) or the American Heritage Dictionary is that they are not typographic resources and these general works tend to either get jargon wrong, or lack definitions for some jargon.

I’ll post more references here when I have the chance to fix my “missing books” problem.—Gniw (Wing) 03:26, 10 December 2005 (UTC)

It's also worthy of note that the Oxford English Dictionary does not mention serifs anywhere in its definition of "Roman." That's something to keep in mind, since we are, after all, defining a term here. --Vontafeijos 05:53, 10 December 2005 (UTC)

I still stand by my statement that “roman” at least was the correct term (versus “serif”) to refer to type with serifs, and when I read about typography it was almost the only term used. I wouldn't put too much trust in normal dictionaries when I look up jargon, even if that dictionary is the OED—To put it bluntly, if the OED does not mention serifs in its definition of “roman” (all lowercase) and it claims to explain typographic jargon, then it only shows that the OED is not a reliable reference for typographic jargon.

BTW, if you used the online OED, I just did a little test by looking up a few common typographic terms (uncontroversial ones too, I might add). It does not even have entries about such common terms as “ascender”, “descender“, “baseline”, “cap height”, “leading”, “em”, “fleuron”, or “dingbat”. I would say the OED is completely useless for looking up typographic jargon.

Some fun for you while you do the tests on the OED: (1) Try to look up “en” and see that there is no such term. There is “en rule”, which has a correct definition, but it does not explain what an “en” is. (2) Try to look up “point” and do the calculation yourself and see how many points it says there are in one inch: its definition would have been correct if it had used the inch instead of the mm, but that definition is no longer always correct in any case, plus if you haven’t read about more traditional typography you will surely think that its definition is wrong or weird.—Gniw (Wing) 06:03, 10 December 2005 (UTC)

Which online OED are you using? I found almost all of the typographic terms, including "en,", that you said were not in the OED, and they all had correct definitions. --Vontafeijos 20:51, 10 December 2005 (UTC)

I used but that’s beside the point, because I did my “homework”—I went to the design bookstore in town—today.

My observations:

  1. “Roman” meaning “with serifs” is alive and well; people still use “roman” to mean “with serifs” as if it were common sense (which wouldn’t surprise me, because it was indeed common sense, among people familiar with typography). For example, p. 62 of Type & typography, 2/e (New York: Waston-Guptill Publications, 2005) states,

    In the context of typography, the word ‘roman’ is confusingly multifunctional: it has often been very loosely applied—used, for example, to indicate an upright letterform or one with serifs, as opposed to an italic or a sans serif. [My emphasis]

  2. However, the more “textbook-like” books seem to be dropping the use of “roman” to mean “with serifs”. There is, however, no concensus as to what these faces should be called: Adobe uses “seriffed typefaces” (e.g., The Complete Manual of Typography : A Guide to Setting Perfect Type, Berkeley: Peachpit Press, 2003), while others might use “serif type” (e.g., Thinking in Type : The Practical Philosophy of Typgraphy, New York: Allworth Press, 2005).

I propose to word this use of the word “roman” as follows:

Typefaces with serifs are nowadays usually called serif type, or sometimes seriffed type. They were originally called roman (despite the word “roman” being also used to refer to upright faces), while “serif type” is a relatively recent invention. These faces are still sometimes referred to as “roman”, but such usage seems to be getting outmoded.

This illustrates Wikipedia’s inherent POV, because common sense, or unwritten knowledge, is not regarded as valid here. It is doing the opposite of preservation of knowledge.

PS: Since you also used an online OED, please tell me where you got the good copy.

PPS: To me, Adobe’s use of “seriffed typeface” (note identical term used in its 1995 book I quoted earlier) is an acknowledgement that “roman” is getting outmoded, but at the same time a refusal to believe it should be called anything else, hence “seriffed typeface” with the generic-looking adjective “seriffed”. In short, it seems that Adobe’s position is that if the correct term becomes outmoded, the thing that previously had a name should no longer have a name.—Gniw (Wing) 00:02, 11 December 2005 (UTC)

But is it common sense, or written knowledge? I agree that the term "roman" as meaning "serif" should be mentioned, but it should also be made clear that the primary use of "roman" is to refer to text without emphasis. This seems to be the case for most laypeople, and it looks like "roman" is only used to mean "seriffed" in the "in" world of typographers and designers, kind of like how journalists are the only ones to use "lede" instead of "lead" for the opening to an article.

P.S. I subscribe to the Oxford English Dictionary Online with full definitions and sources, which you have to pay to use. It's at --Vontafeijos 01:11, 11 December 2005 (UTC)

It was common sense, and it is written knowledge too. However, as is obvious from this discussion, when something becomes more common sense (unwritten) but less written knowledge (explicit definitions), it becomes harder to prove that it is written knowledge instead of just common sense, and people who are not already trained in typography would not know that “roman can mean seriffed” is a valid meaning (and a pretty common one).

Regarding your concern for “laypeople”, I’d say that serif is a typographer’s term itself, and before word processors became common, even roman (any meaning) was a typographer’s term. Add to that that there really is no conflict between my proposed wording and your concern, if you think my proposed wording is still not already clear enough, just post a revised proposal and we can work from there, since I am already acknowledging my observation that the meaning seems to be becoming outmoded.

BTW, some things I feel (kind of OT):

  1. I do feel that the OED’s omission of “roman” as meaning “seriffed” really shows that something is wrong with the OED, because this was very common usage (as seen from a layperson—I was a layperson when I first started reading, and even now I still consider myself just an amateur) and the OED missed this definition, and, because of the OED’s status, this omission is misleading people into believing that this is not a valid definition (like how you felt).
  2. I would say that if the OED omits other typographic jargon (as the free version does) the above would be rather forgivable; however, if the full OED includes other typographic jargon and the definitions are accurate (as you say), this mistake is almost unforgivable because it either borders on history revisionism (i.e., in Wikipedia jargon, a “POV edit”), or shows a kind of incredible carelessness on the editor’s part.
  3. The American Heritage Dictionary’s definition as you quoted is actually rather correct, only incomplete. A complete definition (again, writing entirely from memory), would be something like

    roman type n. 1. Originally meaning type that is both upright and has serifs; 2. Subsequently, any type that is upright, or; 3. (somewhat archaic) any type that has serifs

    Roman type originally referred to type that is both upright and seriffed (partly based on Roman inscriptions). Italic type (based on calligraphy in Italy at the time) was a slightly later invention and was considered distinct from the roman. But the italic became subsumed under the roman as a device of emphasis, before the appearance of sans serif type yet later. The historic development explains why the word “roman” would later come to refer to two different, seemingly-unrelated concepts.

I would say that “roman” meaning “seriffed” seems to be going the way of “gothic” meaning “sans serif”; Wikipedia should be responsible for preserving this meaning before it is too late (when it becomes completely outmoded and hardly anyone remembers this usage).—Gniw (Wing) 06:51, 11 December 2005 (UTC)

Wow, you guys don't mess around. I take Saturday off, and come back to pages of discussion and a special field trip to a bookstore by a writer purporting to be on "Wikileave." I'm very impressed.

Anyway, I certainly concede that Chicago Manual (and NYT Manual and the Legal Bluebook and, obviously, the Federal Rules) are not typography sources. I don't have easy access to the sources you mention, but of course take your word for it.

I would only quibble. First, I think I remember hearing the distinction between "serif" and "sans-serif" typefaces at least 10, maybe even 15, years ago, though of course memory can play tricks. (The Federal Rules I keep referring to—is it obvious I'm a lawyer?—adopted the provision for serif vs. sans-serif type in 1998, and, given how long it takes to amend the rules, must have had that change in the works for several years; they also refer to existing local rules that already incorporated the serif requirement. Given that lawyers generally don't follow the minute-to-minute changes in typography terminology, I think that probably means that use of "serif" was already common by then.) Second, similarly, I'm not sure I see the support for the idea that serif faces were "originally" called roman. Third, I don't think "serif" is an "in" term anymore, as more and more people format their own documents by computer.

One idea would be to split off "roman" and "gothic" into a separate (short!) paragraph (right now "gothic" is tacked on to the lead paragraph), explaining that those two terms at one time meant "serif" and "sans-serif" respectively, but those usages now seem outmoded.

Bentorr 23:50, 11 December 2005 (UTC)

Yes, the discussion got me to dig out my few old copies of U&lc and found that the term “serif type” existed back in 1996; whether it already existed 5 or 6 years even before that I am not sure, since I did not study graphic design and I could only read so much that an amateur could. I don’t remember seeing “serif type” back 1990-ish, but that obviously does not prove that I must be remembering correctly (as what I found in U&lc shows) or I knew all the terms (since I was just an amateur and I still consider myself one).

I still intend to take my Wikileave, hopefully a permanent one, but only after this issue is resolved. (About the trip… since I discovered that many of my books are gone and I don’t remember where they’ve gone, I would need to get myself back one or two reference books for myself in any case, and I thought I might be able to grab a copy of Emigre 69 too…)

About the last two comments,

  • (Second) You may be right. I am pretty sure I am right, but if you guys think this is contentious and want me to provided sources, you probably need to wait 2 weeks before I can go either to the bookstore or the library to look for more sources (unless whatever I got from the store yesterday happen to contain that kind of information).
  • (Third) I already acknowledged that “roman” meaning “serif” is no longer “in”. Adobe acknowledges this in its 2005 book I found, and I have to acknowledge this because I find this usage in yet a few other books (and more importantly the lack of an explicit definition containing “roman can mean seriffed” in those books). What I claim is that while it is no longer “in”, it is still very much valid. I would say it is not even quite “outmoded” yet (but in the process of becoming outmoded) if a book published in 2005 still says roman can mean serif. (The original deletion of the “roman” reference was based on the assumption that this reference is not valid, not that it is not “in”.)

What I really really don't want to see is that after a few years when “roman” really, completely becomes outmoded, and then someone found that “roman” actually can mean “seriffed”, and then proceded to create an article “English roman typeface” (c.f. Japanese gothic typeface) and keep asserting that roman is a kind of font and not a category like “serif type”. Or that people keep rediscovering that roman = serif, adding that reference to the “serif” article, but that reference keep being deleted over and over.—Gniw (Wing) 04:15, 12 December 2005 (UTC)

I wasn't clear on my "third" point. Vontafeijos argued that "roman" was an "in" term, meaning it was only used by typography insiders and was not in common usage. You responded (as I understood it) that "serif" is pretty much an insiders' term too. I was just agreeing with the original point; it's my view (based on nothing but personal experience and anecdotal evidence) that people outside the typography world but who have some basic knowledge of typography know what "serif" means, but do not know the "serif" definition of "roman."

Bentorr 15:56, 12 December 2005 (UTC)

Oh, you mean that kind of “in”. (I was thinking “in fashion”.)

Personally I think encyclopedias exist to let people look up information they don't know, either for research or learning purposes. IMHO “insider’s knowledge” should be in an encyclopedia because of, not despite, something being “insider’s knowledge”.

Let us suppose (just for the sake of argument, disregarding for the moment whether this is plausible or not) that a student is doing a project for art class on typography, would it be more helpful or less helpful if he/she read something about “roman” which seem to mean “seriffed”, but goes to Wikipedia (which he/she trusts immensely) and finds that this definition has been deleted because people thinks that this is not true?—Gniw (Wing) 16:15, 12 December 2005 (UTC)

I'm not really disagreeing with that. I just think it should be clear in this article that this meaning of "roman" is not really in widespread use. I don't object to what is essentially a historical reference.

How about changing the first paragraph of the article as follows, or along these lines. It now reads:

In typography, serifs are the small features at the end of strokes within letters. "Serif" also refers to a font that has these features. A typeface (font) without serifs is called sans-serif (from French sans: "without"), also referred to as grotesque (or, in German, grotesk) and gothic.

Proposed change:

In typography, serifs are the small features at the end of strokes within letters. A "serif font" (or "seriffed font") has these features. A typeface (font) without serifs is called sans-serif (from French sans: "without"). Some typography sources refer to sans-serif typefaces as "grotesque" (or, in German, "grotesk") or "gothic," and serif typefaces as "roman"; however, these terms no longer appear to be in widespread use.

Is that accurate? Is it supported by sources? Does it meet everyone's concerns?

Bentorr 16:28, 12 December 2005 (UTC)

I guess I am not objected to this proposal per se; I suppose it's accurate enough, though I would have liked a bit of historical information (obviously not possible if you guys think that the history is contentious).

Regarding the sans serif reference, I don't know how much should be in here and how much should be in the main article. Apparently, there are many other outmoded names for "sans serif" and, surprisingly, "gothic" is listed in Type & typography as being a "US" term (that obviously has fallen so out of use that US people generally don't recognize it at all). The funniest of all is that the book mentions a British standard (an old one, to be sure), that states that sans serif fonts are "previously" called "sans serif".—Gniw (Wing) 16:57, 12 December 2005 (UTC)

So, OK, drop it from the intro into the main article. Put in a separate section. The intro is pretty long anyway; it could use some trimming. —Bentorr 17:23, 12 December 2005 (UTC)

Since we seem to have reached a concensus on the seriffed part, I have replaced the first paragraph with the proposed replacement. I have kept the sans serif part in the proposed text, however; I think it is better to make the serif article accurate first, before worrying about splitting some of the sans serif details to sans serif.—Gniw (Wing) 06:32, 14 December 2005 (UTC)

Just a quick note on terms. "Roman" is a term that I like to avoid whenever humanly possible, because it has three distinct meanings in typography: (1) a seriffed typeface (as opposed to sans serif); (2) an upright typeface (as opposed to italic or oblique); (3) a typeface with a character set supporting common western European languages, such as one with the ISO 8859-1, WinANSI or MacRoman character set (as opposed to one that supports Central European, Greek, Cyrillic, Japanese, Devanagari, or any number of other languages, language groupings or writing systems).—Thomas Phinney 21:07, 04 June 2006 (UTC)


I had once heard that the word serif came from Sereph -- as in angels. Any truth to this? --Gadren 00:07, 25 April 2006 (UTC)

Unlikely. The OED lists serif as 'of obscure origin', but lists ceriph and cerif as earlier spellings. (Note that they all have the /ɪ/ sound after the r, whereas seraph* has /ɘ/.) It says it may have come from Dutch/Flemish schreef, but that there is little or no evidence to back up this claim. Seraphim itself comes directly from the Hebrew. porges 00:36, 25 April 2006 (UTC)


Why does this article need the attention of an expert? I'd like to remove this tag if there is no pressing need for it. I certainly don't see one. -Vontafeijos 03:36, 2 June 2006 (UTC)

I think it did need such attention (though the failings were mostly not serious). However, I have just gone through the entire article and done minor clean-up throughout. I've now removed the tag. Some of the changes include: adding more dates on when different kinds of serif type were developed; putting in the dominant theory of the origin of the serif in place of a theory I'd never seen before; adding references to Catich and Bringhurst; indicating that the Bouma theory is not widely accepted and mentioning the dominant parallel letterwise recognition theory; and defining "stress" on a letterform. -Thomas Phinney 22:24, 2 June 2006 (UTC)

Fair enough. Thanks for your help! -Vontafeijos 02:52, 5 June 2006 (UTC)

Someone tacked on a "citation needed" tag to the bit about Bouma being replaced by the parallel letterwise recognition theory, which would be all well and good except that the article on Bouma discusses this already. I would think this would negate the need for a citation; however, I'd be more comfortable if someone more knowledgeable were to deal with the tag. danis1911 (talk) 18:41, 21 November 2009 (UTC)

First, other Wikipedia articles are not acceptable sources (WP:CIRCULAR). Second, the article on bouma also lacks any sources to support the claim. Dricherby (talk) 17:07, 13 January 2011 (UTC)


Hey, the first line is either grammatically wrong, circular or both: In typography, serifs are the structural details on the end of strokes that make up letters and symbols in serif fonts (or seriffed fonts) has these features.

I do think a first-time vistor will understand what a serif is (mainly by looking at the example image), but this sentence isn't helping...

Should this be: In typography, serifs are structural details on the end of strokes that make up letters and symbols in some fonts, so called serif fonts (or seriffed fonts). A serif font is therefore simply a font having these features.

What do you think? 14:20, 29 November 2006 (UTC)


There citates serif-like things in far east fonts... How about in Arabian fonts? Do they have anything like that? How about our arabian numbers? How do they work in serif/sanserif fonts? anyway the fontmaker wishes? 19:26, 31 March 2007 (UTC)

There is something like a serif vs sans serif distinction for Arabic fonts, yes. It's more of a calligraphic vs monoline distinction, though. Interestingly, Arabic uses primarily numerals of Indic origin. Not quite clear on what the rest of that question is about, though. Thomas Phinney (talk) 23:35, 3 May 2008 (UTC)

Old Style[edit]

The explanation of 'diagonal stress' is not clear to me. Can anyone improve upon the existing text? LorenzoB 23:28, 20 October 2007 (UTC)


Schreef means wrote in Dutch (past tense of schrijven), but is also a noun, meaning line, border, stroke of the pen, serif related to the verb schrappen (delete, strike through), a much more plausible origin (according to Van Dale Groot Woordenboek, a Dutch authoritive dictionary).

Bruzzel (talk) 17:44, 22 April 2008 (UTC)

Origin: sharp corners hard to carve?[edit]

"Another explanation is that cutting a sharp corner is difficult as any imperfection in the material or carving will be obvious. Further, even minor wear will show. The use of serifs both masks minor imperfections and hides wear."

I'm wondering if anybody has a source for this "other explanation." The wear part has a slight hint of plausibility, but the "difficult" part does not - I've always found it easier to carve a simple sharp corner than to do a serif (which is a more tricky and delicate sharp corner). Of course, I'm no expert stone carver, but I've done half a dozen such projects. Thomas Phinney (talk) 23:31, 3 May 2008 (UTC)

Help with an Incomplete sentence[edit]

I believe that this sentence (from the end of the first paragraph of the 'etymology' section) is incomplete, but I'm not sure what point the author was trying to make, so I'm not sure how to fix it:

He explained that unlike the types of Bodoni's Callimachus, which were 'ornamented (or rather disfigured) by additions of what I believe type-founders call syrifs or cerefs.'

Jtmorgan (talk) 21:14, 7 May 2009 (UTC)

Venetian vs Aldine/Garalde[edit]

Can someone please elucidate the differences between these Old Style subtypes? ⇔ ChristTrekker 20:29, 25 June 2009 (UTC)


It's good that Alex Poole is quoted directly for his literature review of the subject of readability in serif vs. sans serif type. However, Kathleen Tinkel's "Taking it in: What makes type easy to read" is another literature review by an expert in this area that provides some relevant conclusions. Miles Tinker's Legibility of Print also provides some useful information. Finally, Colin Wheildon's studies in 1982–1990, captured in his book, Type and Layout: How Typography and Design Can Get your Message Across – Or Get in the Way, provides conclusions favoring serif fonts. I'm not saying that these are the final word, but they provide expert opinions on this subject as opposed to Alex Poole, who has simply provided a (useful) literature review.

Also, a distinction should be made between legibility and readability. "Legibility" is used quite a bit in this article, without explanation or Wikilinks, and the average reader will simply think that "more legible" means "better" or more readable. In that sense, the material is not that useful for the average reader. --Airborne84 (talk) 17:16, 26 December 2010 (UTC)


The article says that handwritten text doesn't usually include serifs but that capital-i is an exception, to distinguish it from lowercase-L. While many sans-serif fonts do indeed have a capital-i that is identical (or very close to) a lowercase-L, are the horizontals of a capital-i really serifs?

Serifs are small horizontal decorations, whereas in typical handwriting (at least, in my experience; but I don't just mean my own handwriting) the horizontals of capital-i's are much more like the horizontal of a captital-T. No source is given for the statement that hand-written capital-i's are serifed. To me, they seem to be an alternate letter form, as exhibited by 'a' (hand-written as a loop with a vertical, printed similar to a rotated e) and 'g' (open-tailed from the right in handwriting, often loop-tailed from the left in printing). So, are they really serifs? Dricherby (talk) 16:54, 13 January 2011 (UTC)

Would someone please add a short, clear explanation of what "bracketing" means with regard to serif? Thank you! -- (talk) 19:40, 30 January 2011 (UTC)

Readability vs. Legibility[edit]

The readability section is getting clouded with material on legibility. Readability can be measured by a number of factors, including speed of reading. Legibility is how well the differences can be discerned between letters in text. Legibility is not usually discussed in regard to speed of reading body text, for example. These distinctions should be made in the "readability" section. Perhaps the section should be "Readability and legibility." Right now, the section is very confusing to the average reader, I'm sure. --Airborne84 (talk) 02:35, 22 April 2011 (UTC)


I think it's odd that there's no mention here of an Cyrillic analogues. Obviously they exist, yet the only Cyrillic Wikipedia (Belorussian) with an interwiki link here only talks (obviously derivatively) about Latin fonts. The Russian Wikipedia doesn't even seem to have heard of the topic. I'm not sure if I should say "serifs in Cyrillic fonts" or "serif-analogues (à la 鱗) in Cyrillic fonts", seems to me that the Japanese/Chinese "uroko" or whatever can just as well be called "serifs" as "serif analogues". Sort of like saying the Japanese don't have fans, or the Chinese don't have boats, just because their fans and boats don't necessarily share history with Western ones. --Haruo (talk) 20:59, 24 July 2011 (UTC)


The term comes from the French word sans, meaning "without".

Perhaps in the ancient past, but "sans" has been in the English language since at least Shakespeare (Sans eyes, sans teeth, ...) so perhaps this could be removed. — Preceding unsigned comment added by (talk) 09:29, 17 April 2012 (UTC)

Given how common the word is in ordinary English usage, I'd say it's fair to say the term came from French. Or is there a rule that a word can't be borrowed twice? —Tamfang (talk) 19:20, 24 July 2013 (UTC)


The relation/contrast to "sans-serif" is given, but what about the term "semi-serif"? ⇔ ChristTrekker 15:14, 24 September 2012 (UTC)

I think I've only seen it in the Agfa Rotis series (Serif, Semiserif, Semisans, Sans). —Tamfang (talk) 19:19, 24 July 2013 (UTC)

Remove 'controversy' section?[edit]

This section has inappropriate weight at the start of the article, and duplicates material in the Readability section later. Overall does not add value to the article - I propose deleting this section (potentially some of the refs can be added to later sections). Dave w74 (talk) 13:25, 24 July 2013 (UTC)

I agree, I'm not entirely sure it justifies the word "controversy" either. It seems to be a collection of factoids. Bob talk 18:00, 2 October 2013 (UTC)
Agreed. This section just duplicates information elsewhere, and any additional information would be better if it was integrated into existing related sections. Jowiltshire (talk) 10:40, 23 April 2015 (UTC)

External links modified[edit]

Hello fellow Wikipedians,

I have just modified one external link on Serif. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at {{Sourcecheck}}).

You may set the |checked=, on this template, to true or failed to let other editors know you reviewed the change. If you find any errors, please use the tools below to fix them or call an editor by setting |needhelp= to your help request.

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

If you are unable to use these tools, you may set |needhelp=<your help request> on this template to request help from an experienced user. Please include details about your problem, to help other editors.

Cheers.—cyberbot IITalk to my owner:Online 09:05, 1 April 2016 (UTC)

Mitigation of perception?[edit]

Regarding this sentence in the 'Readability and legibility' section:

Hinting information, spatial anti-aliasing, and subpixel rendering technologies have partially mitigated the perception of serif fonts on screen.

According to the Wiktionary, to mitigate means to reduce, lessen, or decrease. Aren't the techniques mentioned at the beginning of the sentence meant to do the opposite of mitigating? Or perhaps they were meant to mitigate problems with the perception of such fonts?

In any case, the sentence is not referenced. Should it be deleted? (talk) 13:39, 18 August 2017 (UTC)

  • There is an excellent paper "The raster tragedy in low resolution". It discusses various aspects of font rasterization in great details, including examples with serifs. — Preceding unsigned comment added by (talk) 09:25, 6 February 2018 (UTC)