Feel better?

Sounds to me like you should be using typefaces that have obliques (slanted romans) and not true italics.

Italics per se are narrower than text faces, that’s a fact of life. That said, there are enough typefaces that have ‘wide-ish’ italics or (as memorized by kazoile) obliques that count as italics but have in most cases the same widths as the roman. Ergo: there is a subset of typefaces that is to your taste.

(Does this help?)

I agree. But, other than for historic reasons, one can argue that because italic is usually narrower and lighter, it will stand out more in an uniform roman text, thus serving its emphasis role better.

Looking at my output, it seems that the italics are generally 95-100% the width of the roman.
The exception are those which are old style.
So I think what you are really objecting to is the currency of the most popular old style "text" faces, such as Bembo, Perpetua, Baskerville (transitional, but still "pre-modern") and Garamond, and recent exercises in the same genre such as Arno, Dolly and Scala.

With Georgia, in which this site is set, the italic may even occupy more space than the roman.

Historically italics did not have any connection whatsoever with romans — they were a whole different species of type. The fact that some cheap typesetter combined the two mayor species apparently puts us in this bind now…

@Nick I guess one needs a certain amount of pixels to obtain a good result, thus the width of Georgia’s italics (?).

@J Tillman. Are you reading printed text or something on screen as you type? and using what application? Some applications don't properly show or even print italics, particularly if they are large opentype families that don't nicely fit the old Windows 4-unit limit on a family system.

I guess what I am saying is, is this a selection problem caused by the users application or is this a font design issue? As Nick mentioned above, genre of typeface can, to some degree, influence the italics. Most modern designs don't have wimpy italics. Identical metrics from Roman to italic is not really a good thing. The glyphs have different shapes so the fit must accommodate them. We all know that putting something in Italics is not a way of saying "IGNORE THIS". Most type designers spend more time designing Italics than they do Roman, even though the amount of usage is far less. All Type designers I know design the italic to work well with the Roman, not to meekly and subserviently coexist with it as an impotent step-child.
Do you have some specific examples of typefaces with Italics that don't hold up? Maybe, if we see what you are looking at, we can be more helpful?

Maybe not relevant in this context, but not so long ago, Actium has been released,
a sans serif that has italics with approx. the same colour as the Roman:

http://new.myfonts.com/fonts/typemafia/actium/

Some comments on it by the designer can be found here:
http://typophile.com/node/75996

@Nick I guess one needs a certain amount of pixels to obtain a good result, thus the width of Georgia’s italics (?).

Except for Unibody!

if you want italics that feel just like the roman, then why use italics at all? if there is not enough of a visible difference, then they lose their purpose of being. some say that narrow italics are meant to actually slow the reader down, i would assume this would be a good thing if you wanted to emphasize something or point out a borrowing from a foreign language.

Hmmm. Following up on the line of Paul's comment, I wonder if you are using italics as intended, or stretching their use beyond what is appropriate.

By that I mean, for short phrases for emphasis, or at most a few lines. "Companion italics" are generally designed to work that way, and if you try to set extended text in a companion italic it is going to be sub-par. You do have a lot of other options, including text blocks indented on both sides, bullet points, lists, etc. If you keep italics to use as intended I think you can find a lot of good options.

Tillman, I partially agree with you: there are many fonts with so much narrow italics. When the glyph width change, there is a cadence break. Although this break is actually desired (otherwise you would not be using italic), it may become excessive when roman and italic have very different widths and are mixed in the same paragraph.

Please note italic naturally tend to be a bit narrower due to its calligraphic shapes –not to mention the fact they are initially conceived as independent sets of types, as Bert pointed. An italic style with exactly the same width of roman could feel actually wider than the roman.

Many contemporary fonts already use a wider italic. Anyway, this is not a choice for all fonts. For example, revival from a 16th Century type would become historically incorrect (and thus feel strange) without narrow italics.

Some fonts I remember have wider italic (although still narrower than roman): Tisa, Guardian, Charter, Skolar, Utopia, Gina, Yukatek and Centro Serif.

It seems to me that the problem here isn't the typefaces but the notion of emphasis. See what I did there? I used italics for articulatory emphasis, a linguistic phenomenon, not for graphical emphasis, a visual phenomenon. I've seen plenty of style guides that blithely state 'Use italics [or bold] for emphasis', without defining what they mean by emphasis, and certainly without recognising that there are different kinds of emphasis. The practice of using italics for emphasis originated in articulatory typography, i.e. italics were used -- among other things: citations, foreign words, Latin abbreviations etc. -- to stress words or phrases, the typographic equivalent of the way we would stress those words or phrases in articulated speech. In technical writing, it is more often the case that emphasis needs to be used to draw attention to some piece of information; this is graphical emphasis, and bold type is more appropriate than italics of any style or width because the heavier weight draws the eye. If a style guide has boxed itself into a corner by using bold type for something other than emphasis, then you need to look to some other eye-drawing solution, e.g. semi-bold sans serif, reserving italics for their traditional functions, which are already too numerous without also using them for graphical emphasis for which they are seldom well adapted.

Thanks to everyone who has commented so far. My position is unchanged. (I told you this was a rant at the beginning!) 1. When I see emphasis using smaller type it makes me want to groan. 2. I think foreign language words deserve the same weight (typographically and figuratively) as English words.

I'm seeing two very different responses here. One group is saying that there are lots of good fonts where the italics are almost the same size and weight as the Roman, and I should just use one of these and be happy. The other group is saying that italics are just smaller than the Roman; it's always been that way. Not to fear. I'm gonna argue with everybody.

To Nick Shinn and dezcomm: Yes there are a lot of good fonts where the italic size and weight is almost the same as the Roman. And I think the type design community should just take a deep breath and bump those italics up the final 5 or 10 percent so that that they are the same size and weight as the Roman. And then we can all be happy. Or maybe just me.

To paul d hunt and Bert Vanderveen: We may have to agree to disagree here. I recognize how it has been. I just think it's time for a change. Italic is different than Roman. Even if they are the same size and weight, they're still different.

To JoergGustafs: As you implied, I am mostly talking about serif fonts. Nonetheless, the sans serif Actium is very interesting and is an example of what I want. But I'm really talking about serif text fonts.

To William Berkson: Yes, I'm talking about what you're talking about, namely short phrases of two to four words.

To Dezcom: Take a look at your post. Imagine that instead of having capital letters that say "IGNORE THIS", you have letters that are actually much smaller than everything else. Does your post seem a little incongruous?
What I am looking at is printed material. In most cases it is a printed version of the specimen PDF. There are a lot of problem fonts; that's why it's a rant. But the one that just makes me shake my head is Greta Text. This seems to me to be a great font in the Roman, not only for newspapers but for general purpose. Why is the italics so small?
http://www.typotheque.com/fonts/greta_text/

To Freiberger: Of the fonts you mentioned, I have been looking at three for some time, FF Tisa, Skolar, and Yukatek. (Yukatek in the still-in-progress form on the Reading site.) Now if those italics were a little wider....

To John Hudson: Your articulatory emphasis is getting a graphical boost due to the font in use here. As Nick Shin pointed out earlier, this has a significant italics. And I stand by my feeling that emphasis with small letters (smaller than the Roman) doesn't cut it.

"...Imagine that instead of having capital letters that say "IGNORE THIS", you have letters that are actually much smaller than everything else...."

If I want all caps, I can choose all caps. If I want either Italic or bold or smallcaps or petite caps or even (heaven forbid) underline, I can choose either. You are arguing about what the typographer has chosen and you dislike, not what is available to some other typographer, if they so desire. The only solution for satisfying you is for you to design, set, and output it all yourself.
The person who chose Greta probably liked the contrast between the 2 styles. Feel free to set all your own work and either buy the types that suit you from the millions available or fashion your own. In this day and age, it is all possible. You can also set all the text in Roman and use the accent color of your choice for the emphasis words. Multicolor printing is much more economical today.

Interesting analysis, John. I think a problem may be that, as you say, technical text calls for a wider palette of options.

Mr. Tillman, I'm not sure what you mean by "small". Companion italics are often both lighter and narrower than the roman, but the combination can achieve the same typographic "color" as the roman, while having a very noticeably different texture. The actual x-height can be the same or bigger.

While I agree that some companion italics are too hard to read, I think it is difficult to isolate one variable as "small", when the different variables interact to get the overall affect.

In the case of my Williams Caslon Text, I made the italic a lot wider than Caslon's original, similar to what Carol Twombly had done for Adobe Caslon. But I also made mine more upright than the original, and upright than Adobe Caslon as well. I think reducing the slope makes it a little more readable.

But there is a delicate trade-off here. If you don't differentiate it enough, it's not going to function to highlight emphasis. And since almost all differentiation from the roman makes a type less readable, if you go to far in differentiation, you start to compromise readability too much.

I agree that getting it right is a challenge, and some, like Joanna are wildly off. But I don't think you are taking into account that you can go the other way too much. The idea of replacing italic with slanted roman, in serif type, has been a flop. The most striking "disproof of concept" is van Krimpen's Romulus, where a wide slanted roman is a flop as a companion to a beautiful roman.

...bump those italics up the final 5 or 10 percent so that that they are the same size and weight as the Roman.

That doesn't make visual sense.
You want an equivalence in emphasis between roman and italic.
But there is no such discrete graphic quality as "emphasis".
There are only things such as area, width, and length.

When the letter l is skewed, it doesn't remain the same length, but becomes longer, and increases in area.
Effects such as this make it appear bigger.
To counteract this and achieve an equivalence of emphasis, it is necessary to make adjustments, one of which, that seems to have proven quite successful, is to make the italic occupy less overall width.

Related:
http://www.typophile.com/node/29998

And, since Hrant isn’t very present around here lately, I’ll report his opinion too:
http://www.typophile.com/node/63847#comment-377131

@Tillman - Maybe you'd be best served by Computer Modern (CM Unicode or Latin Modern), then. That one is intended for math/tech writing, and you get slanted romans as well as italics.

Thanks again to all those who have commented or may comment in the future.

To Nick Shinn: Sorry I misspelled your name in my earlier post. Re: the two boxes, I have no doubt that I used the wrong technical words, resulting in the diagram you showed. From a less technical perspective, how about if I say I am looking for an italic that has the same horizontal distance as the Roman.
You mentioned Georgia in an earlier post. When I looked at the font closely, yes, this is the italic-Roman relationship, or very close, that I am looking for. (You should have hit me on the head harder.) Maybe the italic is a shade too obvious, even.

And the whole point of this rant, re-stated now, is that the Georgia italic-Roman relationship should be more the rule, and not the exception to the rule, for text fonts. Is Georgia considered an unsophisticated font because of this? or is it considered inappropriate for general use because of this? (Obviously, as a Microsoft standard font, it's very widely used.) Why isn't everybody here already? I am genuinely curious about this.

To John Hudson: I will be mulling over various options you have suggested. There's a lot there to think about. But I do not think that this is only a technical writing discussion. It's about all writing.

To riccard0: Thanks for those two links. The first is an interesting thread. And I liked Hrant's "Classic, schmassic" post. Also, now I know how to link to a particular post.

To Té Rowan: Thanks for the suggestion; I will take a look at that.

Without doing an exhaustive comparison of recent text typeface families, I'd say that the general trend is probably in the direction that you desire, i.e. italics are tending to be wider, closer in proportion (and more directly related stylistically) to their roman companions than was historically the case. [Remember that italic types did not originate as designed components of families, as they are used and typically produced now. Italics originated as distinct typefaces, that only gradually assumed supporting typographic rôles. The essentially independent design of italics persisted into the late 18th century, and almost any historical revival typeface will reflect this.]

PS. You might like my Constantia: the italic is only very slightly narrower than the roman.

one can argue that because italic is usually narrower and lighter, it will stand out more in an uniform roman text, thus serving its emphasis role better.

(And others)

If you had spend your university years in a philosophy department rather than wasting them in design, you'd have learned you can argue for anything.

There are very few conclusive arguments. So, we strive for compelling. After such exchanges, usually no one's mind has changed.

Is this thread a search for an argument, or a typeface?

...one can argue that because italic is usually narrower and lighter, it will stand out more in an uniform roman text, thus serving its emphasis role better.

Bill Bernbach's ad maxim, in the contrarian, absurdist, zen style of the 1960s: "When everybody else shouts, you whisper."

Italics are now used for emphasis...

But that's not their only use. When I teach about italic design I emphasize that the purpose of italics is contrast. That can be for emphasis or for quotation (see above) or commentary or to set apart a word or phrase as different somehow, such as a foreign word. In the first sentence of this paragraph I've even used italics for two slightly different types of emphasis. It would actually be more effective to do as John did and say:

When I teach about italic design I emphasize that the purpose of italics is contrast.

There are a variety of techniques that can be used to provide that contrast, and horizontal compression is one of them. So is a lighter weight. Or alternate letterforms. Some of these are more effective than others for a particular purpose. The bottom line, however, is that the contrast must be clear and strong. What bothers me is not which technique is used, but when the contrast is too weak. That's my rant.

I am surprised that nobody mentioned Sabon so far. A Garamond interpretation where italics and romans have equal width. Or to say it with Nick Shinn:

One of Sabon's most distinctive features is the fact that Roman and Italic have the same metrics--a criterion of Linotype "duplexing", as I understand it.

http://typophile.com/node/53558

I second the recommendation of Sabon. When I'm doing legal work, I use Sabon for all my court filings in part because filings require so much italic (not for emphasis, but because italic is used in legal citations). I think Sabon is great anyway, but the wide italic really does look nice in that context.

I disagree with the recommendation of Computer Modern. Don Knuth has made many impressive contributions to digital typography, but that's not one of them. It's ugly and awkward. You can't justify it when there are so many better choices.

Electra is a typeface with two italics: one a sloped roman, one a true "cursive" italic. I don't know off the top of my head whether the sloped roman is available in digital.

In general J. Tillman raises a good point and some of the responses here are underselling the value of what he suggests. When I'm in graphic-design mode, there are plenty of times I'd like to have the choice between two italics.

Forget about sloped romans for a moment, which I agree are too subtle. I love Galliard Roman, but Galliard Italic is too aggressive and stylized for certain projects. So I can't use Galliard. Meanwhile, Charter Italic is more moderate. But Charter Roman isn't as nice as Galliard Roman. If I had a Charter-like italic for Galliard, I'd use Galliard more often. (You could also consider the difference between Georgia Italic and Miller Italic, which might be more apt since Georgia & Miller are derived from the same DNA).

Really it's no different than having the choice between semibold and bold. Sometimes you want one, sometimes the other. So why don't we have more "semi-italics"? Clearly because semibolds are often just interpolated, which is an afternoon project in Fontlab. Semi-italics would have to be drawn from scratch.

@butterick - Why would anyone want an Austin Gipsy with the Chevrolet Tahoe on the market?

You are right, Té, Why would anyone choose to park a Chevy Tahoe on a congested city street when a Mini Cooper would be so much easier? ;-)

Drive a tack with a sledgehammer or a railroad spike with a tack hammer. We choose the tool to suit the job, not to suit some abstract notion of "like better in general."

Of course, I'd prefer a Suzuki Fox (the local name for the SJ410) over the Mini, but that's due to our roads.

You have a REAL Winter!

Oh, plenty of very real winter elsewhere. 'Least no-one's frozen to death here yet. Also, the Mini just can't deal with our F-marked roads.

In the 60-70s, I drove MGBs. They were horrible in snow. Either British Leyland or Morris Garage was not used to anything like REAL winter :-)

I'll try to make time to read beyond
the initial post, but in the meantime:

There are at least two good reasons (beyond
any historical baggage, which I'm no fan of) for
an Italic to be narrower than the Roman:
- Uh, emphasis! :-) Something doesn't have to
be bigger to get attention - it can be smaller.
You just need contrast (and it's important to
note here that slant isn't the only thing).
- In a traditional serif system the Italic has
far fewer serifs than the Roman, so it simply
can't be as wide without looking dorky.

BTW, if you're thinking slanted-Roman, you're
reading my mind. To me -when done right- it's
the ideal approach.

That said, as many here know I'm actually not
a fan of Italics in general! And I think there are
some better avenues of emphasis, which have yet
to be properly explored. One of then is weight,
just not massive weight like with a Bold; I think
Demi weights usually work great for emphasis.
Another one I quite recently thought of is to in
fact increase size, but not based on point size,
instead based on x-height (so apparent size).
http://typophile.com/node/81755#comment-462531

hhp

My take on the original post is thagt it's a complaint about the slight narrowing of italics in ordinary and contemporary typefaces, such as Times Roman or Century Expanded.

Had the original poster seen Bembo with its more historically-appropriate italic, for example, presumably he would have made an off-the-scale comment about this.

However, his rationale is largely valid. Italics are used to emphasize words in line or indicate foreign words. They aren't used to make inexpensive editions of books possible by allowing more characters per line. So italics shouldn't be excessively narrow.

But I think that problem was solved hundreds of years ago, and only italics like the Aldine ones pose a real problem in this connection, not those of ordinary recent typefaces.

> italics shouldn't be excessively narrow.

Sure.
But I think he was complaining about any narrowness at all.

> I think that problem was solved hundreds of years ago

The problem I have in mind hasn't been solved yet...
How do you avoid "skewing the voice" of the Roman?

From my Daidala interview:
"
If you believe that a typeface carries emotional associations in its forms*, then you have to conclude that the conventional practice of having only italics for emphasis is one-dimensional. There can be many kinds of emphasis, many nuances a discerning typographer might want to give snippets of text. So what would be nice is a system of fonts than can nest into each other and convey various emotions, such as stress, sensuality, whatever. What do italics convey? Mostly informality and motion – not necessarily what the text being marked is about.

* And if you don't, how do you justify making yet more fonts?
"
http://www.daidala.com/25apr2004.html

hhp

@hrant:
But I think he was complaining about any narrowness at all.

I agree; I said as much in my post.

As to your suggestion in the Daidala interview: that is an interesting idea, but I can see why, in general, designers have not gone there.

It's hard enough to see that Times Roman italic is Times Roman italic; so, while it is possible to make new typefaces that convey different emotions, making ones that are clearly associated with an existing Roman is significantly more difficult.

Also, while writers are used to using the stereotyped forms of italic and boldface for emphasis, anything else is likely to be stigmatized as gimmicky.

There is some existing precedent for this; besides italics, bold, and small capitals (which I've argued should be thought of as a third case) there is, of course, demibold and extra bold; some typefaces have script versions and others have monospaced versions.

But what first came to mind when I saw your suggestion, and which made me think right away that a "gimmicky" stigma would be likely, was that there would be a weight that makes the given face look somewhat like Ondine. And another possibility would be a weight replete with Valentine's Day cliches.

> It's hard enough to see that Times Roman italic is Times Roman italic

Really? And if so, use a darker weight.
I'm a big fan of using Demi for emphasis... with no slant!

> anything else is likely to be stigmatized as gimmicky.

1) Conservatism will stigmatize anything that nudges its Lazy Boy.
2) As with anything else in text typography, it's all about being subtle,
under-the-radar. The reader should have no idea what's going on, much
less stigmatize it.

I think most designers haven't gone there because
like the rest of the population they're 95% followers.

hhp

Hrant: I'm a big fan of using Demi for emphasis

See my comment earlier on the distinction between linguistic emphasis and visual emphasis. One expresses an articulatory emphasis of a word or phrase, the other expresses a hierarchy of importance. One should be encountered in the text as one reads, the other should draw attention to itself and stand out from the text.

Important I can't stress enough how critical is this distinction.

By the way, the newly released Brill types have a bountifully wide italic, with an only slightly narrower set width than the roman. It also exhibits a number of other unusual features in how it relates to its roman companion. The types are licensed at no cost for non-commercial use.

To me that distinction goes to the level of deliberative
versus immersive reading; meaning that if it stands out
of the text, it can't do with an "Italic" is supposed to.
Which is not to say there can't be different flavors of
non-disruptive emphasis.

Brill: Sounds interesting - I'll definitely be taking a look.

hhp