The writer of this fairly informative set of web pages suggests 1.55 in on left and right, 1 in on top, 1.25 in on bottom. It gives a ratio of text height to text width which turns out to be the golden mean. He claims it's the shortest line length he can get away with, without the page looking weird.

http://www.nbcs.rutgers.edu/~hedrick/typography/

I'm an ameteur as well. It's always nice to have some concrete suggestions, since the expert opinion often seems to be "use the most aethetically appealing choice." I'm a good engineer of documents, but I often appreciate specific recommendations based on someone else's aesthetic judgement.

A pretty good universal solution is to divide the page diagonal into a certain number of equal parts and then use one such part as the top and inner margin (plus some additional binding gutter) and two such parts as the bottom and outer margin.

The traditional Renaissance book used the division of both the horizontal and the vertical edge into 9 parts, so you'd have a 1 + 6 + 2 division of the page. However, this gives you very generous margins that don't work well in today publications. Therefore, you could try e.g. a division into 12 equal parts (1 + 9 + 2).

If you reserve 0.2 inch for the binding gutter and divide the page into 12 equal parts, you'll have the inner margin of 0.692", the outer margin of 1.383", the top margin of 0.917" and the bottom margin of 1.833". Perhaps it's better to calculate in points -- in that case, you'd have the inner margin of 50 pt (plus a 14 pt binding gutter), the outer margin of 100 pt, the top margin of 66 pt and the bottom margin of 132 pt.

Of course, if you'd like to be a more generous, you can try to divide the page edges into 9 equal parts and use the 1 + 6 + 2 principle.

A.

I don't think you can answer this independent of the typeface involved. I personally find that Bringhurst's (and Felici's) recommendation of an ideal single column width of 66 lower case characters--2 1/2 lc alphabets is very important to readability. And he is right that more than 75 to 80 characters starts to slow the reader down.

This being the case, most 'letter' sized pages with Times New Roman have much too long a measure. So to make layout work on letter sized paper is a challenge. You can use a very wide typeface like Century School Book or even Courier--which not accidentally works perfectly as far a line length on a letter sized paper with comfortable but not wide margins.

Generally, you are better off using a wide left margin and putting titles in it, or going for a magazine style of two or three columns with a narrower typeface.

If you are going for a wider measure and a narrow typeface, then ample leading between lines and short paragraphs with extra space in between helps.

Thanks for the advice Scott. I wonder about having equal margins on the left and right. It seems somehow less elegant than asymmetrical margins, no?
I wonder if anyone else would eb willing to share the proportions that they use?
Thanks

William has offered the best advice, there is no point in having elegant margins if the measure is too short or too long.
Tim

Here's a thought-- if you're talking about one fat column with a margin on an 8.5x11 sheet, aren't we talking mainly about ruled note-paper and double-spaced typed manuscripts, really? Magazines in this size virtually always break up the space into columns, modules, etc., as suggested above. So, I would guess that although you could find a nice set of proportions in the one-column, 8.5x11 environment, that's really kind of forcing things from a typesetting point of view.

FWIW, I tend to see that kind of page as a magnified version of something smaller-- i.e., you'd be better off setting type in something closer to 14 pt in most cases, then reducing for publication (or just to 75% on your screen). With 1-inch margins, e.g., 12-point TNR often occupies at least 80 characters of linear space-- way too big.

You're right. I often make two-sides-of-one-page handouts, and on those I use equal margins, but on anything else I'll usually use unequal margins. Since I think about class handouts most often, that was what came to mind first. If memory serves, the source of my advice may have been addressing a similar situation.

Scott

Thanks everyone for your help.

I wonder though, if the paper dimensions - 8.5x11 - are not conducive to nice single column typesetting, what do you actually do, given that that is the paper standard for North America. When you write a business letter, or any kind of correspondence which will be printed on a piece of paper this size, what margins do you actually use?

Interestingly, Frank Lloyd Wright actually turned the paper horizontally, the bottom half of the left hand side was used for his mark (a black square) and his signature, and the right half of the sheet he used to make a nice size text block, but this seems inappropriate for a business/professional context.

I've been trying to use smaller type (11 pt), but maybe sticking with 12 pt would be bettr given the unfortunate measure that seems required.

My intention is to take two or three suggestions and try them out to see which ones appeal most.

Thanks everyone

I should try to address this, since I'm supposedly a teacher of professional and technical communication. Here's what I think people do to sort of fight against the less-than-ideal environment of 8.5x11:

1. Use 1.25 in margins on L and R. In fact, MS Word defaults to this, although usually publishers and editors ask for 1 inch all around. Then again, they often ask for double-spacing.

2. Use massive top and bottom margins for short letters, you know, like we all learned how to "center" a business letter between the top and bottom of the page when using a typewriter. Sticking a letterhead or logo at the top (or even the word "MEMORANDUM" in a huge size for memos) is also used to break up vertical space.

3. Add space between paragraphs.

Notice that 2 and 3 above are about breaking things up vertically; I think it's just one way of balancing positive and negative space on the page field. Paragraphs may still be very wide.

4. Tolerate a too-big sheet and just go on with life. 99% of this kind of communication is meant to be pragmatic, servicable, etc., with no real consideration of artfulness.

"For pros": Set up, as someone may have suggested above, a left-hand column of about 2 or 2.5 inches width, for logo, masthead, contact info, etc. Effectively this narrows your "body" column dramatically. Then you can set stuff in that second column with healthy top and bottom margins without it looking all skinny. This is often done on commercial and organizational letterhead, where there's a budget to do so, so it's sort of a different animal than the run-of-the-mill correspondance 1-4 refer to.

Hello,

You should try applying Villard's diagram to your page, you can read about it on Tschichold's "The Form of the Book". I used to draw the diagram in Illustrator and then transfer the margin values to InDesign (or Pagemaker, Xpress, etc.) but it was pretty boring to draw all those lines, so I've managed to determine the equations of the lines and come up with an Excel file. All I have to do now is enter the dimenstions of my sheet of paper and the number of times I want to divide it and it gives me the appropriate margins based on Villard's diagram. I can upload the file if you're interested. Another way of determining your margins is to use the method of Paul Renner as described by Jost Hochuli in his "Designing Books", but I've found the proportions to be too arbitrary, and never quite understood their origin, so I prefer Villard's (also Tschichold's) method. :)

Something important I forgot to post, the values given by Villard's diagram are determined by the number of divisions you make. The number of divisions is determined by your measure (width of your column) and for that I use Bringhurst's "average character count per line" table (p.29 of "Elements of Typographic Style"). One last thought, the values given by the diagram are not definitive, you will always have to make some adjustments in the depht and width of your columns, but it's a good starting point.

Rafael, I am curious about your excel file and/or equation.

I don't feel the need to either fill up the page or have equal margins. White space is a welcomed friend, not an evil enemy. X-height and leading have their effect as well as the chosen face.
There is no formula. You can use the classical scheme outlined above. You can also use a grid, dividing the page into 3 or 4 units wide and omitting the leftmost unit. Mostly, you need to train your eye by experimenting. This is always the better teacher than any formula. It takes longer at first but once you train your eyes, they pay dividends forever.
Also, grids don't have to be even, they can be geometric or arithmetic progressions as well. See Karl Gerstner's "Designing Programs."

ChrisL

Tiffany, the equations are simple, basic analytic geometry (though I had to take a look at a math book to remember them). I'm also uploading two pdf files of a layout I'm working on, I've used Villard's diagram as the starting point of my grid.

If by any chance you have access to a copy of Müller-Brockmann's Grid Systems in Graphic Design you can apply Villard's diagram to his cover, it fits perfectly recquiring only some minor adjustments (though he never mentions using the diagram inside his book).

Tschichold dedicated a lot of his time to rediscover Villard's diagram, he was also a persona non grata between swiss designers (specially those responsible for the Neue Grafik magazine), I suppose this was one of the reasons Müller-Brockmann didn't mention Tschichold's research, but this is only a speculation. Here you can read an excerpt of Tschichold's book.

As to dezcom's comment, I think it's really a matter of choice, I'm really addicted to precision, so I'd rather use a calculator and control every single unit of the page than to trust on my eyes only. Both approaches are valid. There's a nice discussion about these different methods of work on Lucienne Robert's The Designer and the Grid published by Rotovision.

http://pontomidia.com.br/rafael/villard_diagram.xls
http://pontomidia.com.br/rafael/sessoes_grid.pdf
http://pontomidia.com.br/rafael/sessoes_layout.pdf

Thank you, Rafael.

Interesting take on the Müller-Brockmann-Tschichold connection, or rather lack thereof.

Rafael, thank you for the sites... I recently joined the site here and posted a blog... Paul Renner - Who influenced his work and what other people had he influenced? You seem to know things along these lines so i thought you could help answer this question.