I think your professor has the right idea about forcing both the word and the footnote onto the next page.
There are several ways to do this.
Most cavalier is to introduce an extra paragraph break.
You could also increase word space selectively to "gain" an extra line.

Henrique,
When someone setting type is setting justified composition (squaring up the lines of type evenly on both sides), they are adding a tiny amount of word space to push the text out evenly to both margins. They do this to give the page a particular aesthetic that book readers are quite used to seeing. Just as they have not committed a crime, your minute adjustment to move a word to sit on a page with its footnote not only is not a crime, but fulfills a functional task as well. Fear not, the Type police will not arrest you for chosing the lesser of the two evils :-)

ChrisL

Chris, I believe that Henrique is referring to whole lines that move when the footnote itself is placed on the page, unfortunately. (Think about it: a footnote takes up it's own line(s) plus some intervening space, so it can displace a disproportionate amount.) So, the issue isn't one of tightening up to gain a word or even two. It's probably more like two or three lines worth.

This is a real challenge. I would look over the spread and even the spread before and see whether there are any paragraphs that can be easily pushed here and there, coupled with maybe running a spread a line short, to get enough of the problematic lines pushed, so that the footnote and reference can both be pushed to the following page.

Alternately, related strategies could be explored to pull enough lines to keep them together on the original page.

Basically, this kind of problem requires looking further back to find room to move.

Increasing the point size by a tenth or two for a single page (or spread) would be a last resort for me. But your professor's probably right. As much as one would hate to do it, it probably would be invisible to most. I suppose if it's invisible to your own eye, then go for it.

Of course, if you can move the lines with just a tenth or two point size increase, you can probably do it with careful word-space strategies as well.

-- Kent.

Yes, Kent. I always go back and look for paragraphs with short last lines that can be pulled up with very minute negative tracking.

ChrisL

I'd simply shorten that page by one line without any adjustment to the other lines, the way it's customarily done to avoid widows and orphans.

Changing the text size and thus the line height messes up the baseline grid in case there is one -- which should be the case for a proper document.

The usual convention (at least in the States) in a book with footnotes is all pages have the same depth, and unavoidable whitespace is put between the text and the note. Now if footnotes are few and far between, this won't work, and you have to revert to making up pages as you would in a book with endnotes.

The laying out of pages part of composition is just as important as the H&J. To me, the obvious solution is to go back to earlier pages in the chapter and look for a place you can "make a line," so the problem word appears on the "next" page regardless. In this case, "saving a line" earlier in the chapter will work only if you were almost there anyway.

There are any number of ways to make or save a line, depends on both the copy as set, and the layout program. But you have just encountered an example of what I try to tell designers who've never set type: it is all compromise, and these compromises cannot be solved by applying a hierarchy of rules. If they could, we would write a computer program to do the work; case-by-case human judgment wouldn't be needed.

Of course, like any puzzle, one solution may create another problem. Just keep working.

BTW, the one thing I haven't see mentioned is to slightly widen the margins on the offending spread. You can easily take the margins from, say, 25 picas to 25 picas three points, and no one will be able to tell. If that saves a line, you're home free. Alternatively, narrow the margins, again by 3 or so points, which might make a line.

Welcome to the world of composition.

* * *

Edit: You might also want to make a note of any changes you make to solve pagination issues. I don't know if they teach this in school, but in the real world, there are "Author's Alterations" and "Editor's Alterations," which occur after you finish. Actually, you didn't "finish," that was just "first proof."

It can easily happen that an author or editor will change the copy so you'll want to take out an earlier fix. BTW, after first proof, you'll want all your reworking to keep the copy on each page the same, if possible, Likely the publisher or author has just paid considerable money to have an index made up from first proof, and changing what words fall on what pages could affect the index.

However, sometimes an alteration means the text just has to change pages. For that reason, I prefer to change the characteristics of the page -- i.e., the margins, over the copy that (at one time) falls on a page -- i.e., the type size.

Did I say welcome to the world of composition?

Those are all great tips, charles_e!

Wow! Thanks, guys!
You've given me some invaluable help here!
I've read Bringhurst's book, but I hadn't realized I could use the same techniques to avoid orphans in this case, which is kinda obvious to me now.
Nick, Chris, Kent, Henry and Charles, thanks a lot!

Henrique