Dude, you are overthinking this issue, based on a number of unwarranted assumptions, such as “ It is natural to assume that the larger-font variant will be designed to have somewhat larger page size.” Why is this a “natural” assumption? Have you ever heard of large-print editions?

.

Large print edition? This is exactly what I had in mind. And the good ones I saw were of larger format. (And: if your opinion is that the page size will remain the same, this just means that your answer is 0%, dudillo! So it is not any indication of “unwarranted assumptions”. Especially remarkable answer, since the hypothesis that “there is an optimal number of characters” presumes that the answer to my question is exactly 100%….)

A client? I specifically said: the choice of page size and layout is completely in your hands.

If you can locate the guidelines mentioned below perhaps they might contain something useful:

The National Association for Visually Handicapped (NAVH) provides the NAVH Seal of Approval to commercial publishers for books that meet their large print standards. The standards call for:

Maximum limits on size, thickness, and weight
Minimum limits on margins
Type size at least 16 point, preferably 18 point
Sans serif or modified serif font recommended
Adequate letter and word spacing
Flexible binding recommended to allow open book to lie flat

- http://en.wikipedia.org/wiki/Large-print

> If you can locate the guidelines mentioned below perhaps they might contain something useful:

Thanks!
http://www.sfgov2.org/ftp/uploadedfiles/mod/programmatic/NAVH06.pdf

Unfortunately, this contains only hard limits.

Besides character per line, there is usually a recommended limit to line length (for normal reading within arm's length, about 7") to avoid excessive head turning when eye movement is not enough to encompass the line. Reading should not be as if you're at a tennis match. I guess the same thing would apply to posters and billboards: minimize or eliminate head turning for the normal viewing distance.

[to follow]

> to avoid excessive head turning

And long line lengths can also be a problem for people who read with a magnifying glass.

I know a couple of elderly folks with eyesight that has deteriorated to the point that even large-print books must be read with a magnifying glass. (They have macular degeneration.)

Duderino,

It is not my opinion that some large-print editions use the same page size as normal editions; it is my experience—hence, it has been empirically verified in some instances. It is not uncommon that people who require larger type are somewhat old-school, and are used to reading newspaper and multi-column-formatted magazines. Such folks tend to eschew the whiplash encountered when trying to read very wide columns.

So, you can presume all you want: presumptions can be unwarranted, as well.

"But thinking about this more, the[re] may be some other power [or] law governing the relation between the font size and line length."

The power of the reader, I think.

There are two directions and one place. The past was dedicated to the publisher's knowledge of their readers, the future is dedicated to the reader's abilities as granted by their choice of devices. The place is the present. If you are talking in the present about composition ideals of the past, the publishers knew the laws. When you are talking in the present about composition ideals for the future, the laws start with the devices.

Another way of talking about this, is that the "font size" is not the parameter we look at when determining the optimal size-to-length relationship of text type, it's the x-ht.

But is the length of extenders that unimportant?

BTW what kind of linespacing does screen type "like"? I mean compared to print.

hhp

Perhaps leading should be measured in angle of vision too.

> the[re] may be some other power [or] law

Just in case: power law ≡ x ↦ xᵗ (for a fixed value of t). It is equivalent to saying that when x increases by k%, the result will increase by t·k % (for small values of k).

> the "font size" is not the parameter we look at when determining the optimal size-to-length relationship of text type, it's the x-ht.

I intentionally used a vague term so that when answering, people have an opportunity to choose whatever measurement they consider relevant to the question. Update: actually, what I had in mind in the original post was using the same font, just of different font sizes. (My experience is only with optically scaled fonts; so I would prefer this assumed in the discussion.)

But your refinement makes a lot of sense: when considering different fonts, which of font measures is better to use in the context of my question?

> It is not my opinion that some large-print editions use the same page size as normal editions; it is my experience

Duderabimus, if I need to repeat what I already said, here it is; all you are saying is that the answer may be 0. Since I never said it could not, I do not see how what you say exposes any presumptions in my question.

It certainly is true that in practice when smaller type is used, line lengths in characters tend to be longer.

When discussing the "ideal" line length, though, generally the discussion is about a book that is optimal in other ways, such as having the ideal type size (possibly around 14 points) and it's an aesthetic characteristic.

When a range of widths - 1 1/2 alphabets to 3 1/2 alphabets, for example - is specified, while part of that is due to an "ideal" not being well defined, that also includes space for the one to use varying with the type size.

14pt? That's huge. What kind of book are you thinking about?

Calypso could work.

@John Hudson:
14pt? That's huge. What kind of book are you thinking about?

You could well be right, and 12 points would be a better value.

The kind of book I'm thinking of is a self-conscious art book of the sort that William Morris had printed.