There are so many variables involved, I don't think it's possible to isolate one. If you were to conduct a thoroughly empirical experiment, you would vary word space against each of these in multiple combinations:

-line length
-type size
-type style
-type face
-type weight
-tracking (letterspacing)
-leading
-word length
-language
-ink color/paper stock

and that's just in the type specs. You would also have to vary text, and readers' age, education, sex, etc.,, and the reading environment, if you wanted to come up with a legitimate general principle.

You could however focus on a specific relationship. One that concerns me as a type designer is the amount of contiguous white space in glyphs: for instance, consider a word ending in "t" and the following space: the "t" in Helvetica has two protusions on the right, the "t" in Futura just one. how does this affect the size of the following word space? The experiment would involve designing a typeface with alternate versions of the t, and seeing how that effected word space. Another thing to vary is the amount of protrusion of the crossbar in f, and t, and the arm in "r".

Thanks Nick,
you read my mind. I'm in the process of designing
some sort of visual demonstration of typographic
variables, comparing one against the other. I was
sticking to line length, wordspace, global letter-
spacing (tracking) and type size, although now
you mention ink colour and paper stock, I wonder
if I can integrate that. Hum.

You mention word length, which is interesting.
I was assuming, because a word is usually
bigger than a wordspace, that as the length of
a line increases, the amount of printed area
in the line increases more quickly than the
unprinted area, and this is why the wordspace
should be increased. But this is a very
tentative supposition, and I've been trying
fruitlessly to prove it to myself mathematically.
I think I might be on a fool's errand ...

The wordspace in relation to specific glyphs
is also something I've been thinking about in
a very fuzzy way. I tend to see it in relation
to the end of one sentence and a following
sentence starting with "T", which is pretty
common. Sometimes it can get quite gappy,
depending on the font.

Thanks again for your help, you've made me
think about the problem differently. I was
getting pretty blinkered.

>the amount of printed area
in the line increases more quickly than the
unprinted area, and this is why the wordspace
should be increased.

I don't know the answers here, but it seems to me more than one goal and standard is involved. For example, increasing word space in longer lines may help evenness of color, an aesthetic consideration, but may harm readability, which is influenced by consistency of rhythm of the letters and spaces.

In 'The Finer Points in Spacing' Dowding gives convincing examples of where excessive word spacing hurts readability, because the eye doesn't hold to the line of text as easily. On the other hand, his own word spaces to my eye are so pinched that they make reading more difficult.

Good luck on your research & do post your results!

> On the other hand, his own word spaces to my
eye are so pinched that they make reading more
difficult.

Amen to that. I think his love of tight spacing
turned into a bit of a self-defeating fetish.

When I have something worth looking at I'll
post them. This will probably take me a few
weeks though.

Your topic is very interesting. I'd love to read the final piece. My first attempt at an essay when working on my MA was about how space is used in typographic design. While the essay itself is absolutely full of flaws, perhaps my bibliography might lead in some new directions.

---

Bibliography

Bringhurst, Robert The Elements of Typographic Style. 2nd edn.
Canada: Hartley & Marks, 1999.

Dowding, Geoffrey Finer Points in the Spacing and Arranging of Type. 2nd edn.

Gerstner, Karl Compendium for Literates.
Cambridge, Massachusetts: MIT Press, 1974.

Gerstner, Karl Designing Programmes.
London: Alec Tiranti Ltd., 1964.

Hartley, James i{Designing Instructional Text.} 3rd edn.
London: Kogan Page, 1994.

Kunz, Willi Typography: Macro+ and Micro-aesthetics.
Switzerland: Verlag Niggli AG, 1998.

Morison, Stanley First Principles of Typography. 2nd edn.
Leiden: Academic Press Leiden, 1996.

Schriver, Karen A. Dynamics in Document Design.
U.S.A.: Katherine Schowalter, 1997.

Spencer, Herbert i{The Visable Word.} 2nd edn.
London: Lund Humphries, 1969.

Spiekermann, Eric and Ginger, E.M. Stop Stealing Sheep & find out how type works.
California: Adobe Press, 1993.

Sykes, J.B. ed. The Concise Oxford Dictionary (of Current English) 7th edn.
Oxford: Oxford University Press, 1982.

Tschichold, Jan Asymmetric Typography. Trans. Ruari McLean.
London: Faber & Faber Limited, 1967.

Twyman, Michael

Thank you Tiffany. I love that Karl Gerstner.
Am I correct in thinking you went to Reading?
Cos that's where I am.

I realised over the weekend that my project
has two threads, a theoretical one and a
more practical one. I'm trying to design a
way of showing three type space variables
against one another. Wish me luck!

You're all invited to the graduation show,
by the way. It's in the middle of June.

Alex, this is indeed a great topic.

Patterson and Tinker did a lot of solid empirical research in the 50s and 60s, but I don't remember anything specific about word spaces (you might check). As Nick says, the massive inter-relation of variables is very daunting. And I agree with William too: I have my doubts about making word spaces longer in proportion to line length. It might actually do more harm, not just in the way William notes, but also because the longer the line the more you depend on the leading to find your way back to the beginning of the next line, and bigger word spaces make that harder. Lastly, the more horizontal room a line takes the slower the reading.

If Spiekermann would give a reason, we could analyze it - maybe there's something I'm missing?

hhp

Hello Hrant,

thanks for the Patterson Tinker information, I'll
go and have a look at that. As for the wordspace/
line length conundrum, maybe I'm looking at it in
the wrong way. Perhaps it's more helpful to see it
as shorter lines having smaller wordspaces. I can
see how this might be the case on a newspaper,
for example, where short lines with minimal
leading necessitate tighter horizontal spacing to
maintain a more coherent horizontality. Whether
there's more to it than economic necessity though,
I don't know.

I can't help wondering if Erik Spiekermann's
position is one of principle rather than
practicality. Having said that, when I look at
the end of my previous post and compare it to
the longer lines of your post, my post seems more
gappy. I don't know anymore whether or not I'm
seeing things because I want them to be true.
I've spent too long thinking about this problem
: )

Alex, as you are doing research on this I think Erik Spiekermann would explain to you at greater length the reasons behind his views. This could well give you information that will help you in testing his idea. It can't hurt to e-mail him a request through his web site.

It recently occurred to me that there is another factor involved in word spacing: kerning.

When you have a word starting with a "needs kerning" combination of characters, it requires a certain amount of word-space preceding it, in order that the first letter doesn't look like a single word. This is particularly apparent for the nasties: T,V,W and Y followed by a round lowercase letter.

Hence the huge spacing in 19thC typography.

In a tightly fitted and kerned face, you can even put kerning between the space character and letters such as A,T,W,W,Y, and period and comma.

"Hence the huge spacing in 19thC typography."

Nick,

I just saw notice of your Punch article on the SHARP-L list. Think it was SHARP. Skimmed it real quick and bookmarked it. But...

My understanding of the reason for hugh spacing in that very maligned and misunderstood century was that at some point comps were getting paid by the nut quad, therefore, more incentive to nut it out. Not really an aesthetic or technical choice. True or false? Or, you know, somewhere in between.

Gerald

I think if your line gets to the point that it needs additional word spacing to be legible your line is too long and you should redesign. Long lines, big word spacing = swiss cheese.