Which alignment on the printed page is better: Ragged-right or justified? It seems that ragged-right is preferable, at least in some circumstances.
Today, I’m re-posting a piece that I first published on April 23, 2009, on the now defunct Content Wrangler site and then moved it to this blog as legacy material that was buried in the dark links of history…
I’m revisiting that post for two reasons: To my surprise, this has been one of my most popular posts in terms of search queries, so apparently this is an interesting topic. And I’ve discovered an additional argument with a twist that was new to me…
But first, let’s rewind…
Last year, I wrote:
How do you argue for the preference of ragged-right over justified alignment in print? Searching the web, I soon came across pages which mentioned research, but it was harder to actually find it.
- The National Center on Educational Outcomes put out the NCEO Technical Report 37 which summarizes several arguments and references on the topic in “Table 3. Characteristics of Legible Type”, see the entry on “Justification”. Among them are:
- Margaret Gregory’s and E. C. Poulton’s article “Even versus Uneven Right-hand Margins and the Rate of Comprehension in Reading”, Ergonomics, Volume 13, Issue 4, July 1970, pages 427-434. From the abstract: “… made no difference for good readers, but for the poorer readers the justified style resulted in a significantly worse performance.”
- Steven Muncer, et al’s article “Right is Wrong: an examination of the effect of right justification on reading”, British Journal of Educational Technology, Volume 17, Issue 1, January 1986, pages 5-10. From the abstract: “… with reading material presented in right-justified format and in ‘ragged’ uneven line format, subjects performed significantly worse on right-justified material.”
- David R. Thompson’s paper “Reading Print Media: The Effects of Justification and Column Rule on Memory”, paper presented at the Southwest Symposium, Southwest Education Council for Journalism and Mass Communication (Corpus Christi, TX, October 6-7, 1991). From the abstract: “… best score for recall was recorded in the flush left/jagged right.”
- The UK government agency RNIB’s “Clear print guidelines” on designing printed information that is accessible to people with sight problems: “… avoid justified text as the uneven word spacing can make reading more difficult.”
- The SEC’s “Plain English Handbook: How to create clear SEC disclosure documents” (PDF), see p. 50: “… spacing between words fluctuates from line to line, causing the eye to stop and constantly readjust.”
- … and a thoughtful blog post by Ken Adams with an argument by Ellen Lupton, curator of contemporary design at Cooper-Hewitt National Design Museum, and author of “Thinking with Type: A Critical Guide for Designers, Writers, Editors, & Students”: “… subtle word-spacing and letter-spacing algorithms are needed to make justified text look ‘good’.”
Now I’ve found more arguments:
Karen Schriver’s book Dynamics in Document Design of 1997 is both comprehensive and excellent in explaining the motivation, the tools and the history of good document design. Be warned, however, that it deals almost exclusively with printed document design. Online design is the afterthought that takes up Appendix C, pages 506-517. (That detail right there tells you pretty much what kind of a book it is… 🙂 )
Schriver essentially argues that ragged-right vs. justified is the wrong question – imposed on us by software options, I want to add. According to Schriver, the real issue is word spacing.
Regular word spacing makes for faster reading and more accurate comprehension, in both ragged-right and justified text. Much of the software we use for writing gets word spacing in ragged-right alignment reasonably right without too much trouble. The problem with justified text is that it requires a sophisticated balance of letterspacing, word spacing and word hyphenation which much software apparently doesn’t get right automatically. Instead, it…
…creates a disturbing visual illusion known as “rivers” – paths running vertically through the text that connect the blank spaces between words on adjacent lines. (p. 270)
Here’s an illustration of rivers, from the Wikipedia article on Sentence spacing:
An example of the "river" effect in justified text
So, the bottom line is: If you have rivers in your text, consider ragged-right alignment to do your readers a favor – or invest extra time in spacing your lines nicely.
– If you know of any other arguments or sources that can help us tech writers with alignment and justification, please leave a comment!
Filed under: style guides, technical writing, writing | Tagged: justified alignment, karen schriver, ragged-right alignment |
Kai's Tech Writing Blog 
Six or seven words per line, and double spaces before sentences? No wonder there are “rivers”.
There’s nothing wrong with justified text when the column is fifteen or so words wide, and has single spaces after each full stop. The argument that it’s harder to read is, I think, based on spurious evidence. Ragged text just looks ugly and frustrates me when I’m trying to read, and I’m an excellent reader.
Just an additional data point.
Thanks, GuanoLad, for your comment. I must admit I chose the image for sheer convenience, because it was readily available. It’s not a perfect example of the river problem in otherwise well set text.
As far as I can tell, the quoted evidence and the tests it is based on are solid, however, and shouldn’t be dismissed lightly. It sounds like Gregory’s and Poulton’s finding applies to you where it “made no difference for good readers”.
Have you read Toni Morrison’s “The Bluest Eye”? If you did, you will have noticed that Morrison employs both, ragged-right and justified, to distinguish different narrative positions. So on top of everything else, it can also be a stylistic medium.
The point is excellent readers have no problem either way — this is a concern in making content accessible to all. Not only is word spacing an issue with justified right, but the uneveness of ragged right makes it easier for the reader to scan to the next line because the ragged right provides visual clues or markers of the line the eye just left. It’s easier to find the next line with ragged right. Ragged right rules, and it’s perfectly justified!
Thanks for your comment, Geoff! I really like your sign-off line about ragged right being justified – it even inspired a tweet! 🙂
Excellent post! Hopefully I remember to come back to this next time I’m trying to explain to someone why I dislike justified text.
I admit that it can be done well – those cases where there actually is a sophisticated balance of letter spacing, word spacing and word hyphenation – but I frequently find it distracting.
As a below-average reader with ADHD, distraction is my enemy and the quote from the SEC is spot on:
“… spacing between words fluctuates from line to line, causing the eye to stop and constantly readjust.”
That time spent readjusting can be enough for my mind to wander or for me to lose my place. Poor hyphenation can also be a problem, where my mind completes the word automatically, then reads the tail of the word as a new word.
Whilst it’s mostly the unven word spacing that causes problems, I also think that ragged-right text can be helpful when I’m finding my place in the paragraph; a pattern is formed on the right-hand side that serves as a visual reference point when scanning or glancing at a line. The line I should read next is delineated mostly the first word of the current line, but also by the pattern of the paragraph.
One readability problem that left-aligned text can have, though, is the awkward appearance of lines that “stick out” due to gaps left where long words didn’t fit. Somehow this is distracting as well. It’s probably also what gives left-aligned text its reputation for being uglier than justified text.
Ideally, I’d like to see “gentle raggedness” on the right hand side, and software algortithms that offer light-handed hyphenation of left-aligned text would be welcome.
Thanks, Nathan, for adding some valuable first-hand evidence to what’s been a mostly theoretical post so far. I’m glad you found it useful!
I hadn’t thought of using the ragged-right paragraph pattern for orientation, but I think it works just as well or better than the less significant identation at the beginning of the paragraph.
Now, if we could only get the SEC to effectively battle distraction in financial matters, too… 🙂
I admit my personal preference is for ragged right, so I was happy with the results of your research findings. I have to add, though, that in addition to the “river” problem, I’ve also encounted the opposite–major problems with the extra tight kerning on some lines of right-justified text that make a line of text read as though it is a single word. This causes the reader to have to re-read or slow down to discern breaks between words. As an editor, I am occasionally required to mark up copy to highlight unacceptable breaks–for example some clients don’t allow their company name or product names to be broken across two lines of text. When copy is ragged right, I don’t usually have to worry that adjustments to copy spacing like that will result in unacceptable word kerning elsewhere on that line or paragraph, except to maybe cause a very short line immediately before the fix. With justified-right copy, I do have to re-review all the copy in that paragraph to make sure that fixing the unacceptable break hasn’t now resulted in kerning problems elsewhere in the paragraph.
Thanks for your comment, Mary Jo, and sorry for the very late reply!
I know the issue with extra-tight kerning, especially when reading a foreign language. I have the hardest time with twitter hashtags that contract several words – and have come up with amusing misreadings when mentally breaking the words in the wrong places!
I hadn’t thought of the perpetuating kerning issue yet, but I can totally see how fixing kerning in one line may create problems further down in the same paragraph. So it seems even “professional” kerning can create problems in justified paragraphs!
I prefer ragged right in all my office memos or other technical documents written by me or colleagues. The reason is that with justified text, it’s possible to lose track of which line you’re on because they all are the same length. With ragged right, the variability in length lets me keep track of lines easier. I didn’t see this reason in the comments. Then there are other word processor-specific reasons, such as it does weird things to last lines sometimes, and that the kerning never seems to be as good as in a textbook, which gives the “rivers of whitespace” problem.
Thanks for your comment, Fred – I used to like well-done justified text as a young and voracious reader. But since I’ve been writing more, I’ve come to appreciate ragged right for similar reasons as you do!
[…] not justify. There is a certain consensus that justification of text on web pages cannot compare to text justification on printed works and […]
[…] and right-aligned text both have a “ragged” left edged which has been shown to impede reading speed and comprehension. The straight (or “hard”) left edge and ragged right edge combination of standard […]
We are have this discussion in the office right now, thats how I came upon your post. I am designing a magazine in which my account exec says it is hard to read when the type is justified. As a designer I believe it to be cleaner looking. Since I am designing it to be read we are designing the type ragged. Thanks for your post.
Glad my post was of help, Brooke. Good luck with your design!
One problem is that word processors like MS Word do whole-space justification rather than micro-justification. My nearly 30-year-old word processor, XyWrite, which fitted on a low density floppy disk, could do micro-justification. It may not have had all the bells and whistles that MS Word has, but it definitely had more pistons and cylinders.
Good point, Steve. Yes, I’ve noticed that justification in MS Word seems to deteriorate the longer the words in a sentence are. Personally, I’m mainly in software documentation, so I haven’t really had my hands on any decent type-setting or DTP software that might be out there.
This study seems to have been overlooked by a lot of folks researching the ragged right vs. justified effect. In summary, no difference in reading speed or comprehension was found for various forms of alignment. This directly contradicts the earlier research by Trollip and Sales in 1986.
Coll, J. H., Fjermestad, J., & Coll, R. (1998). An eight experiment sequence to determine reading equality. Information & Management, 34(4), 231-242.
Thanks, Matt. I wasn’t aware of this particular study, but it seems that findings vary depending on the different variables the studies put in place, such as age, education, situative circumstances. I was aiming for a consensus of most of the studies I could find – which admittedly still allows for the “most correct” study to be the outlier of the bunch…
One interesting thing Coll, Fjermestad & Coll did was to analyze groups for differences in age, sex, education, work experience etc. that could potentially act as mediators for reading ability and/or prior knowledge and thus have an impact on speed and comprehension.
Since they did not find any differences they of course couldn’t do an analysis on which factors might be relevant but the groups themselves did not show any significant differences on these factors.
These earlier studies of course were done with printed pages. More recent research has looked at e-delivered text (i.e. web pages, e-readers etc.) Some interesting conclusions have been drawn regarding 1 column versus 2 column text etc.
Gaps remain in the research however so while we have lots of cool stuff to look at… more research is needed! Hopefully at some point I will be able to add to the conversation myself.
To me ragged text always looked dirty, like the author never made the effort to typeset the text properly. It has the same effect on me in interrupting the flow of reading as orphans and widows in typeset text.
I am therefore glad that there is now research showing that the issue is not with ragged or justified text but with letter spacing. With excellent tools for scientific and technical writing, like LaTeX and proper hyphenation, this is a non-issue.
I have decided that when preparing documents for publication in multiple formats (the web, ebooks, tablets, telephones, etc) the ragged format is better due to the possibility of rivers in alternate formats. Ebooks have a feature that allows one to select the size of the text (an especially useful option for people with limited vision) so this type of publishing will need to use the ragged format. In a printed edition I would prefer right justification but attention to kerning needs to be addressed.
It would appear that with “modern” publication methods right justification will have to be left behind in most instances.
Thank you for the interesting article.
Thanks, Joe. I’m not sure how well aware I was of ebooks, tablets and smartphones as publication devices when I wrote the post almost 6 years ago. But I agree that the more recent and diverse devices potentially aggrevate the rivers issue. In general, it seems that online help layouts get more minimalist and simpler, and right justification doesn’t quite square with this trend.