Slashdot Mirror
Are Two Spaces After a Period Better Than One? (arstechnica.com)
Researchers at Skidmore College conducted an eye-tracking experiment with 60 Skidmore students and found that two spaces at the end of a period slightly improved the processing of text during reading. Ars Technica reports the findings: Previous cognitive science research has been divided on the issue. Some research has suggested closer spacing of the beginning of a new sentence may allow a reader to capture more characters in their parafoveal vision -- the area of the retina just outside the area of focus, or fovea -- and thus start processing the information sooner (though experimental evidence of that was not very strong). Other prior research has inferred that an extra space prevents lateral interference in processing text, making it easier for the reader to identify the word in focus. But no prior research found by [study authors] Johnson, Bui, and Schmitt actually measured reader performance with each typographic scheme.
First, they divided their group of 60 research subjects by way of a keyboard task -- the subjects typed text dictated to them into a computer and were sorted into "one-spacers" (39 regularly put a single space between sentences) and "two-spacers" (21 hit that space bar twice consistently after a period). Every student subject used but a single space after each comma. Having identified subjects' proclivities, the researchers then gave them 21 paragraphs to read (including one practice paragraph) on a computer screen and tracked their eye movement as they read using an Eyelink 1000 video-based eye tracking system. [...] The "one-spacers" were, as a group, slower readers across the board (by about 10 words per minute), and they showed statistically insignificant variation across all four spacing practices. And "two-spacers" saw a three-percent increase in reading speed for paragraphs in their own favored spacing scheme. The controversial part of the study has to do with the 14 point Courier New font that the researchers presented to the students. "Courier New is a fixed-width font that resembles typewritten text -- used by hardly anyone for documents," reports Ars. "Even the APA suggests using 12 point Times Roman, a proportional-width font. Fixed-width fonts make a double-space more pronounced."
First, they divided their group of 60 research subjects by way of a keyboard task -- the subjects typed text dictated to them into a computer and were sorted into "one-spacers" (39 regularly put a single space between sentences) and "two-spacers" (21 hit that space bar twice consistently after a period). Every student subject used but a single space after each comma. Having identified subjects' proclivities, the researchers then gave them 21 paragraphs to read (including one practice paragraph) on a computer screen and tracked their eye movement as they read using an Eyelink 1000 video-based eye tracking system. [...] The "one-spacers" were, as a group, slower readers across the board (by about 10 words per minute), and they showed statistically insignificant variation across all four spacing practices. And "two-spacers" saw a three-percent increase in reading speed for paragraphs in their own favored spacing scheme. The controversial part of the study has to do with the 14 point Courier New font that the researchers presented to the students. "Courier New is a fixed-width font that resembles typewritten text -- used by hardly anyone for documents," reports Ars. "Even the APA suggests using 12 point Times Roman, a proportional-width font. Fixed-width fonts make a double-space more pronounced."
Double-spacing is a hangover from manual typing, and most of us who learned in that era learned by typing with something resembling Courier. Most typewriters couldn't handle proportional fonts or adding extra space after a period, so double-spacing was the way around that. When I see something double-spacing, I recognize that person as someone who is generally old enough to have learned on a typewriter (or the first generations of word processing), and who doesn't engage heavily with IT. Those people are also likely to be less distracted in reading and thus capable of reading faster. The "3 percent increase" for them reading with double spaces is hardly significant.
They're fun to "find/replace" when some old type-writer-using geezer has put them in one of our document templates. Another favourite of mine is the guy who puts spaces after an opening bracket ( like so).
There are different kind of searches. For example, we have got a database with six million product titles. Suddenly we may get two identical to human eye titles, but which differ just by a single, almost unnoticeable, space. Or two spaces, or three.
This should be a precise search, I have to find exactly the title, which I need. No error is allowed.
Certainly, I can solve this programmatically. But why to add more entropy, more disorder, more lack of predictability into the human language, which is a mess even without it.
Because you have humans using the system.
Which is why you "ignore whitespace" and code it as an option wherever it's needed. If anything, it actually makes searching FASTER, because an exact-whitespace match takes longer to find.
When you then put in Unicode, other languages, non-breaking spaces, paragraph marks, and you're working on human-entered data, you're really onto a loser from the start if you have coded anything on a byte-for-byte matching process.
Also, your system works against you in more ways than one. Someone creates an entry with one space. Someone else doesn't see it so they create it with two. Now you have two entirely different entries with different data referring to different database rows, but both "look" identical.
Ignore whitespace, and the problem solves itself.
Oh, look at that.... Slashdot replaces double-space after period with single space. This is definitely two spaces to the left. And now three spaces to the left of that.
When even Slash has it in code they haven't updated properly in decades and can't have UK pound signs (you get this junk:
£), you know you're onto a loser!
Not really, computer cannot really do this automatically except very complicated language-specific programs with lots of exceptions, because the "." is also used within sentences for abbreviations.
Sentence boundary disambiguation has always been a rather annoying problem.
The thing with using paragraphs, sentences, etc, is that it allows one to bite off bits of information. We already use em-dashes to mark off particular clauses we wish to emphersise -- to this point -- but now argue about spaces. White-space helps the reader catch the large-scale of the text. Setting a sentence off in double-space is one devise that does this. Even in the font i type here, it is easier to pick sentences off.
Paragraphs are likewise set off by a blank line, or first-line indent. Where first-line indent is not permitted, then one should use blank line separation. Quotes and examples should be set in an indented paragraph, where the quote is the paragraph.
Bold text is appropriate if one wants the text to stand out on the page at first glance. Highlighted text that is meant to be emphersised in context should be in italic.
OS/2 - because choice is a terrible thing to waste.
2 spaces and the Oxford comma! You one space kids with your missing commas can get off my lawn!
Please call it the Serial Comma. My, Oxford trained, English teacher was quite specific that putting a comma between the penultimate item and the "and" was wrong.
I'll see your Constitution and raise you a Queen.