Apache OpenOffice (AOO) Bugzilla – Issue 93876
spellcheck phrase when part should fail otherwise
Last modified: 2014-09-22 22:09:51 UTC
For Spellcheck, a word should be any string including one with a hard or soft space in the middle. We might describe these as phrases, but they aren't always that. Linguists count a string as a word whether it's set solid, hyphenated, or with a space, because they behave grammatically like words. Some of them should not be approved by a spell checker unless all parts are present. Common instances include names. More than one space might occur. Examples: von der Lippe (a family name) stare decisis (a U.S. legal principle) I looked in 4 good print dictionaries. None of them has _decisis_. None of them has _der_ as a separate word other than as an abbreviation. Only one has _von_ and 3 don't; and the one that has it puts it into the context of personal names. Writer's Spellcheck doesn't recognize _decisis_. However, I did not look in a legal dictionary per se for _decisis_ by itself. One authority on whether a word can have a space inside it is Webster's Third New International Dictionary of the English Language Unabridged (Merriam-Webster), Explanatory Notes, section 1.1 (at page 16a), and The Writing of Compounds, sections 1.1 and 1.2 (at page 30a). Since Writer's Spellcheck does accept _von_ and _der_, it seems likely they were included because of their use with other strings, and not because of their use in isolation. The better method would be to include them separately only if they would be encountered with different neighbors, otherwise to include them only as parts of words that are spelled with intervening spaces. A word counted for good spelling does not have to be a word as defined by word wrap. The space that might be included in a word should be either hard or soft. As an experiment, I typed "stare decisis" (without quote marks and with a soft space) into an unsaved document and added it as a word to the standard user-defined dictionary. Nonetheless, when I ran the spellcheck (by F7), _decisis_ was displayed in the dialog as Not in dictionary, so that _stare decisis_ was not automatically accepted. I'm using OOo Writer 2.4.0 without Java Runtime Environment on Linux Fedora Core 4 on a Pentium 4 laptop. I didn't see this feature. Sources: von der Lippe's name is at <http://www.valley.net/~townsend/Lippe.html>, as accessed 9-13-08. Dictionaries: W3 (above); Shorter OED ([4th ed.]); American Heritage (3d ed.); Random House Webster's Unabridged (2d ed.). Somewhat related: Issue 12541. Thank you. -- Nick
Reassigned to SBA
Reassigned to requirements.
No complaint from spellchecker about decisis after its addition to standard.dic AOO410m14(Build:9760) - Rev. 1573062 2014-03-01_04:11:01 - Rev. 1573123 Debian
But, in English usage, I don't think "decisis" should be recognized except in "stare decisis", so that "stare decisis" would be correct but "decisis", if alone, would be wrong; and likewise for other strings with spaces inside of them where one of the space-bounded strings does not occur by itself in English usage.
stare decisis -> only decisis is underlined by the spellchecker (stare is already known by the spellchecker). Only decisis is added to standard.dic. And your request is to have decisis alone still underlined when not in conjunction with stare? I fail to understand this and doubt if this issue is meaningful.
This is for writers of advanced texts, such as legal briefs, college papers, and scholarship, in which it's somewhat common to use phrases that are borrowed from foreign languages into English. Many have spaces. Some phrases behave (in the view of linguists) grammatically like words. Besides "stare decisis", examples include "per se" and "inter alia"; probably another hundred or so that are relatively common could be added to a dictionary. If "decisis" is not a word by itself, it should not be in the dictionary, but "stare decisis" should be. Likewise, if "se" is not a word, it should not be added to the dictionary, but "per se" should be, and much the same can be said for "inter alia". The programming may not be very difficult. If the first substring of a spaced string, when it is by itself, fails a spellcheck, the spellchecker can then go on to evaluate the balance of the string the same way. If two or more consecutive substrings include any that failed, the spellchecker then checks the pair of substrings as one string. If it passes the last test, it would then edit the underlinings. It might search for specific spaced strings throughout a document or selection (approving all) before returning to the beginning to check all other strings for errors, which solves a problem of identifying spaced strings against a dictionary. A space may be represented by a line break.
(In reply to Edwin Sharp from comment #5) > I fail to understand this and doubt if this issue is meaningful. "Stare decisis" is a word that is part is a specific lexicon, so it is difficult to see why the error is meaningful as you are unlikely to use "decisis" without "stare" preceding it. But with a more common word, take "per se," this is a little more problematic. "Se" is not a word, but "per se" is. However, it isn't hard to imagine someone making the typo "Se went to the store," when you really mean "She went to the store." Writer does not count this as a spelling error if you were to add "per se" to the dictionary because it seems to take "per" and "se" as separate additions to the dictionary. I tried this out in Microsoft Word to compare, and while adding "per se" to the dictionary resulted in "per" and "se" being separate entries--and therefore "se" on its own not being counted as a spelling error--Word was able to pick out in it's grammar check that I probably meant "She went to the store." Writer, on the other hand does not have an internal grammar checker, and so that error would go unnoticed. Some other words compounded with a space that may pose a similar problem include: "terra firma," "ad hoc," and "per capita."