Skip to content

Language issues

This is just a few quick notes about some of the issues of language and conventions that come up when dealing with subject wikis.

One of the most basic questions is what language to use. English is the only language I know where all the requisite terminology and ideas have been developed sufficiently, so that’s what I’m using, but there is still the tricky matter of what language variant to use. The current choice is American English as the default. This is partly opportunism: the bulk of traffic comes from the United States, so catering to American English is probably better for the bulk of traffic.

I’m not aware of any software features in the English MediaWiki that can automatically handle the minor spelling changes between different English language variants. I did recently discover that some pretty clever stuff has been done to handle issues of traditional versus simplified Chinese on the Chinese Wikipedia. Probably, the differences between different variants of English (the standard Queen’s, American, Canadian, Australian, and others) is too little.

I have made generous use of redirect features to automatically redirect between different language variants for the spellings of terms, but the current solution is unsatisfactory.

Even fixing a particular language variant doesn’t solve the problem because the need for mathematical terminology tpyically requires one to distort language anyway. For instance, I need to use “closedness” for describing whether something is “closed”, rather than the more pleasing “closure”, because the latter is used for the act of making something closed. Then, there are issues of hyphenation and capitalization. By and large, I have tried to follow rules that are reasonably consistent and unambiguous and also match up with the way mathematicians typically use the words. The butchery of language is likely to displease many. On the other hand, the good thing about a wiki structure is that the tools for navigating and searching for information can be so good that spelling discrepancies and weird-sounding word formations aren’t a hindrance to finding information quickly.

More disturbing are issues of convention within the subject. For instance, many people on the pure side of group theory use the “right exponentation” notation for group actions, while representation theorists and many people in other parts of mathematics use the left action convention. Having a single convention for the wiki might be appealing, but it doesn’t seem quite right because results in a particular sub-discipline should be stated in the language that people of that sub-discipline are used to.

The solution I’m gradually coming to for this is roughly as follows: In most cases, try to have a statement that doesn’t get notational at all, so that people do not have to worry about issues of left and right. In addition to this, have notational versions of the statement, and give them in left-right pairs — for each left action convention version, give the corresponding right action convention version. For instance, consider the definition of pronormal subgroup.