User talk:Martin Kealey

Hi. Redirecting revocableness was highly inappropriate. We don't omit words just because they sound a bit clumsy! Equinox ◑ 13:19, 16 March 2012 (UTC)

I didn't omit it, I recommended a more common form.

Human languages have generative grammars, which means they have an infinite number of possible words. Since we can't include them all, when do we omit them? When they have no citations? When they are commonly regarded as "typos"? When they are forms only used by non-native speakers? When they've been obsolete for a century?

Would you accept disrevocable? revocabling? revocationable? revokable?

I note your considerable efforts to insert the various inflected forms of English words, particularly adding "-ness" inflections, but it strikes me as a boiler-plate exercise that would be far better explained by a single entry "Meaning of the -ness suffix", aided by improvements to the search functionality that would include base forms when searching for inflected forms.

Martin Kealey (talk) 21:23, 16 March 2012 (UTC)

We include words that are attestable, i.e. ones that have been used, not ones that merely theoretically could be used. Look at WT:CFI. If we redirect words, the software will not "see" them properly for categorisation etc. Also an attempt at "blanket" coverage could be misleading in a few cases, since a word might end with -ness and yet not be using the -ness suffix (compare how silly doesn't mean "in a sil manner"). Equinox ◑ 21:40, 16 March 2012 (UTC)

Thank-you for the reference to WT:CFI; indeed I should have started with that, and I apologise for not having done so.

Attestability (attestableness?) is quite appropriate, however surely that would suggest that citations are in order? (Or are Wiktionary & Wikipedia held to different standards?)

Having said that, it still strikes me that the one-page-per-inflection is suboptimal for affixes that generally have only one meaning, such as "-ness". Certainly have separate pages if they have even shades of different meaning, but if the only possible meaning is readily inferred from the standard meaning of the affix, a simple redirection to the base form should suffice.

But that's a policy WT:REDIR change to be sorted out before starting to make content changes, so I won't be making any more redirections any time soon. Martin Kealey (talk) 22:00, 16 March 2012 (UTC)

We would ideally have citations for all words, but this is a tiny project compared with Wikipedia, so a lot of it hasn't been done. There is WT:RFV for challenging suspect words. Yes, most dictionaries seem to include the -ly and -ness forms under the headword without any actual definition part, but since we're not paper and have unlimited space, writing "state or quality of being __" isn't that big a deal. I make a lot of use of copy-paste and programmable scripts etc. or my fingers would long since have fallen off. Equinox ◑ 22:05, 16 March 2012 (UTC)

Doing it with templates

Have a look at these: WT:Sandbox/attestableness WT:Sandbox/likeableness WT:Sandbox/solubleness and in particular, check out the "source" -- a single template invocation for each page. Martin Kealey (talk) 03:33, 18 March 2012 (UTC)