Talk:Table of loglangs and quasi-loglangs
The table is excellent. And (talk)
It would be good if some of the cell items (e.g. Families) were links to pages on the family and some headings (e.g. Family and SSM strategy) were links to pages listing and explicating these values. And (talk)
- I agree. Things like "A+B+ chunk" are not self-explanatory. At the very least, the cell items should be explained on the page itself, in a supporting section. That would be a first step toward separate pages. Maiku (talk)
I love the adjectival Family names!
The term "speakable" is potentially ambiguous. In the light of the Saussurean distinction between Tongue (Langue) and Speech (parole), a speakable lg is one capable of producing parole -- at first blush the notion of an unspeakable language might sound nonsensical, but in fact there's many an incomplete conlang that has grammar without lexis and so is incapable of producing parole. This is probably not the sense intended here; my initial thought is that "speakable" means "oral", as opposed to "only written" (like UNLWS). But would a signed, chiretic, language be speakable? I think we'd sort of want the answer to be yes. So rather than speakability (apart from in the Saussurean sense), I think the better parameter is "primary realizational medium for phonology", which may be phonetic (as with all the languages listed), chiretic (no loglang that I know of, but Rikchik could count as a conlang example, if chireticality includes tentacularity), or graphical (UNLWS). And (talk)
- "Speakable" here was simply meant to distinguish languages envisioned to be human-speakable (having a phonological (utterance) design component), in contrast to languages such as UNLWS, or ordinary logical notations. I wasn't really thinking about langue/parole or Rikchik. We can take our time and think of ways to handle these wider distinctions. Maiku (talk)
"Year" should be "Year of Publication", possibly with a separate column for "Years of Active Development"? Or even if not these exact two columns, it's still potentially important to distinguish dates of publication from dates of development.
And (talk) 01:40, 9 April 2020 (UTC)
- This is another area that can be hashed out. There are two ways to look at Toaq (Dzu); under one view, there was one long period of development (2013-2017) culminating in modern Toaq, and under another, there were two discrete languages (2013, 2017) of which one is the obsolete predecessor language of the other. Neither view is invalid, but I lean toward the latter view, as I feel it will work better on the wiki (basically I would rather see a short article for Toaq Dzu rather than squeeze Toaq Dzu into a section on the Toaq page). Likewise in the case of the Morneauvian languages, there are several discrete stages, which are largely separate "skins" of the exact same language, but I feel there are enough interesting differences among these "skins" to merit short articles for each one. Maiku (talk) 13:13, 9 April 2020 (UTC)
What makes a constructed language a member of a family? It would seem obvious that Lojban is a Loglanic language; however, Lojban does not have very many cognates with its predecessor, Loglan (being a relexification of the latter). So, then, it would seem that conlang genealogy has to do with inherited grammar as well as lexicon. This makes decisions about family membership more difficult. If Toaq happened to be built on the foundation of Lojban's grammar, would it too be Loglanic? Selguha (talk)
- On further thought, let me sketch an answer to the first question: a family is a group of languages in which a majority of words are (A) shared with an ancestor language; (B) derived from words in an ancestor language through phonological processes; and/or (C) derived from words in an ancestor language through one-to-one relexification, where word definitions in the language of description are kept unaltered. Selguha (talk)