On Wed, 17 Jan 2007 19:04:07 -0500, Jim Henry <[log in to unmask]> wrote:

>Some years ago my brother and I, and some friends,
>played a conlang party game that went like this.  Each
>player, on his turn, would coin a word in the game-language
>and demonstrate its meaning to the other players with
>charades, drawings, pointing out examples in the
>environment, etc. -- no use of English allowed, or
>English allowed as a last resort (the rules varied).
>There was a short translation challenge sentence
>and the first player to express it in the game-language

Neat!  I once started a conlang with a friend using the same kind of
process, but without any kind of underlying game.  The phonology turned out
different enough from English, although at this point it wasn't formalized:
all words monosyllabic, C*(G)V(G), /ptkbdgfsSxmnNrljwiIuEoa/, interesting
clusters like /Sfmu/ 'spades (the suit)'.  But we soon found ourselves not
up to discussing fine points of syntax or keeping records without English,
so moved to more traditional methods of development (and later revised away
most of the words from that first phase). 

I take it your games were fairly social, aimed more at developing the
language than blinkeredly racing to a translation of the target sentence?  

>The main problem with it was that linguistically
>unsophisticated players (all of us, in some sense,
>or all of us except me, perhaps) tended to
>follow English too closely in the phonology
>of the invented words, the syntax, and perhaps
>the semantics.  I've recently thought of a way to
>recreate it as a card game, with more structure
>to guide players in creating a language that
>hopefully doesn't relex English, and perhaps
>demonstrates some mechanisms of language
>change, as well -- at least as far as phonology
>and grammar are concerned; I haven't got
>so good a handle on the semantic relex

Also a neat idea.  I've had thoughts of a card game that generates a conlang
as well, but nothing this fleshed out.  

OT (in two senses!): here's another nebulous and undeveloped idea I once had
for a linguistic-flavored card game, using normal cards.  In each round of
play, an input card or structure of cards would somehow be generated.  Each
of the players would then play a card or structure of cards from eir hands,
and these would be ranked using a hierarchy of OT-style contraints, the
player who produced the best candidate winning the round.  There would be
some way to adjust the constraint ordering in your favor, maybe by bidding
in an auction before the hand, or maybe dynamically during the play.  

>My basic notion for Glossotechnia (my tentative new name for it)
>as a card game is this: there would be two decks.  The main deck
>includes cards like Phonemes (k, t, p, a, i, u, etc.), Syllables (CV,
>VC, CVC, etc),
>and Syntax cards (Subject-Verb-Object, Verb-Subject-Object, etc.).
>When you play a Phoneme or Syllable card it
>goes face-up in the middle, in a kind of phoneme table
>arrangement, and you're helping define the phoneme
>inventory and phonotactics of the game-language.
>When you play a syntax card (like "VSO" or "head-modifier"
>or "prepositional") you help define the syntax of
>the language.  You can play a new syntax card that
>replaces a syntax card of the same kind already
>in play (e.g. "SVO" replacing "VSO").
>The Phoneme cards consist of most of the phonemes
>of English, plus a few like /y/, /2/, /e/, etc., which many
>Americans will have a nodding acquaintance with
>from high school German, French or Spanish classes.
>If I were making a deck to take to the Conlangs Conference,
>of course, I would include more exotic stuff. :)

How about phonology cards that introduce whole contrasts at once? 
Especially for exotic stuff -- having to put in palatalized and labialized
and pharyngealized and devoiced and creaky and breathy and ... variants of
everything would quickly swamp the deck; instead, while (say) a
"pharyngealisation" card is on the table, pharyngealized variants of
everything are in play.  But you could use tamer feature cards as well;
they'd make the inventory more likely to be symmetric (if anyone cares), and
they'd provide a mechanism to do sound-changes on whole series at once
(replace the "mid-high vowels" card by the "high vowels" card).  

>Some of the most common phonemes in the world's
>languages occur more than once in the deck,
>as do the three most common word orders and the
>CV and CVN syllable shapes.
>Then there are Sound Change, Grammar Change
>and Meaning Change cards (also part of the main deck,
>shuffled in with the Phonemes, Syllables, and Syntax cards).
>They let you do things like split a phoneme by context,
>drop a phoneme (merging it with another already in
>play), replace one phoneme with another; add
>or drop inflections; extend or restrict meanings of words.

"add or drop inflections" meaning "fuse independent words together" or
"grammaticalize",  respectively "replace with a periphrastic expression"?

>Then there is be the translation challenge deck -- a
>collection of translation challenge sentences; everyone
>would draw one at the start of play and translating that
>sentence into the game-language would be their goal
>to win the game.  There is another challenge
>card set face-up in the middle and no one can win
>with their private challenge until the group challenge
>has been translated.

I suppose this deck would have to be large, so that after repeated play
players don't come to know all the sentences in it and therefore know what
new words people will be trying to create.  I was thinking it might be more
fun to let the players generate a pool of goal sentences themselves at the
start of the game, but that's subject to the same problem, only worse.  

The opposite style of goal would also make an interesting game: instead of
an English sentence to translate into the language, give some random
phonological string (probably without word divisions marked, like
[lo?otaixetandZiSoendala]) and let the objective be to create a language in
which that string is a grammatical sentence.  This completely reverses the
focus of the game, though.  Being good at charades or drawing so that you
can convey the meaning you want becomes unimportant, but the choice of
phonemes and the like becomes strategically significant.  

>If there are at least two phoneme cards and at least
>one syllable card in play, a player can also coin one
>or more words on their turn, using the old rules
>(demonstrate the meaning with charades, pointing out examples, drawing
>pictures, using previously coined
>words of the game-language etc. but no use of English).

But I presume the other players are guessing using English?  That right
there might be the greatest force imposing English-like semantic divisions:
it's more natural, especially for the "linguistically unsophisticated", to
guess an English word than an arbitrary region of semantic space.  

Suppose I'm trying to coin a word with a particular semantic range in mind
but my opponents will only guess nearby English words.  What then -- do I
relent and just accept a close English word for my semantics?  do I consider
it close enough and end the guessing round, but later indicate the semantic
shading I had in mind?  do I somehow try to indicate 'broader', 'narrower',
whatever, until someone has it exactly?  Especially if there's some benefit
to getting your words guessed as quickly as possible (e.g. getting more
definitions per turn), players might tend towards semantics equivalent to
English out of efficiency, or laziness.  

>But you can only use the phonemes  in
>play and the syllable shapes in play.  So if e.g. it's
>the second round and people have played the k, t, i and o
>cards and the CV and CVC cards, you could coin words
>like "kito", "tok", "tiki", etc.
>If later on the "e" card is played and then someone plays a "phoneme merge"
>card to discard the "i" and say that /i/ merges into /e/, then words already
>coined with "i" in them change it to "e".  "keto", "teke", etc.
>Maybe there is a time limit on coining your word(s) -- if you
>can't make the other players understand you within, say,
>2 minutes, the next player gets to start their turn.  But if you
>can demonstrate the meaning of two or more words in the
>time limit, fine.  (Maybe it's easier to coin words like "this"
>and "that", or pronouns, in groups than one at a time.)
>Maybe if you manage to define your word just using
>the game-language, with no charades or drawings
>etc., you get to draw another card at the end of your
>turn (increasing the size of your hand); but when
>your time is running out and you have to resort to
>English to define your word, you have to discard
>a card at the end of your turn, reducing the size of
>your hand.
>There is a similar bonus for the player who first translates
>the group's challenge sentence -- they get to draw an extra card,
>increasing their hand size.
>There are a couple of "new mission" cards that let
>you discard your translation challenge and draw a
>new one, or swap your challenge card with someone
>else, and some other Action cards -- one lets you
>draw four, play one and discard three,
>another gives you a free pass to define a word in
>English with no penalty, etc.
>I've started making the main deck, but haven't done
>much on the translation challenge deck yet.
>Probably I'll be play-testing it with my brother and
>some other friends in a couple of weeks.

I'd be interested to see logs of game sessions, if they're not too much of a
pain to keep.  

>Any comments or suggestions -- especially on how
>to use card-game mechanics to prevent or
>discourage semantic relexing of English?

I haven't come up with any card-game mechanics better than your "semantic
extension" and "-restriction" cards, which can be used to combat English
relexing after the fact.  

Maybe players could be somehow encouraged to use existing words in extended
senses to render the words in their target sentence -- this would be a
lopsided approach, though, only causing semantic broadening and not narrowing.