You don’t have to go that far down the list of common words to start encountering content words. I don’t think it’s that implausible that there are simple answers we’re overlooking/overthinking in some places.
EDIT: For example, I don’t think we have “have” yet, which seems a bit odd to me? Surely something in the text should have something at some point?
I’m aware of this!
I’ve been mentally assuming green, because some sort of paired red & green controls on a door would be conventional.
I am not sure. We know from the game completion library message that the translation is not literal; phrases are similar but worded differently.
The main reason it matters is:
CALMAO nearly KINJ enough to SHALOK SESL BISYLEN.
Can we make more sense of this phrase if BISYLEN is “has moved” or “moved” rather than “gone”? What in the world is the mysterious SESL? (A possessive form of “that”??)
Lionese doesn’t seem to use “have” as an auxiliary verb. But I agree it’s surprising that “to have” hasn’t appeared in its ordinary sense anywhere in the text.
I was envisioning that for words like these where probably nothing else depends on it, if we ever finish the rest of the translation, we could have a sort of “cleanup” phase where we put all of the inconsequential but flavourful decisions to a poll. But it’s not inconceivable that we might figure out some other reference in the meantime which could point towards one colour or the other.
That’s fair, but I think there’s also value in assigning tentative translations now, so that decipherers can see the remaining important words in a more complete context.
It would be nice to think so, but in fact I only remember one word of this stuff and that’s a word I lifted from someone else’s story.
(A very obscure SF story from the 1980s. You will not be able to find it.)
Not even BLONG?
Nope, I forgot that I used that one. Until it was mentioned here, I mean.
What if SHI is “low”?
A SHI passage leads northeast, and the cavern FANIT away to the west.
A low passage leads northeast, and the cavern curves away to the west.
The dwarf doesn’t seem to understand you. “I don’t understand,” it SHIFAN.
The dwarf doesn’t seem to understand you. “I don’t understand,” it mumbles.
The DORBLEN is SHIAR here, covered with ZATHAL HEMBO.
The ceiling is lowest here, covered with limestone falling.
SHIP PACKAGE
lower PACKAGE
Less clear are:
The dwarf throws a nasty knife. You SHIAN the knife with your sword.
You down(?) the knife with your sword.
The meaning fits (swatting the knife down to the ground), even if I’m not sure of the perfect word.
It SHISYLE from the cavern, looking MRONSOR NO thankful.
Could it be:
It (shuffles? sidles? retreats?) from the cavern, looking wary yet thankful.
While we’re thinking about corpora, here are all the hits for “enough to VERB DET NOUN on” in the Corpus of Contemporary American English (~1B words):
ENOUGH TO UTILIZE THIS TECHINIQUE ON
ENOUGH TO TRUST THAT LIGHT ON
ENOUGH TO TAKE THIS THING ON
ENOUGH TO TAKE THESE PEOPLE ON
ENOUGH TO TAKE THAT CALL ON
ENOUGH TO STOP ANOTHER SELL-OFF ON
ENOUGH TO SHOW THIS TORTURE ON
ENOUGH TO SHED SOME LIGHT ON
ENOUGH TO SEE ANY MARKINGS ON
ENOUGH TO RUN SOME TESTS ON
ENOUGH TO RIDE THIS PLACE ON
ENOUGH TO REMIND SOME PEOPLE ON
ENOUGH TO RAISE SOME EYEBROWS ON
ENOUGH TO PUT THIS SHIP ON
ENOUGH TO PUT THAT LATCH ON
ENOUGH TO PROVIDE SOME BACKGROUND ON
ENOUGH TO PREVENT ANOTHER ATTACK ON
ENOUGH TO PLOT THAT POINT ON
ENOUGH TO PLAY SUCH TRICKS ON
ENOUGH TO PLAN THESE ROBBERIES ON
ENOUGH TO PLAN ANOTHER TRIP ON
ENOUGH TO PIN THIS BUG ON
ENOUGH TO MAKE THIS POINT ON
ENOUGH TO MAKE THIS DETERMINATION ON
ENOUGH TO MAKE THIS DECISION ON
ENOUGH TO MAKE MUCH DIFFERENCE ON
ENOUGH TO LOG MANY HOURS ON
ENOUGH TO KILL MOST LIFE ON
ENOUGH TO KEEP THIS FILM ON
ENOUGH TO IMPOSE THIS PENALTY ON
ENOUGH TO HOLD BOTH PANS ON
ENOUGH TO HAVE SOME CONSENSUS ON
ENOUGH TO HAVE ANY OPINION ON
ENOUGH TO HAVE ANY EFFECT ON
ENOUGH TO HANDLE THAT TASK ON
ENOUGH TO GUARD MOST PLAYERS ON
ENOUGH TO GRAB THIS ALBUM ON
ENOUGH TO GIVE SOME THOUGHTS ON
ENOUGH TO GET THIS REPLICATOR ON
ENOUGH TO GET SOME MEAT ON
ENOUGH TO GET MUCH FORCE ON
ENOUGH TO FIND SEVERAL CLIENTS ON
ENOUGH TO FACE ANY ENEMY ON
ENOUGH TO DO SOME RESEARCH ON
ENOUGH TO DO SOME REPAIRS ON
ENOUGH TO DO SOME EXPLORING ON
ENOUGH TO DECIDE SUCH THINGS ON
ENOUGH TO COVER THOSE DISTANCES ON
ENOUGH TO COVER THESE EXPENDITURES ON
ENOUGH TO CLOSE THAT GAP ON
ENOUGH TO CAST SOME SUSPICION ON
ENOUGH TO CAPTURE THESE INSTANCES ON
And the Corpus of Historical American English (~475M words):
ENOUGH TO UNDERSTAND WHAT INDIANS ON
ENOUGH TO TRY ANY TRICKS ON
ENOUGH TO THROW SOME LIGHT ON
ENOUGH TO SEW THAT BUTTON ON
ENOUGH TO RAISE SOME EYEBROWS ON
ENOUGH TO PUT ANY SALT ON
ENOUGH TO PLANT BOTH FEET ON
ENOUGH TO MAKE THIS DECISION ON
ENOUGH TO MAKE MUCH DIFFERENCE ON
ENOUGH TO KEEP THESE FELLOWS ON
ENOUGH TO KEEP MOST PEOPLE ON
ENOUGH TO JUDGE EACH CASE ON
ENOUGH TO IMPOSE THOSE VIEWS ON
ENOUGH TO HAVE THAT WOMAN ON
ENOUGH TO HAVE THAT APPARATUS ON
ENOUGH TO GET MUCH FLESH ON
ENOUGH TO FORCE THAT RECOGNITION ON
ENOUGH TO FINISH THAT LILY ON
ENOUGH TO FIND MORE TRACES ON
ENOUGH TO FIND ANY INHABITANTS ON
ENOUGH TO DO SOME REPAIRS ON
ENOUGH TO DO SOME EXPLORING ON
ENOUGH TO DEMAND ANY ATTENTION ON
ENOUGH TO CHOOSE SEVERAL REPRESENTATIVES ON
ENOUGH TO CAUSE SOME UNEASINESS ON
ENOUGH TO BLAME SUCH SITUATIONS ON
None of these is attested more than once, suggesting that there really isn’t a common or standard idiom fitting that pattern.
Can you run the query for “ENOUGH TO x YOUR y ON”?
Good catch! I hadn’t realized it wouldn’t count “your” as a DET.
Contemporary:
ENOUGH TO GET YOUR HANDS ON ×8
ENOUGH TO HANG YOUR HAT ON ×2
ENOUGH TO GET YOUR FOOTING ON
ENOUGH TO COMMUNICATE YOUR CONCEPT ON
ENOUGH TO BET YOUR LIFE ON
ENOUGH TO TIP YOUR CHAIR ON
ENOUGH TO SET YOUR HAIR ON
ENOUGH TO SEND YOUR BALL ON
ENOUGH TO REST YOUR WRISTS ON
ENOUGH TO PUT YOUR BLOG ON
ENOUGH TO PLANT YOUR LIPS ON
ENOUGH TO PLANT YOUR FEET ON
ENOUGH TO PAY YOUR BILL ON
ENOUGH TO MAKE YOUR OPINION ON
ENOUGH TO MAINTAIN YOUR BALANCE ON
ENOUGH TO KICK YOUR GLUTES ON
ENOUGH TO KEEP YOUR SITE ON
Historical:
ENOUGH TO SUPPORT YOUR MOTHER ON
ENOUGH TO PUT YOUR MIND ON
ENOUGH TO PASS YOUR GENES ON
ENOUGH TO LAY YOUR TROUBLE ON
ENOUGH TO KICK YOUR GLUTES ON
ENOUGH TO KEEP YOUR FAMILY ON
ENOUGH TO IMPRESS YOUR PERSONALITY ON
ENOUGH TO EXPRESS YOUR COMMENTARY ON
ENOUGH TO DRY YOUR FACE ON
Which shows one significant outlier. Unfortunately, we already have a word for “hat”, and “hands” would be plural.
When I started this, I didn’t want to add any potentially-wrong translations to the transcript and give the impression that they were confirmed, and I didn’t think I would have the time to maintain a separate dictionary or annotate the transcript with confidence levels for each word. But since @EJoyce has now compiled all the vocabulary into a handy spreadsheet, I’ve gone ahead and added a lot of the lower-confidence translations to my translated transcript.
Here is the latest version of the transcript, and here is a link to the diff for anyone who’s trying to keep anything else in sync with it. Most of the added translations are ones that have already been suggested in this thread; I’ve gone with SHI=“low” following @evouga’s suggestion, and with DOBLEN=“floor”, DORBLEN=“ceiling” (I think there’s something with the ZATHAL being stalactites and stalagmites, but I’m not certain enough to put it in yet.)
I’ve left CORSET as “dragon” for now because I’m not certain what we might want to replace it with. I’ve also left the scoring message as it appears in the Inform library because I’m not sure what we’re proposing to replace it with, but I’ve reverted one translation of BISYLEN in case the translation we had was incorrect.
If we’re getting into corpus statistics now, at what point does bringing out fancy NLP stuff become fair game? I know there are plenty of reasons to be sensitive about language modeling as a concept lately, but we also seem to be in a situation where “statistically modeling the probability of various tokens appearing at certain points in a text” is actually directly relevant.
Personally, I have no issue using language models for things like that. Others might feel differently, but my take is that language models currently get a bad rap mostly because they’re being applied to entirely the wrong problems. Predicting which word best fills in a blank is the right purpose for them, and it’s something they were doing a very good job of before OpenAI found the cheat code for easy marketing.
It should really be “deflect” (if it isn’t “parry”)
Some loose ends:
“The bearer of this document KENOL 6000 shares of Red Hat stock. KOUNAM.”
I don’t see any likely choices for KENOL other than “owns.”
KOUNAM is some kind of politeness construction; we know that
Zao gu-kounam ve
means “would you like to.” Looking online at pictures of stock certificates, I don’t really expect any one-word trailing sentences on one (except “signed”), but from context I think the best we can do here is to translate it as “please.”
The smoke twists and MOZEM, flooding the pit with PLORNY DVASN.
VAS is “bad” and VASI is “nasty,” so we’re looking for another word along those lines for DVASN. “Toxic”?
As for the PLORNY, we also have:
The smoke PLOR around the pit, but it doesn’t seem to NEIDO any.
The meaning here is clear, but it’s hard to think of a noun and verb that have the same root and that both fit. My best guess for both PLOR and PLORNY is “swirls.”
Finally for NEIDO: we need a word that means “dissipate” but that can also serve as the adverb NEIDII. There’s also a very good chance that NEID and LEID form an antonym pair. I think I still stick to my previous guess of NEID = “thin.”
EDIT: I should also mention that XOLNE takes a participle in both of its appearances in the text. Almost all Lionese words have a single English counterpart; and yet it seems almost certain that XOLNE means “doesn’t look to be.” Have we missed a one-word solution?
I had “plume” for PLOR(N), but “swirl” is more common as a verb so that’s probably better.