AaronOfBarbaria
Adventurer
It is true that intent can only be read at the table, but changing whether an action is allowed or not based on how the player describes their intent seems incredibly inconsistent and unfair to me - especially when the end result is saying "that action would be fine for your character to do, if you weren't so daft as to actually tell me the reason you picked it was because you are trying to get your character to pass this challenge successfully."I think in both these points it comes down to intent that can only be read - but is sometimes pretty obvious - at the individual table at the time: whether an experienced player is doing action X (e.g. avoiding the gaze of a Basilisk on meeting one for the very first time) only because she-as-player already knows it's the right thing to do when the character has no in-game reason to know to do this?
Lan-"sometimes the best monsters to meet are those the DM has invented herself"-efan
As for your basilisk example... your assumption that the character doesn't have an in-game reason isn't a fair assumption. The character might have no idea that lizard is or isn't actually a basilisk, but has simply heard tell of cave-dwelling reptiles that turn people that look at them to stone and is playing it safe. The character might also not be adventuring in Gotcha Land, so there has likely been some evidence suggesting petrification of people leading up to the actual encounter... or it could just be that the character sees a very large lizard with too many legs to be natural covered in wicked looking spines and with it's eyes literally glowing (seriously, look at the art for a basilisk) and has decided that if anything in the whole world could turn him to stone if he looks at it, this horrific thing can.
Or to say that much more briefly: a basilisk is a very bad example to use for a variety of reasons.