Linguistic mechanisms for less wrong cognition

I’m work­ing on a con­lang (con­structed lan­guage) and would like some in­put from the Less Wrong com­mu­nity. One of the goals is to in­ves­ti­gate the old Sapir-Whorf hy­poth­e­sis re­gard­ing lan­guage af­fect­ing cog­ni­tion. Does any­one here have any ideas re­gard­ing lin­guis­tic mechanisms that would en­courage more ra­tio­nal think­ing, apart from those that are pre­sent in the oft-dis­cussed con­langs e-prime, loglan, and its offshoot lo­jban? Or per­haps mechanisms that are used in one of those con­langs, but might be buried too deeply for a per­son such as my­self, who only has su­perfi­cial knowl­edge about them, to have rec­og­nized? Any in­put is wel­comed, from other con­langs to crazy ideas.