Let us make an a word “totality” to convey to “everything to which a word refers and that it conveys”. If a person were to create a new totality and assign it to the word “pain” then nothing previously said about pain would necessarily hold relevant to subsequent uses of the word pain.
In fact, to show that any previous relevancies still held true at all to a new totality one would have to show how the two totalities are necessarily indicative of an identical concept such that both before and after the words would be used interchangeably. So one could for example preserve the consistency of reference to a word such as “pi” by assigning the new definition “the value of the arccosine(-1)” instead of “the ratio of a circle’s circumference to its diameter.”
When a human defines any English word, he is changing the totality until proven otherwise. For this reason if one person writes an essay on “meaning” as does another the works may be about something totally different if they don’t use the same totality (which usually necessitates identical or interchangeable definitions) for the word “meaning.”
Even biases in behavior are included in totality. For that reason it would be perfectly feasible to put the totality of the word “toothpick” into the word “meaning.” This is because rational people would then no longer find the meaninglessness (the attribute of not having toothpicks) of an argument a reason to disregard an argument.
Likewise there can be a totality of “meaningful” such that there is not necessarily a practical advantage of listening to meaningful things. And there can be many definitions of “right” such that no logical hedonist has any reason to do what’s right any more than he would want to do what’s “wrong.”
It appears to me that there is presently no set clear definition nor totality associated with many words used in philosophy. If there were then it would be a simple matter to determine what things are "right", "fair", "meaningful", "justified", etc. If we had a definition and totality of "right" then we could perform experiments to measure what actions produce "rightness" by that given definition (not that this would have any application until such an application was illustrated). This would change ethics to either inapplicable data or game theory.
Even so we don’t actually need the word “meaning” to think about the practicality much like a person could avoid using the symbol π (and instead use the definition) while still deriving the relevant math. Regardless of what “meaning” is if the existence of god, for example, is untestable then I can’t notice any effect no matter what I do so I have no reason (as a hedonist) to try to do what I would predict a god would want. In the same vein, regardless of what is “meaning” is the Copenhagen interpretation of quantum mechanics is untestable then I have no reason to base any choice on the Copenhagen interpretation yet I still may read about it if I find this fun.
What does this mean? Ethics itself, if not defined to include game theory, has no relevance to anybody's choices. Political philosophy, if not defined to include game theory, has no relevance to anybody's choices. Questions of "justice", "fairness", "duty", "superiority", "legitimacy", "natural rights" play no role in choices until proven to be optimal in algorithms of choice. Also investigations into meaning which don't have a practical definition (such as: "anything a human can conceptualize and predict behavior of") will be of minimal use.
There are also many applications from this on human behavior. Such as impressions of fault, deserves, pride, guilt, embarrassment, etc.
If you find this interesting and care to read more you may visit semanticreasoning.org