I don’t think so. I think there are strong objections to the single home theory, but they don’t touch the idea of a word being at home in a language-game, having a role or a function. It’s easier to see in the negative: if you’re working on a bit of carpentry and you have the wood, hammer, nails, screws, drill, ruler, sandpaper, and so on, then the soldering iron doesn’t belong here. — Srap Tasmaner
I think you are changing the subject by switching to the negative. The issue is the affirmation that a word has a home (or possibly more than one home). By proceeding in the negative, as you suggest, all you can do is keep saying 'this is not the word's home', and 'that is not the word's home', so on and so forth. You would always be left with a multitude of possibilities such that if meaning is directly related to the home, as Sam26 suggested, we could never have certainty of meaning.
I know that we are not necessarily looking for certainty, as you say, we are simply looking to accomplish a purpose. However, the context of Sam's post indicates that the issue we are dealing with is the question of distinguishing one "sense" from another, and in the case of logic, certainty is the purpose. So the problem is well exemplified with the way that people use the word "know". There is an epistemological sense of "know" which implies justified. Justification requires logical proof, and this requires that a word's use be limited by a definition. Ambiguity and the possibility of equivocation nullifies any attempt at justification.
So you might really be talking about something completely different from me. You are saying, so long as we can exclude misuse of the tool, we can proceed with the tool in a vast multitude of correct uses. Excluding misuse will exclude the possibility of mistake, and the tool will always serve the intended purpose. But you do not appear to be considering the fact that word usage has at least two sides, the person who hears is distinct from the person who speaks. And, the person speaking cannot exclude the possibility of mistake by the person hearing, in the way you propose, because where the word "doesn't belong" varies from one person to another.
Look at Sam's interpretation of Banno's chosen word, "congenital", a few posts back, in relation to my interpretation. I think that this word does not belong in that context, there is no language game which supports that use, and Banno is wrong to use that what in that way. But Sam, in my opinion fabricates a game which supports that use, and claims that Banno is simply within the rules of that game. As demonstrated by this example, and multitudes of other similar examples which abound in this world, your proposal, that we might just decide that a word "doesn't belong here", is completely inaccurate, because someone else will come along and use it there. And this incompatibility between one person saying it doesn't belong, and another saying it does, will result in the word not serving the intended purpose, and misunderstanding.
The homonym business — eh, it’s almost semantics. The one argument against it would be that in introducing a word into a language-game it does not already have a role in, you’re relying to some degree on people’s understanding of how the word is used elsewhere — either for the metaphor, or by making a case that there’s a strong analogy between the known use and the new one. It would be hard to pitch a known word as an empty vessel you can add a new meaning to at will. (A somewhat outlandish metaphor can do the trick. Timothy Williamson got mainstream philosophers to talk about “luminosity”.) — Srap Tasmaner
I don't agree with this at all, and I believe that this is why this issue is so "tricky". I think we have to distinguish between two very different "ways" of "introducing a word into a language-game it does not already have a role in". If you are relying on peoples' understanding of how the word is used elsewhere, then you are not actually introducing the word to a new game, you are forming an extension on an old game. This is the sort of overlap which Banno referred to with "family resemblances". But this is where the game analogy breaks down and fails, though people like Banno will refuse to accept this fact. What Wittgenstein represents as a game, is one specific way of using the word. If we allow now, that "a game" consists of two distinct ways, even if one is related to the other by a family resemblance, we contradict the premise of "a game". Therefore distinct uses must be distinct games despite the reality of "overlap" This is the age old issue in Plato's Parmenides, of the incompatibility between One and Many.
That is the one "way" of "introducing a word into a language-game it does not already have a role in", allowing that the two games have a relationship (family) with each other. And the problem is that this really negates the effectiveness of the game analogy. Meaning is attributable to this relationship between games, not to any game. We now have to assume something within language, which is very significant and important to meaning, which is outside any particular game, as the relating of one game to another. This is equivalent to "the whole is greater than the sum of the parts". We have to allow that there is something which makes a whole a whole, which is not a part of the whole. It's a sort of dilemma, and the solution is to reject the analogy. Language-games are proposed as the parts, but the whole which is "a language" is not a congregate of such parts. Therefore the proposal is unacceptable.
That "way" we might call the natural way. The other "way," we might call the logical way. The logical way is to strictly define the word, making the usage specific to one particular game, thereby excluding all relations with other games. Excluding relations with other games is very important, to avoid the tendency to equivocate. This way is exactly opposed to relying on peoples' understanding of how the word is used elsewhere, because that way of understanding consists of a multitude of relations between games (which "game" fails to capture because the understanding lies in the relations, not in the games) and this sort of understanding is extremely conducive to equivocation.
With respect to the two "ways", the logical way is consistent with the "language-game" analogy, but the natural way is not. So the language-game description really fails to capture the true nature of natural language, being based in the logical way which is opposed to the natural way.
One point from the other direction doesn’t seem to be brought up much: must a word have a single use in a language-game? Why couldn’t a word have multiple uses in the same language-game? — Srap Tasmaner
In essence, this is exactly why the game analogy fails in accounting for natural language. "A game" as demonstrated by Wittgenstein is a single type of usage. To avoid violating, or contradicting that premise, a double usage cannot be one game. So a usage is a game, and this principle allows for the reality of logical proceedings, free from equivocation. However, natural language is directly opposed to this, deriving meaning from a multitude of very distinct usages. One might portray these distinct usages as distinct games, but that's really a step in the wrong direction. The real production of distinct games is the artificial process of creating distinct logical premises, and distinct logical proceedings. The natural process of deriving meaning is a comparison of individual, particular, instances of use, which cannot be portrayed as games. They cannot be portrayed as games because the game representation assumes that each instance of usage proceeds with the intent of establishing a rule for general usage. Natural language use does not often proceed with the intent of establishing a rule, it just proceeds with the intent of accomplishing the purpose in that particular instance. So when a person compares distinct instances of usage to derive meaning, this is not a matter of comparing distinct games, because that intent, of demonstrating a game is not necessarily there in these distinct instances.