I'm not ignoring the part that you bolded. The player has inserted him-/herself into the situation of the character - and hence has the problem to solve.
What you are ignoring is that he has to solve it as the character itself would solve the puzzle and not as the player would solve it. That means that a stupid PC might not be able to solve it, even if the player could.
The question of what knowledge the player should use is not addressed at all in Gygax's essay, is it?
Yes it is. There is a paragraph about stepping into the person of the PC and playing the game as if you were that PC. The PC's knowledge is very obviously a part of that. The later paragraph about solving it as the PC and not the player is also very obviously telling you which knowledge to use.
I know that it was pretty common in the early days of D&D for players to memorise the Monster Manual - that's one of the reason why new monsters (and new trap ideas, and new spells, and new magic items, . . .) are so important to the game, at least in that period. They are a way of surprising players, and presenting them with a problem to solve - rather than expecting players to pretend to be surprised when they actually are not (eg pretending to be surprised by trolls' regeneration, when in fact they're familiar with it), and to pretend to solve a problem to which they already actually know the solution (eg pretending to try to hit upon fire as a solution to the troll problem, when in fact they already know that fire is the key).
This goes to the contrast that Gygax draws between role playing (in effect, acting) and role assumption. Here are what I think are some key passages from the essay in this respect:
No. No it doesn't contrast that. The ability to create new things has nothing to do with how one plays the character. To the character, every monster not encountered before is new.
Role assumption is stepping into the PC and playing that PC as if it was a real person with all the flaws and strengths of that PC. Acting is going off of a pre-set script where everything was pre-established. Nothing about that difference is contrasted by you making a Borc instead of an Orc and changing the way it looks.
A role-playing game should be such that players begin the personification portion as role play, and then as they progress the activity should evolve into something akin to role assumption. This does away with stilted attempts to act the part of some character. In place of this, players should try to become that person they are imagining during the course of the game, and conduct the actions of their characters accordingly. . . .
Stupid acting stupid is conducting the character's actions accordingly for the person the player of a 5 int PC is imagining.
Combat, survival amidst threatening conditions, or stalking an opponent are typical means of adding excitement and suspense into the whole. These are action oriented portions of the game activity which call for little role playing but a fair amount of role assumption. The magic-user character (and thus, the player of that character) must know his or her spells and how to utilize them efficiently. The explorer must know outdoor craft. Whatever the situation, setting, or character being played, skill not theatrics is what is called for here. . . .
Right, and choosing the right stats for the character is a large part of being able to make those choices properly. The paragraphs you are choosing don't override the others that limit character choice based on stats.
Role-playing games are different from other games in that they allow participants to create a game persona, develop this character, and enhance his or her skills and abilities. While some considerable amount of acting is most beneficial to play, this is by no means the sole objective or purpose. The fun of such gaming includes all the other elements mentioned, plus the interactive relationships which develop between the various characters of the players participating. In the well-balanced game, role playing should quickly become role assumption . . .
The existence of other goals or purposes does not invalidate or negate the game persona portion. They have to take that portion into consideration, at least as far as this article is concerned.
Not every game of this sort must be completely balanced with regard to all of these aspects. Such a decision is entirely in the hands of the game master and the players. If a particular group desires to stress acting, or combat, or problem solving, or any other singular feature of the whole, that is strictly up to the individuals concerned. How they enjoy gaming, and what constitutes fun, is theirs alone to decide. . . .
And here he is saying, "But all this is just my opinion. Your opinion counts more."
There are a few interesting things here. First, he emphasises the breadth of feasible approaches - a greater emphasis on his preferred style (of player skill-based "role assumption") or on his less-preferred style (of player theatrics or "role playing", which would include pretending to be surprised by trolls when one actually knows what their abilitiies are). Although Gygax has strong preferences, he is not stating a purist position about what he thinks is the only way to play D&D.
You are not understanding him properly. Assuming a persona is not acting. Let me give you an example. When you watch Sheldon on Big Bang Theory, the actor is acting only. He is scripted with virtually everything he does, though some small amount of improv probably happens. Once I saw the actors on a show being interviewed. They were asked a question and then asked to answer it as their characters. What they did there was not acting. It was assuming the persona of their character. They had no script, yet the result was virtually indistinguishable from acting. They remained true to their characters, even though it was entirely improv and only persona based. The latter is what Gygax says should be done, and the former avoided.
Second, look at what he says about combat. In combat, theatrics do not loom large. But a skilled player is able to assume the role of the MU, by knowing the spell list and making sensible choices. Likewise, he assumes that a player of a ranger ("explorer") will actually draw upon knowledge of outdoor craft to solve ingame problems (eg thinking of ways to use trees to make rafts, or vines to make ropes, or to signal via fire and smoke, etc). This is his preferred approach. (Though, I stress again as he himself does, not the only viable approach.)
Look at every PC and NPC wizard he ever made or ran for. They have very high int scores. He is assuming that the wizards he mentions in the article aren't imbeciles, but rather highly intelligent PCs who are capable of making those choices when the player assumes that persona.
When playing in Gygax's preferred style, if it important to the table that 5 INT matter to character decision making beyond the language rules and INT checks, then the GM should be filtering information provided to the different players on the basis of character INT, so that when the player takes on the situation of the PC, the fictional position is true to the PC's INT.
No. From his prior paragraphs, his style is clearly different. He states the table can change things from what he recommends in his prior paragraphs, but that statement doesn't become his style.
I also think it reminds us that there are multiple ways to do RPGing, and it shows us an author with a strong preference nevertheless acknowledging that others will enjoy doing it other ways (and even recommending that modules, tournament sessions etc advise prospective purchasers/participants which playstyle they are suited to). In this respect, I think it fits with [MENTION=97077]iserith[/MENTION]'s position in this thread, that to insist that a character having 5 INT must give rise to a question of "How do I roleplay that" is to take one preferred approach (broadly, the one that Gygax identifies as theatrics/"role playing" in the narrow sense) and dismiss other completely legitimate and viable approaches (such as Gygax's preferred "role assumption" approach).
Yes. Do it your way is a key point, but that doesn't change his prior paragraphs to mean something else. He feels that you should step into the character and play that persona, making choices as the PC only, which limits the player in his ability to make key decisions if the PC is stupid.
When Gygax talks about "theatrics", I don't think he is just getting at literal improv acting. He is getting at the idea that playing one's PC means self-consciously adopting another persona and posing the question "What would this other person do?".
Hence his call for "role playing" to quickly become "role assumption" - the persona of the PC is developed fairly quickly, and the PC is oriented in some sort of challenging fictional situation, and now the player's job - having put him-/herself imaginatively into that situation - is to solve the challenge using the resources that the game makes available to him/her.
If your PC's INT is 5, that's one resource you don't have. Your PC's 5 INT might also have contributed when you took the early steps of establishing a personality for your character, which helps shape the PC's goals and methods, and hence helps give content to the character's fictional position.
Yep. That 5 int shapes the goals and methods by limiting the player to playing stupid in most situations.
But other than these ways, Gygax is not expecting that the 5 INT should be a constant consideration in making action declarations for your PC. Because that would be engaging the situation from the point of view of a player ("How do I, this real person in the real world, roleplay a 5 INT") rather than from the point of view of the character ("How do I, Thrud the Magnificent, solve this problem that confronts me?").
Wrong. It's the exact opposite of that. By putting yourself into the shoes of the PC and asking yourself, "What would an imbecile do here," you are stepping out of the view of a player and into the view of the PC.
Again, if it's important for the table that Thrud's 5 INT contribute to the way Thrud (and hence Thrud's player) encounters that problem, that falls onto the GM to factor it into the way information is dispensed.
The DM should never have to play or control the PC unless some sort of magic is at play that is causing it. If it's important for the 5 int to contribute, it's the player's job to roleplay it.
Speaking now for myself rather than as a Gygax interpreter, I think it's enough that Thrud's player doesn't have INT to draw on as a resource, and hence can't declare knowledge check, certain sorts of investigation or perception checks, etc, with much prospect of success. I don't also see the need for the GM to factor into his/her framing of Thrud's situation. But if I wanted the INT 5 to matter in that respect, I would absolutely put it onto the GM, not the player - as I have repeatedly stated upthread.
As has been pointed out, you can build around that and close up that little flaw........which I view as a flaw with the game itself.