I *do* define "puzzle" very broadly, and it's not so I can split hairs. It's because if the game offers something for the player to meaningfully interact with, then the player (a) cares about what the outcome is, and (b) may fail to get one outcome vs another. (That's pretty much how I define "meaningful".)

That applies to Galatea. If you want to call Galatea "puzzleless", I'm okay with that; but the discussion is then about how Galatea presents outcomes which don't feel like failures, even when the player gets sidetracked or doesn't know how to approach a given subject. And then those techniques are equally applicable to "traditional puzzle" scenarios.

Hm. So if Galatea, or a conversational NPC of that complexity, were a gate guard -- she would de facto be a puzzle. (Presume the player is outside the gate and Chapter 2 is inside.) You could imagine some late-conversation states in which the guard allows the player in; and others in which the guard mentions a back door, or offers to arrange a meeting with the person you're trying to meet, or various other ways to move the plot forward. The player is going to perceive that as a puzzle even if *every* outcome leads to Chapter 2, because the player's impression comes from first impact.

The main problem I see is that the puzzle serves several roles in an IF game at present:

It's a gating mechanism -- allowing certain content to be metered or privileged.

It's an achievement -- players (often) want to overcome challenges for a number of different reasons, including pride, score, etc.

It motivates exploration of the world model -- the player's attention is focused on aspects of the world model that might prove useful in overcoming the puzzle at hand.

It can facilitate characterization and plot -- the PC's personality can be established through responses to action, and world-model puzzles generally require varied actions to solve.

Any non-puzzle mechanism is going to have to somehow fill those roles, in which case you have the "if it quacks like a duck" issue. Either that or you'll have to dispense with or modify one or more of those aspects, which then takes away a major tool used to make an interactive experience compelling.

Not that it can't be done, but I think we'll need to have our definitions straight first.

Jeremy: Does that mean your answer to the title's question is "yes"?

Matt: Just a clarification: I'm not arguing that puzzles are inferior to other game mechanics, just that some variation would be good.

Puzzles being so hugely dominant is as much a 'problem' of choice of words.