I think part of the reason ExploderWizard feels that 4e is "railroady" is that the published adventures (at least the early ones) were really focused on the encounters.
If people are going to judge an edition's Merit based on adventures produced by WOTC for that edition, then late era 2e, 3e, AND 4e suck
I tend to agree with JeffB here. I'm thinking of a couple of 3E modules - Expedition to the Demon Web Pits, and Bastion of Broken Souls - and as written, they are nothing but strings of pre-scripted scenes which the players must move their PCs through more-or-less in the order written or the adventure won't work as written. Bastion of Broken Souls even has NPCs (the angel-gate, the exiled god) who won't negotiate and will always fight to the death.
When I ran Bastion of Broken Souls I just ignored those directions to the GM. When I use 4e modules, I ignore similar directions. 4e doesn't raise any special problems here.
Thus we get to the root of the problem. There is literally nothing for the players to accomplish that isn't simply pushing a button (rolling a die).
Huh? I've played a lot of classic D&D (B/X, AD&D) and action resolution in those systems involves dice-rolling too. For instance, in the back of Moldvay Basic there is a discussion of how a GM might resolve a player jumping over a cliff in the dungeon to escape from overwhelming enemies, and it suggests a low percentage chance of landing safely in a stream below.
That doesn't make B/X a railroad or devoid of meaningful choices - the choice whether to stand and fight, or to jump, seems reasonably meaningful. 4e is no different in this respect.
I ran a 4E campaign for over a year, sandbox style and never needed a single skill challenge. There was certainly skill use when required. More complex situations were simply roleplayed out. The chief determining factor in the resolution of those situations was active player input. Meaningful contributions by the players is what maintains interest and investment in the game.
Obviously player input and meaningful contributions are what maintains interest and investment in the game. That is completely orthogonal to skill challenges - which are about the method for determining the outcome of those player contributions. (As opposed to GM fiat, which is an alternative resolution technique but not necessarily an antidote for railroading.)
B/X has more helpful DM advice in this respect in its meager 128 pages than the whole library of 4E DMGs.
I actually don't reacall that B/X has any advice on avoiding railroading. It is focused more on various techniques of action resolution (in the absence of a uniform resolution system) and dungeon stocking. But if there's stuff I've forgotten I'm happy to be reminded!
Who said they are?
I never said 4E encourages GMs to ignore action declaration. It does (at least as the published adventures I have read indicate) limit those action types depending on the framed scene.
This is like saying that because Keep on the Borderlands doesn't tell me the difficulty of negotiating with the evil priest, but does tell me AC and hp for that NPC, therefore the players can't talk and can only fight.
Gygax took for granted that the GM would use the appropriate action resolution techniques to respond to player declarations of action. 4e is no different. If the PCs want to negotiate with an NPC, the game has the mechanics for resolving that.
If you have played with GMs who ignore player action declarations - and so, for instance, just ignore players who declare negotiation attempts on the part of their PCs, and only accept declarations of combat actions - then I have some sympathy. How did those GMs cope with 3E and earlier modules which only provided combat stats for NPCs?