You're not disagreeing. I said the same thing!
I went further and said that in paradigmatically maps-and-notes-based play (ie classic D&D, OSR-ish RPGIng) I don't even think the idea of "railroading" is really apposite. The flaws of those games are boring dungeons, killer dungeons, Monty Haul dungeons etc.
In a different sort of RPGIng maps don't have to be railroading either - in my Classic Traveller game I used floorpans of the Annic Nova when
I adapted that old module to run a Alien-type scenario. That wasn't a railroad because the players' didn't have goals about
what they would find in the vessel or
how to make it through the vessel. So the particular geographic layout was essentially neutral vis-a-vis the players' goals.
But in examples like
entering the castle or
making it into the tower via the catacombs or
finding one's way out of the dungeon after being teleported away by a crypt thing then the geography is fundamental to the players' goals. And at that point it absolutely can give rise to railroading.
The same thing becomes true in relation to NPCs and NPC reactions. Deciding in advance that the NPC likes donuts and doesn't drink wine seems like most of the time it will be harmless: if the players want their PCs to get on the NPC's good side by giving a gift, they're going to have to take a stab or ask a friend of the NPC what s/he likes. Provided it doesn't become too tedious that all seems like harmless, maybe even fun, colour.
But if one of the PCs is a baker, or is a vintner, then the GM's choice there takes on a whole new significance and runs a very signficiant risk of being a railroad, or a shutdown of that PC, or something similar in that neighbourhood. Because the baker PC presumably uses the provision of baked goods to make friends; likewise the vintner PC presuably likes to engage others with his/her love of fine wines.
In saying all this I thik I'm mostly just elaborating on what
@hawkeyefan has been saying. But I think I'm also adding a gloss that
we can't say, in the abstract, that some techncque (eg maps-and-notes)
is or isn't railroading. It depends on the particular context of play. But I think we can say that some techniques are not well-suited to some fairly common approaches to play: eg NPC-as-puzzle (ie the analogue of maps-and-notes in social resolution) may not be well-suited to a game that wants vibraint, verisimilitudinous, rich and engaging social encounters. With the OP as Exhibit A as to why.
I'd probably still disagree about the nature of maps vis a vis railroading and player goals, at least in the abstract. A map can be a railroad, for sure, but that's down to map design. The maps that are probably index a preference on the part of the designer to limit options and the force the action in particular ways. Obviously I missed you general thesis that it's mostly a matter of can and not will, to which I wholeheartedly agree.
As far as NPCs goes, I also agree with the above, but I think the example your using, of bakers and baked goods, is perhaps somewhat trivial as far as railroading goes. If the NPC is abstemious then that is a pretty significant part of their character. That doesn't make them not gracious though, so it wouldn't have to affect the player necessarily. It's also the case that a hundred different gifts might do in that case, and the fact that one linked to a PC won't in that case doesn't really register for me as railroading. Not because of the character or choices particularly, but because it doesn't make sense to me that the whole game world is full of people who love muffins just because one of the PCs is a baker. Don't get me wrong, I agree with the spirit of what you're saying, and it's possibly nitpicky of me to engage with the specific example.
If I'm reading what you mean as NPC-as-puzzle correctly, which I think broadly means that the NPC has one handle and it's the job of the PCs to figure out what it is, then I'd agree that it's a poor way to inculcate vibrant social interact. That's not really a character though, it's a cardboard cutout. I don't approach NPC design or play that way at all.
Well I don't know much about how you run D&D because I don't think I've seen you post much actual play.
That's mostly on purpose. I find that examples of actual play tend to sidetrack conversations. If I had one that fit a convo perfectly I'd use it, but mostly it feels like trying to fit the evidence to the thesis for me. Also, my memory for the fine details of play is poor for the most part, at least as far as being able to recount them as blow by blow recaps. I envy your ability to summon up that level of detail seemingly at will.
<snip>
To finish this post: if your maps and notes are not to resolve action declarations, what are they for? I can tell you what mine are fore - eg as in the example of the Annic Nova. They're to support framing. The same thing happened when we played Wuthering Heights: we needed to know how long it would take to carry a body from Soho to the Thames and so I Googled up a map of London. From that we could get a time range, which then interacted with the rule that a ghost (ie the ghost of the PC's body which was to be dumped in the Thames) manifests a certain number of minutes after death.
But I know from this and other threads that
@Lanefan and
@Maxperson are using maps-and-notes to resolve action declarations. This comes through absolutely clearly in the most recent discussion of the castle to be entered.
What do you use maps and notes for?
I think there's something pretty interesting to be unpacked here. When you say you use them to support fictional framing, I feel like I agree with you, as in that's what I use them for. However, if you asked me what well framed fiction was specifically good for, I'd probably say to aid in adjudicating actions. The two ideas seem pretty closely linked to me. There must be some nuance there that I'm missing in order for you to present those a two quite different processes.
One difficulty, for me at least, is that NPCs and physical spaces don't play the same way, or at least I don't use them the same way. I don't usually bother with physical maps for anything smaller than a real dungeon, and even then, I don't really see the connections of the physical space as something that limits player choice in a negative way. I mean obviously it does limit choice, when there are only two corridors you have only two choices, but that seems trivially obvious. Maybe it's because I don't have room contents in the way that a published module does that I'm struggling here. My 'dungeons' tend to have sorts of inhabitants, and sorts of possible treasure, and there may be some loosely strung together encounters, but those aren't tied to rooms. The idea that X is waiting in room Y has never made any sense to me, as it makes the place enormously static rather than responsive to the players actions. As soon as the players hit a dungeon, the inhabitants are in motion, and where the players might encounter X, Y or Z, has everything to do with their choices and nothing to do with the map. My maps are just a tool to keep me colouring inside the lines when it comes to obeying the laws of physics, really.
As for NPCs, what I actually do in play varies a little. I will have notes of some kind for important NPCs, but in other cases I don't. In the latter case the fiction guides my responses and thus adjudication. So lets say the players encounter a guardsman of some kind. If I don't have notes, his responses and personality are going to be defined by what a 'usual' exemplar of that role would be in the fiction. A poorly paid guardsman snoozing next to a dilapidated warehouse door is one thing, and the highly trained and motivated guards outside the kings bedchamber are something else. In each case I don't have anything predetermined about responses, I just react to whatever the PCs decide to do as an approach in a way that makes sense for the role of that NPC. That is modified in each case by any success or failure states from PC actions that might apply going in (a failed stealth check, a successful performance check for a disguise - stuff that sometimes gets rolled for before engaging with an NPC).
I don't ask for SI rolls to start, I ask for declarations. There, the answer to
what do you do? sets the parameters for how the actions unfolds. My description prior to that question usually contains some information that the players can use to guide their actions - the guard seems nervous, or the clerk seems busy and annoyed. If I have notes, those also often have some drives or personality traits that I'll use to frame an ongoing interaction when they matter. My first hurdle for social interaction is usually about objections. If the PCs are asking for something and there's no reason the NPC shouldn't comply, then there's no rolling involved. In the case of a nervous guard, the fact that he's nervous is probably the initial hurdle, set in the context of him doing his job. To go back to your bakery example, if I had an NPC who from my notes was abstemious and also a dick, that will inform his response to the offer of pastries. It won't dictate his response mind you, because that completely negates all the nuance that exists in various specific player approaches and declarations, but it will inform his response.
I'm not sure where that leaves me on the question of what do I use notes for relative to framing versus adjudication.
