I disagree in the bolded text. In order to go where the action is, you need to know what the PC's action is. Not the PCs motivation.
If the PC attempts to haggle for some Calishite silks, I (the DM) don't need to know that he wants to purchase them for his mother. To play the NPC haggling, I need to know what the NPC's motivation is.
I've been out of the conversation for a stretch, so I'm just commenting on a few bits here that I've seen with a quick scan.
On this above, when [MENTION=42582]pemerton[/MENTION] invokes the principle of "go to the action", he is referring to one (or both) of two specific types of action:
a) System agency: The premise that the game itself is fundamentally built around. If you are a PC in Dogs in the Vineyard, being a player means you have already bought into the play paradigm of being a flawed, vulnerable-but-stalwart, gun-toting Paladin in a Wild West that never was, meting out justice and keeping the peace in towns shot through with sin. If you decide that your Dog has lost faith, has had enough of this impossible life of service and wants to hit the trail for a Boom Town. Alright then, retire him and briefly memorialize him for all of us with that story. But make a new Dog that is embedded into that premise. Because this game isn't called Lost Dogs and Boom Towns where we follow your fallen Paladin through his pursuit of fortune.
b) Player agency: The premise, themes, and tropes that the player has signaled they are interested in via the PC build mechanics.
Its not enough to have NPCs with impulses and drives of their own. Well, its enough if "the action" is
"whatever spills out of an honest, premise-neutral, organic-outgrowth (but cognitive bias afflicted...this we cannot do away with no matter how much we feel our model bears fidelity to an impartial, properly parameterized fantasy world simulation)
rendering of a person/place/thing in accordance with those animating factors."
But that isn't "the action" that pemerton is alluding to. He's alluding to the centrality to every moment of play of (a), (b), or (often) both. You could call this "premise logic", "genre logic", "drama logic"...what-have-you. Regardless, it is the prioritization of that (as "the action") over and above the prioritization of what I have bolded above (as "the action") that is fundamental here.
So when you're framing scenes or dynamically changing a situation post-resolution or tallying up the fallout after the dust has settled, that prioritization is the driving factor behind your GMing. Now the typical response to this by folks whose RPG mental frameworks are steeped in Sim priorities is incredulity like; "well, nonsensical, irrational NPCs or an incoherent setting seems inevitable with that prioritization." I promise you, games that feature "drama logic" (let's just go with that) don't eschew sense or coherency.
Just one other bit right quick.
A Star Wars game could trivially emerge in a "play to find out" fashion through either (a) or (b) above:
(a) The system's premise itself (and the machinery, resolution mechanics/PC build mechanics/reward cycles, therein) is about discovering the legacy if your heritage and either redeeming it...or falling prey to it.
(b) The player has a Relationship statement about his father that offers both a d4 (which creates complications) and a d8 (which pushes toward success) to dice pools:
"My aunt and uncle always say 'I have too much of my father in me.' They don't speak of him beyond that. I sense a strange pull to discover his fate that goes well beyond curiosity."
Boom. Off we go. We can easily "play to find out" how a scrubby water farmer can get to Jedi Knight or Seduced By the Dark Side or (hell) killed by Sand People/Hoth Yeti from that.