Part 1:

Since your question is written from the perspective of a Marine, I will try to use a Marine's quotes to answer your question.

For the average Marine coming into the Pacific theater, their pre-conception of the enemy would be formed from the many (exaggerated) stories from combat veterans who had served in previous engagements. They would not be coming to the realization after engaging the enemy, but would already have a preconception that the Japanese would not surrender.

E.B. Sledge, in one of the precious few detailed perspectives we have from the enlisted ranks, wrote about his own preconception during his first day of D-Day combat on Peleliu in his book "With the Old Breed":

"The Japanese counterattack was no wild, suicidal banzai charge such as Marine experience in the past would have led us to expect. Numerous times during D-day I heard the dogmatic claim by experienced veterans that the enemy would banzai: 'They'll pull a banzai and we'll tear their ass up. Then we can get the hell offa this hot rock and maybe the CG will send the division back to Melbourne."

As General Merrill B. Twining (who mostly fought in Guadalcanal) writes in his memoir "No Bended Knee": "Information spreads rapidly in the naval service, whether it be the official word or unofficial scuttlebutt (rumors and gossip exchanged around the ship's drinking fountain). Sometimes scuttlebutt precedes the official word. Often, it's the 'straight dope.'"

But where did these preconceptions come from? From the first real major sustained combat engagement that the Marines performed: Guadalcanal. On Guadalcanal, at least, the Marines started to come to this realization about the enemy's view of surrender within five days.

Taking an enemy's surrender, especially during combat, entails significant risk , and it only takes a small number of well-publicized incidents for an average marine to simply not bother taking that risk.

Roughly, there are four ways to take prisoners:

1. Accepting voluntary deserters during periods of quiet,
2. Involuntarily capturing soldiers during combat operations,
3. Voluntarily capturing individual soldiers who surrender when outnumbered,
4. Accepting the surrender of whole units of behalf of the commanding officers.

I am assuming you are referring to the two "in-combat" methods of taking prisoners, but I'll touch on the others as well.

As it stands, there were, all throughout the conflict, multiple instances of taking prisoners. The Marines actively attempted to broadcast messages of surrender to stragglers in Guadalcanal. In the best, and safest, case, prisoner-taking came in the form of starving groups of unarmed Japanese voluntarily walking with their hands up to enemy lines to give themselves up to groups of heavily armed soldiers. From below, Twining writes about these broadcasts:

"In one case, a Japanese platoon at Gifu [a Japanese nickname for Mt. Austen] that heard our broadcast urging them to surrender decided they were too ill and weak even to walk to American lines."

However, the Japanese began to establish a reputation about themselves among Marines as very dangerous to keep as prisoners, especially if involuntarily taken, and always posing an ever-present risk to their captors. In such a situation, moral questions get put aside. Twining hints extremely carefully about a dark incident during the conflict on Guadalcanal:

"General Vandegrift had received a letter from a missionary located in the remote interior of Guadalcanal who had rescued a downed Japanese aviator. Although both injured and wounded, the pilot had somehow survived when his fighter crashed in the jungle. Well cared for by his rescuer, he began to recover. As the pilot gained strength, he became increasingly hostile, until the missionary felt threatened with the loss of his own life. What would the general recommend that he do? I was certainly happy not to take part in that decision. We all agreed that it was strictly a matter [for the personnel officer] and the rest of us kept clear."

Despite this, there was never a point at which there was any expressed policy to avoid taking prisoners. In fact, it is the opposite. Prisoner-taking is not exactly a moral question. Intelligence sections have a high demand for prisoners in order to extract detailed information about troop movements, etc. Even if the prisoner does not talk, it is still possible to learn many things just from the appearance, belongings, dialect and attitude of the prisoner themselves. The Marines, in the sense of the military organization, never wavered in their practical desire to take prisoners. Marines, as individuals, however, were much less enthusiastic about being the one that takes the personal risk to get a prisoner.