LogFAQs > #962381504

They recorded it a lot like how the dialogue read in their heads lol.....
See Manga 2 Anime & Comics 2 Cartoons & Novels 2 Audio

Cheese was adored & often voice work was made to match perceived quality in animation or graphics or maturity level. The is I think if your audience is supposed to be a child, then talk at them like they're a child. I think they were emulating stuff like Star Wars or Power Rangers too....

Elder Scrolls & Fallout, for example, have yet to break this rule. Characters all kinda look like shit. The recordings kinda sound like shit. Makes for cohesive experience and encourages combat/exploration.

GTA does the same thing. NPCs in general, especially throwaway NPCs, tend to have a lower polygon count and a lower quality audio in their voice to match.... The player may end up viewing them as annoying or unimportant, and flee off to find the boss of the level, or try to kill the NPC, which are exciting & addictive staples of the gameplay. They also tend to lean on lower quality recordings so as not to raise the bar too high on what the player expects from the rest of the game. Also, by keeping it cartoony, they can match better with other multimedia... like cartoons for example. What voice work was done in the SNES days was equally cheesey if not way way worse too, so they were largely working from that point.