Reduced contrast is generally achieved with film by developing less than "normal" -- this produces a lower slope on the linear portion of H/D plot (that is, less increase in negative density for a given increase in log exposure), which we see as less contrast. In Zone terms, by developing less, you keep the low zones nearly the same, but the higher exposure zones develop to lower values.

Going along with this is the need to expose more in order to provide correct high values in the positive (print or scan). Together, exposing more and developing less are referred to as a "pull" -- the opposite of "push" in which film is knowingly underexposed and overdeveloped to compensate (perhaps because the light level is too low for the available film and lens combination, or perhaps in order to boost contrast on a low contrast scene).

One relatively common application of pull processing is to produce a "high key" image -- low contrast development combined with higher than normal exposure, then printing at high contrast to bring out only selected features of the image.

In older images, another factor may enter into a "soft contrast" image: lens flare. Prior to the early 1940s, lenses did not receive anti-reflection coatings (the process hadn't been invented yet), so each glass surface interface resulted in light scattering, which (especially with a bright area in view relative to the main exposure) tended to fill in shadow areas and reduce overall contrast. This is also why pre-War lenses were usually designed to avoid high element counts -- triplets, 4-element Tessar types and 5-element Heliars, the largest element count that was at all common before 1945 was the six-element Xenon (I have one, it's an awesome lens, but it does flare some against the light).