Human ears were seemingly manufactured to be most sensitive to frequencies of sibilance in the human voice (3000 Hz), and surprisingly insensitive to lower and higher frequencies until the volume is quite high. This gives us the best advantage to understand the MESSAGE, even with much distracting background noise. Stuff like droning air conditioners and clanking dishes are effectively attenuated. We are also set up to respond to audible sensations in a roughly logarithmic fashion. This is necessary because the range of sound energy produced by everyday objects varies by a factor of over 100,000 (quiet room to noisy street traffic). A linear ear would be either be overloaded all the time or deaf to anything but shouts. As a matter of fact, ALL our senses are logarithmic for similar reasons.
This logarithmic nature means that the noise produced by two horns is perceived as only slightly louder than one horn. We also require about a 10 TIMES increase in energy to perceive something as TWICE as loud. A drop to 1/10 the energy is perceived as HALF as loud. These energy vs. perceived loudness properties were studied and quantified in the 1930's telephony work by Bell Labs.
Engineers designated the smallest detectable change in volume, in a very quiet room, by an average adult as 1 deciBel (dB) in honor of their employer. This change represents an energy increase or decrease of about 12%. An easily perceived change in loudness is about 3 dB (about 30% change), while TWICE/HALF as loud is about 20 dB (1000% change).
Another interesting feature of humans is our ability to mask noise, or ignore sounds entirely. This is because our senses are also RELATIVE devices. We are very poor at measuring the ABSOLUTE level of sound, light, touch, etc., but very good at noticing the DIFFERENCE between two sensations. Applied to the ear, it has the ability to adjust it's noise threshold so that noise is only perceived if other coherent (not random) sounds are at a similar volume. If the coherent sounds are much softer than the noise, we won't hear them (like using white noise in a office setting to mask conversation). Conversely, if the coherent sound is much louder than the noise level, we won't hear the noise, even though it is still there. Dolby has been exploiting that hearing defect/feature for over 30 years.