Through development from birth, humans have an automatic tendency to read lips. It is part of the way we interpret communication subconsciously. When we are in a loud room trying to talk to someone you lip read. If we are talking to someone with an accent and is a little harder to understand, we lip read. Even if you are not in a loud room or talking to someone with an accent, we still lip read. It's a part of human nature from the first language you learn until the day you die, you will lip read subconsciously (or consciously). Whenever you are able to see the person who is communicating with you, you use their lips (and facial information/expressions) to better enhance the words coming out of their mouth.
The McGurk effect is a phenomenon introduced by cognitive psychologist Harry McGurk and it demonstrates how much the brain relies on the use visual speech information when communicating through auditory speech. The McGurk effect is an illusion where you are hearing one sound, being shown a second sound but what your hearing is being overridden by what you are seeing. Here is how to test out this phenomenon; Get two friends and have one stand in front of you, and one behind you. Have the friend standing behind you repeat "ba, ba, ba..." and in the same time have the friend in front of you mouthing "va, va, va". Although to begin you do hear your friend saying the ba syllable behind you, but as you focus on the friend in front of you that "ba" sound turns into a "va" sound. In the labratory of Lawrence Rosenblum, they found that the visible "va" overrides the auditory "ba" 98% of the time. That's important to consider because that is proof lip reading is an important part of auditory communication.
The McGurk effect can occur in many different ways, another example is hearing a sound, being shown a second sound, but hearing a third undisclosed sound. For instance, you have two friends like in the first experiment, one behind and one in front, and they are both saying two different syllables "ba" behind and "ga" being mouthed in front. While you are focusing on your friend in front, you are perceiving the "ba" as third noise, "da". In this case, your being compromised between what you're seeing and hearing and this is why. If you articulate those three syllables, "ba", "ga", and "da" slowly you will realize that "ba" "da" and "ga" are all correlated in the sense of how you use your mouth to articulate those words.
If you do want to test this phenomenon and you are alone, there are videos on youtube of people performing the McGurk effect, they have an audio going on in the background while they are mouthing a different syllable/word.