I have an audio recording that has a very low volume. When I try to amplify and normalise using audacity, the distortion and noise make it impossible to understand the words that are being said. It's a lecture... just words with no background music or anything but with very low volume. Any help would be very much appreciated. Thank you!
Let me combine this in a Community Wiki answer, based on the two comments given on the question.
Using a tool such as Audacity, try applying a highpass filter (say, 400 hz cutoff frequency). This will remove the low frequency environmental noises that tend to have high amplitude and low information.
When applying the high pass filter on 400hz, and you want to further optimize for speech only, you might add a low pass filter at 2500hz or 3000hz (experimentation needed. Quality of the sound will go down, but intellibility may go up.
If you only understand the words, you can indeed apply the filters as suggested, prior to amplifying the signal.
This is often called "Narrow Band" in telephony, and this is defined as being between 300 Hz and 3400 Hz. (Wiki Source)
You can experiment with a high-pass between 200-400 Hz, and a low-pass between 2500 Hz and 4000 Hz. It is impossible to determine here which setting will yield the maximum result.
The higher the high-pass, the "thinner" the voice will sound. The lower the low-pass, the "muddier" the voice will sound.
It is impossible to determine if this helps at all, but this would be a good start.