where the writers are
Music and Prose: The Lives of Notes

And words.

Let's start with a note—C sharp (or C#), for instance.

In music, enharmonic equivalents are notes, intervals, or key signatures that have alternative names.  For example, the note  C#  has an enharmonic equivalent in D flat (or Db); the key of C# major has an enharmonic equivalent in Db major.  In equal-tempered tuning, the frequencies (that aspect of sound that determines pitch) of the notes comprising the major scale in C sharp major are identical to those of the major scale in D flat. If the two scales are played one after another, the ear won't be able to distinguish differences between them. They will sound exactly the same.

Yet, no composer who imagines a piece in Db would ever write it in C#.

Why?

The two keys do not exist in isolation: they establish specific and different types of contexts.

Some of a musician's understanding of these contexts has to do with education, the rest, personal experience.

Our first music lessons teach us that in ascending scales starting on C, we sharp notes; when they descend from C, we flat them. If we associate musical sharps with an object that can pierce, or a person who is shrewd, we may come to hear sharped notes as piercing or bright. And if we associate musical flats with a surface that is smooth and even, or a voice that is colorless, we might hear a darkening in flatted notes.

Once we apply a key signature to a scale, its behavior changes, Suddenly, there are new rules for flatting and sharping notes, which are dependent upon the key. And as our ears develop, we begin to sense subtle differences in tone between flatted and sharped notes, despite their common frequencies.

We can't escape that context; nor can we escape our humanity, our subjectivity...

...with either music or words.

Our first reading experiences may be with single words that are accompanied by illustrations of their meanings, but our acquisition of language comes first from those we hear speaking around us. As a result, we learn to associate words not only with their sounds, but also with the specific contexts in which we hear them.

As our reading ability develops, and our vocabularies increase, we learn that words do not exist in isolation, that they are complex and multidimensional.  Words may share properties, but once they are placed within a context, those properties will be both influenced by and dependent upon the words around them, making one choice better than another. If the context is a familiar one, the choice becomes clearer.

That is why many writers would never substitute thump for thud, or like for enjoy.

And why a composer who imagines a piece in Db would never write it in C#.

When our associations with a particular word are so strong, that only one word will fit, function, and sound right within the context we've created, that is the word we must choose.