Do/why do languages trend away from synthesis over time?

Do languages become less synthetic over time? Some languages certainly have; why does this happen?

Synthesis, in terms of linguistics, refers to the morpheme-to-word ratio. A language that is more synthetic has more morphemes per word. In Turkish, for example, one word “gelebileceksin” means “you will be able to come”. It takes us 6 words in English to say that, but 1 in Turkish, thus Turkish is more synthetic than English.

Latin, for example, is decently synthetic. A word like “amaverit” means “he will have loved”. However, over time, as Latin became the Romance languages, it became less synthetic. Words like “habere” (to have) came to be used to express verb tense; instead of one word “domi” meaning “at home”, the word “ad” was used to mean “at” and then word for “home”.

English lost noun cases in favor of word order and prepositions expressing grammatical function; in other words, it lost synthesis.

Why does this happen? Does it ever happen in reverse where a language becomes more synthetic? I can think of no examples, but I don’t know a lot about this in the first place…

I know this question is very “linguistic” and maybe non-Jeruba types will find this boring or confusing, but I’m really trying to find out if this is a real trend and why it happens!