This is known as the Whorfian Hypothesis, aka Sapir-Whorf theory. In generalized-to-the-point-of-inaccuracy terms, the idea is that language constrains thought. It’s one of those ideas that we can perceive as intuitively correct but that does not stand up to experiment.
There are, for example, languages that don’t have words differentiating green and blue, and others whose counting numbers don’t include specific words for numbers larger than two. Some languages have no words for cardinal directions but use terms like “mountain-way” and “ocean-way.”
Experiments do seem to support a weak version of Whorf - people from cultures with “missing” words can differentiate between green and blue for instance, but it seems to take a bit longer. There’s also a paper indicating that people who don’t use cardinal coordinates have a better innate sense of orientation when, eg, walking corridors in an enclosed building.
I’d personally fall between the weak and strong position because I do not believe in free will and do believe that semantics are a significant driver of behavior, but that’s a step beyond where most of the current research is. There’s research into free will, but none that I’m aware of that pulls in cognitive semantics as a driver.