The wish to begin from science and data is not a mistake. The world is there to be measured. Experiments can surprise us. A table of observations can correct a story we liked. Anyone who has had to change a belief because a measurement would not yield has already learned something this course never denies: things are not made true by our speaking them.
It is also true that many modern disciplines have grown precise by refusing vague speech. Chemistry does not advance by eloquence. A lab notebook that will not name quantities is not yet a record. The arts of language have no quarrel with that severity. They are not a bid to talk the thermometer into a different number.
Often the student who wants to start from science is reacting to classrooms in which language was treated as decoration while the “real” work was supposed to happen elsewhere. That reaction has a point. If a sentence cannot be checked against what was done, it is not yet a scientific sentence. Data are a kind of severity toward our wishes, and that severity is a virtue.
What is true, then, is this: evidence can correct speech. What has not yet been shown is that evidence can replace speech, or that a student can be formed in a science without first being formed in the words by which that science is taught, recorded, and handed on.