Even so, the scientist does not live in the experiment. The scientist lives in reports, definitions, hypotheses, lab talk, papers, and teaching. A measurement that cannot be named cannot be shared. A law that cannot be stated cannot be tested by anyone else. Data do not walk into a community; sentences do.
Consider what actually happens after a result. Someone must say what was counted, in what units, under what definition of the thing counted. “Significant,” “control,” “error,” “cause,” “species,” “energy” — each of these can be used as a tool or as a fog. The student who treats words as secondary will still use them; he will only use them without noticing when they have stopped meaning what he thinks they mean.
A graph can be made to argue without saying that it is arguing. A table can hide a definition in a caption no one reads. A method that cannot be written so that another person can follow it is not yet a method of a public science. It is a private knack that happened to work once.
The cost of treating words as secondary is not that science fails on Tuesday. It is that the student cannot say what the result was, cannot tell a definition from a slogan, and cannot see when a picture has been asked to do the work of a claim. Method without an art of speech becomes a private skill. Science, which has to be taught and checked by others, cannot remain that.