One of my students last week told me that on her first day in Cambridge she got lost. She had a map and did what she could to read it, and while it took her two hours to get home, she saw a lot of Cambridge and had on the whole a very pleasant time.
Getting lost is not something we do very much these days. We have the GPS. And that is a bit of a shame. It is good, sometimes, not to feel in full control. The same, clearly, is true of language-learning. A learner’s English will at best be an approximate map of the whole language, at a more or less crude scale; parts of it will be inaccurate, parts incomplete, and in some wide tracts it will be a case of here be dragons (such was my own feeling, for example, with conditionals in Italian – using the word if, I knew, would frequently summon the subjunctive, and that was something I didn’t want to have to deal with for quite some time).
In the British Museum there is a map drawn on buckskin of the lands between the Missouri and Ohio rivers, which represents an attempt to buy land concessions by European traders from Native Americans. It dates from 1775.
The map is drawn by Native Americans, and includes features only relevant to their own mental geography – rivers, Native American settlements, paths of bison migration, and so on; it entirely disregards a number of substantial European settlements in the area. There are European maps of the same region which show only European towns and no Native American settlements.
We make maps to suit our own vision of the world. Language, like the world, is not a finite set of resources to be exploited or garnered or mastered, but a set of possibilities which can be arranged or activated in a variety of ways. How you navigate it is not arbitrary, but neither is it particular restricted. My student eventually got home, no worse and somewhat better for the experience; and in the same way learners will not be much poorer for a bit of meandering in the by-ways.