In many respects, including how it is learned, counting is much
more like language than either swimming or playing chess. Gifted
Martians aside, most of learn to play chess through a process of
explicit instruction, often supplemented by reading about and studying
the game. Goddard's account shows that this is not how a child learns
to count. It is patently obvious that it is not how we learn
language.
Just as Goddard's child begins with the counting rhyme, and from
there moves on to counting, controlled counting, addition, an
understanding of transfinite numbers, &c. through a prolonged process
of imitation and `going on,' a child learns language, as Quine reminds
us, `at his mother's knee.' The child starts with imitation, he is
encouraged, he is gently `pushed' this way, and `pulled back' from
that way. Thus, the child is introduced into the full array of
practices which make up language. There is no sharp delineation
between learning to speak a language and being a full-fledged speaker,
no break marked by acquisition of some bits of knowledge-that. At no
point in the process is the child required to learn rules of meaning
or even to be able to recognize them. Most of the time, the child is
never even given the rules which constitute language. The child's
introduction into the practices which make up language is not,
pace Quine, accomplished by some sort of behaviorist
training. The child mimics, but by itself, this does not constitute
learning. In order for learning to occur, the child must `catch
on'--catch on to the practices, to their standards and criteria;
catch on to using these practices for purposes of her own; catch on to
the power these practices give her to affect outcomes and changes in
her environment. In short, the child must catch on to all the things
Dummett tells us language can do.
We succeed in speaking to one another, in arguing, in presenting reasons, in pleading, and so on, only because we each hold ourselves and others accountable for acting in accordance with the standards of the practices which constitute a language. This is simply to say that we are subject to correction. It is this constant reminder from the linguistic community of what we say and what we don't that keeps us on track, that allows us to use language to do all the things that it can, including making changes in what subsequently happens. Language is possible because we form a community of users engaged in a shared set of practices; as members of that community, we are responsible both for acting in accordance with and for `enforcing' the rules.33
But to say that we enforce the rules is only to say that we show
each other how to go on. It is not to say that we recite the rules to
one another.34 It is not because I
can recite my grammar or theory of meaning, or because I can be
cajoled into acknowledging rules of my language that I am able to
speak it. I can use language--speak and understand it--because I do
what I have been taught to do. I begin in imitation and end in
mastery.35
If we want to nod in the direction of implicit knowledge and say
that the rules are internalized--something I would advise
against--this can only mean that the child becomes part of the
linguistic community, and thus subject to all its criteria and
standards. It would be less misleading to say that it is the child
who is internalized by becoming a full-fledged member of the
community to which the practices belongs. This occurs when the child
learns how to act, not when he gains some or discovers some bits of
knowledge-that. But if this is so, then language, like counting and
measuring is a practice grounded firmly in knowledge-how--i.e., a C2
skill.36