Practices

In ``Interpretation and Reality: Two Queries for Krausz,''¹⁷ Bernard Harrison and I use linear measurement to show how practices figure in the mastery of a skill and its rules.

...what will [teaching someone the practice of linear measurement] involve? Teaching him to carry out measuring operations with rulers and measuring-tapes? Not only that. Someone who knew only how to go through the motions of measuring, even if he had been trained to say aloud or otherwise note the numbers corresponding to gradations on the yardstick, would not understand what measurement is. That person would have been converted into a kind of measuring-device himself: a tool, an element in the practice of linear measurement, not its master. To be a master of the practice ...that person needs to understand the point of the practice: on the one hand, the purposes it serves in our lives, and on the other, what it is about it that enables it to serve those purposes. Only then will the learner be able to deploy the practice in the service of purposes of that type of his own.
¹⁸

Building on this, the following is a rough attempt at defining a practice.

1   There must be criteria, or rules.

1.1  
The agent's behavior must accord with the criteria.

1.2  
The agent is accountable for his behavior's being in accordance with the criteria. In other words, the agent's behavior is subject to correction and modification.

1.2.1  
The behavior must be teachable and learnable.

1.3  
The criteria must be public.

2  
There must be a point to the activity.

3  
The agent must understand the point, and be able to use the practice for his own purposes.

C2 and C3 differ from C1 because the activities in C2 and C3 fall under practices. A skill is an activity which has an associated practice. Therefore, C2, but not C1, concerns skills which may, for Dummett's purposes, be usefully compared to the skills in C3, which are skills with an explicitly theoretical component.¹⁹

But, one may wonder whether I have left something out of my analysis of practices. Is there, it might be asked, an additional requirement along the following lines?

4   The agent must know (explicitly or implicitly) the rules, and be guided by them in the strong sense that she acts on them, i.e., goes through the double operation of considering and executing.

If we accept 4, then C2 collapses into C3,²⁰and Dummett's case is complete. Can I avoid this? I argue that the cost of adding 4 is simply too high--and too high for Dummett himself. It would require that we ride roughshod not only over ordinary intuitions about differences between various sorts of skills, but over Dummett's stated views, viz., that some skills are merely practical, while others are something more. In the next section, I show why 4 is not, in general, a condition of practices or skills.