Variable Names

You have already seen several programs which contain an anonymous variable, i.e. a variable which either is given as an underscore _, or whose name begins with an underscore (e.g. _Tail). If you consult and then list a program which contains such a variable, SICStus shows an underscore only in both cases.

Anonymous variables is something which you have a reason to use quite often, as it is a Prolog convention to use a name beginning with an underscore for variables whose value is not used in the clause in question.

Prolog is a declarative language, whose control structure relies on the unification of heads and goals. That is, facts and rules are accessed by matching predicates and variables. When the goal of a query matches a head of a rule, the variables in the head are instantiated (or if the query is called with a variable, the variables of the head are unified with the variables in the query). The goals of the rule are called with these variable bindings. The most important function of variables is thus that they pass on values or bindings.

When we write a procedure, i.e. all the facts and rules which define a particular predicate, it is not always the case that each argument, or part of an argument, is used by every individual clause. This was seen by the list/1 procedure on handout 7:

list([]).
list([_Head|Tail]):- 
         list(Tail).

{Warning: [Tail] - singleton variables in member/2 in lines
14-18}

It is good programming style to use an underscore followed by a name rather than just an underscore when you write the program. The reason for this is that by chosing your variable names carefully, you make your programs easier to read.

Remember the ancestor predicate from the chapter on Recursion.

ancestor(Old,Young) :- 
        parent(Old,Young).
 
ancestor(Old,Young) :-
               parent(Old,Middle),
               ancestor(Middle,Young).