TypeError: unbound method add() must be called with BinaryTreeinstance as first argument (got nothin

[Dan Stromberg]

I'm getting the error in the subject, from the following code:
def add(self, key):
"""
Adds a node containing I{key} to the subtree
rooted at I{self}, returning the added node.
"""
node = self.find(key)
if not node:
node.key = key
# placeholder
node.left, node.right = self.__class__(parent=node),
self.__class__(parent=node)
return (False, node)
else:
if random.random() < 0.5:
print('node.left is %s' % node.left)
return BinaryTree.add(self=node.left, key=key)
else:
print('node.right is %s' % node.left)
return BinaryTree.add(self=node.right, key=key)

The above add() method is part of a BinaryTree(object) class, whose
subclass is RedBlackTree.

We need to explicitly call BinaryTree.add() with an explict self, to avoid
inappropriately calling RedBlackTree.add().; BinaryTree.add() is being
called with a RedBlackTree instance as self.

The debugging print and traceback look like:
node.left is 0 -1 red
Traceback (most recent call last):
File "app_main.py", line 51, in run_toplevel
File "test-red_black_tree_mod", line 328, in <module>
test()
File "test-red_black_tree_mod", line 316, in test
all_good &= test_duplicates()
File "test-red_black_tree_mod", line 194, in test_duplicates
tree.add(value)
File
"/home/dstromberg/src/home-svn/red-black-tree-mod/trunk/duncan/red_black_bag_mod.py",
line 919, in add
(replaced, node) = super(RedBlackTree, self).add(key=key)
File
"/home/dstromberg/src/home-svn/red-black-tree-mod/trunk/duncan/red_black_bag_mod.py",
line 376, in add
return BinaryTree.add(self=node.left, key=key)
TypeError: unbound method add() must be called with BinaryTree instance as
first argument (got nothing instead)

Why is it complaining that .add() is getting nothing, when node.left isn't
None? As you can see above the traceback, it's got a value represented by
"node.left is 0 -1 red".

python 2.x, python 3.x and pypy all give this same error, though jython
errors out at a different point in the same method.

Thanks!