Children's Attributions 9
Similarity positions (graphs 1 through 4) postulate that children initially use humans (or
some humans) as an analogical basis to understand God’s beliefs. At the very least, we can
distinguish two stances: a strong and a weak one. The strong stance is perhaps best represented
in Piaget’s work (1960). There are two possible interpretations of Piaget’s understanding of the
development of God concepts. In graph 1, an infallible parent (who is capable of knowing what
is inside the box without having to see it) is used as the basis to understand God until quite late in
development. At some point, children start to recognize that parents can entertain false beliefs
but they do not transfer this characteristic to God, since at this same point they start to learn that
God has special qualities such as omniscience. For example, children would initially say that
both agents believe that rocks are inside the box, then, only by age 7, they would start to say that
humans believe that crackers are inside the box, and God believe that rocks are inside.
Conversely, in graph 2, a normal human being is used as the basis to understand God until quite
late in development. Then, children start to learn that God possesses certain special
characteristics that set God aside from common humans.
The weak stance postulates that children initially use humans as a basis to understand
God’s beliefs but start to differentiate them earlier in development than Piaget postulated -
before reaching the age of seven. In other words, we are envisaging the possibility of Piaget
being wrong simply in terms of the onset of the differentiation. In graph 3, an infallible human is
used as a basis to understand God. In graph 4, a normal human is instead used as the basis. This
explains why both the human and the God line stay flat for some time in the first instance, and
climb initially in the second. Although these positions are not well established in the literature,
they are possibilities that one should consider when dealing with cross-cultural data. For