According to Davidson, this polarity was structured along a single axis opposing mass merchandising and highly specialised boutiques with "conventional and often non programmed single line stores" in between. Such a polarity could be observed for every product line. Reducing the perceptions of store types to a single dimension proved too limited, and Tigert refined the polarity concept by structuring it along two dimensions: small stores vs. warehouse technology on one hand, narrow product line vs. mass merchandising on the other hand. As for Davidson, such a polar structure of stores could be observed for every product line. Both authors provided only a theoretical framework for the polarity concept, but did not attempt to validate it empirically. Thus we should talk in both cases of a concept of normative polarity, i.e. perceptions of store types by customers as they are perceived by an expert. This scheme proves highly stimulating for a deeper investigation of the polarity concept, that would not be based on normative appreciations anymore, but would be derived from customer perceptions of store types and competing stores inside each type.