If codes were simply a matter of strong convention, we as analysts of political advertising would be faced with a simple task. It would be theoretically possible to collect the code's correlation between signifiers and signified and construct a kind of dictionary for all the nonlinguistic codes of the political ad. A survey of an electorate could be used to determine the semantic space occupied by each sign (e.g., lexical multidimensional space as in Osgood, Suci, & Tannebaum, 1957). There could be a kind of Rosetta Stone of political codes. Armed with this knowledge, an analyst would simply decode the ad by looking up and summing the semantic units that correspond to the signs in an ad.