Quantitative versions of content analysis place an emphasis on validity and reliability, and both of these concepts require another variable that correlates with the one you have measured. In the case of validity, this should be something related construct validity, such as discriminant or convergent validity, which could be some variable other than the data from a second coder.

Reliability, however, requires two separate measures of the same thing, so I cannot see you could assess reliability without a second coder. You could, however, claim that the coding system has such a high degree of "face validity" that a second coder is not necessary.