Political parties in Dahomey have received little attention from the scholarly community. This is an unfortunate fact because Dahomey's political parties have had an interesting development since at least 1949. Thus, first hand observation of a political party system moving through time from infancy to its present state has been forfeited. Although it would be difficult to point to any one reason for this neglect, it appears that it is a result of a general scholarly bias towards the study of only those political parties in countries where party decision making and party activities seemingly have a significant effect upon the political system.

For a scholar wanting to do a cross-cultural study of political parties the above sounds a tragic note. With the party as the unit of analysis, it means that the scholar is left with a universe of political parties with many incompletely defined members. Moreover, it becomes difficult to make the important comparisons and empirical generalizations necessary for such a cross-cultural study. Emphasis tends to be towards an exhaustive study of the "important" parties to the exclusion of others.

We find in the literature surveyed for Dahomey, that few propositions on political parties are presented and that no empirical data beyond election results is advanced. We do not even find a single case study devoted to Dahomey's political parties. Instead, we find that the political party becomes synonymous with the party leader and, if that party is in control of the government, the leader becomes synonymous with the government.

After evaluation of all the available literature on political parties in Dahomey during 1950-62, 434 pages were indexed from 19 documents for inclusion in this bibliography. A total of 124 pages from 11 documents was in English and 310 pages from 8 documents were in French

A look at the table of index codes and their frequency of application will reveal the strengths and weaknesses of our information file. The distribution of the codes is largely self-explanatory, but a few comments upon their application may be in order.

The second most frequent indexing code is "party leaders and officials" (code 360) and appears on 67 pages. One might argue that this reflects the extreme importance of the party leaders for political parties in Dahomey. However, it seems that the abundance of substantive literature on the party leaders is both a function of a particular leader's role and a function of little information on political parties. This becomes clearer when one considers that the most frequent indexing code is "government structure and political history" (code 680) which has been assigned to 87 pages. The literature tends towards brief accounts of political history of news events. The use of leaders such as Apithy and Maga, upon whom information is readily available, in explaining the party process may have been easier than researching actual party goals, activities, and organization.

One can also get a good idea of the usefulness of the literature by examining the frequency distribution of indexing codes combined into the nine major categories. We find that codes for the "party system" (8--) are used more frequently than any other set. One should keep in mind that for Dahomey this reflects election results, coalition formations, and history of the party system. All of this makes interesting political history but does little towards revealing the processes involved. Much the same could be said for categories of "political environment" (6--), "social and economic environment" (7--), and "party origins" (1--), insofar as they are handled in the literature for Dahomey.

The two categories, "party activities" (2--) and "party goals" (5--) are somewhat helpful. However, they tend to be associated more with the party leaders, by way of code 360, than with the actual political parties. Reliance on this particular code combined with code 350 (party members in government posts) explains the prominence of the "party composition" codes (3--) in the literature. Unfortunately for the purposes of the ICPP Project, the "party organization" codes (4--) are used only 34 times, and there are even fewer references to the "study of parties" (codes 0--).

Aside from outlining the major characteristics of the literature indexed, we should devote some attention to the gaps shown by the less frequently used codes. "Party activities" codes (2--) were seldom evoked in indexing. Emphasis in the literature on party leaders produced information on activities such as "influencing government policy" and "party discipline," but "raising funds" and the "social activities" of the party are hardly mentioned.

Under the "party organization" category, 3 out of 10 codes were barely used. The literature is so deficient that the indexer has not been able to formulate an adequate idea of what Dahomey's political party organization is like. In the "party goals" category (5--), only "issue orientation" (530) receives much attention. Here again, this is more a product of the party leader than it is of the political party.

Political parties appear to attract the attention of scholars interested in Dahomey's politics. However, the political party as a unit of analysis quickly recedes from the scholar's attention as soon as description and explanation begin.

From the data quality report of 19 documents listed in the second table some further observations can be made. We find that most of the literature is written by academics, based on secondary sources or none at all, has no quantification or empirical propositions, is not footnoted and is considered of medium quality overall.

In evaluating the tables of indexing and data quality codes, one can say that the state of the literature for the time period 1950-62 is such that research work on all levels of analysis is needed. Data obtained from primary sources and first hand observations would be invaluable. Some works published since the bibliographic research was completed in 1968 have improved the state of knowledge materially (see, for example, Dov Ronen's Dahomey), but these recent publications do not figure in the above evaluation.