1.8 Methods
1.8.2 Data analysis methods
The collected names have been analysed in various ways depending on the article. It must be emphasised that the line between data collection and analysation methods is often blurred. For example, above discussed topic of compiling etymologies for names could belong to the analysis methods as well.
Locating the chosen names in the map is a method that is implemented in each article. The aim of this method is to determine where the name was used.
Most of the name bearers are linked to some specific places, e.g. villages or towns. This applies particularly to documents produced for fiscal purposes.
Medieval letters and court documents do not always indicate origins of persons mentioned and therefore, more background work must be done to determine, where a name bearer lived or originated from.
Villages and towns where the bearers of collected names resided can often be located with sufficient accuracy based on both old and contemporary maps.
Again, onomastic literature, such as those presented above in section 1.8.1., has been helpful in situations where a settlement name searched for does not exist anymore. Most of the locations mentioned in sources concerning modern Estonia, Finland and Karelia can be recognized and placed on a map but
Russian sources contain cases that are more difficult. It follows that some of the locations have remained unresolved despite all efforts. Names with no reference to any location are excluded from the maps.
The distribution of studied names and name elements is further analysed mainly by two comparison methods. The first one is called in onomastics as typological-geographic research, which means analysing the emergence and distribution of names or name elements (Ainiala et al. 2012: 48). In Article I, this method is used to examine if the village names studied have parallel cases outside the study area and accordingly, to indicate those regions that the study area had connections with before and during the Middle Ages. This method is implemented in each article, but in the first four the comparison is qualitive, i.e. assessed visually. In Article V, name data is computationally clustered to determine regional patterns in personal name usage.
The second comparison method involves comparing the distribution of chosen names to the results of other scientific fields such as archaeology and historical linguistics. This way, it is possible to visualize how the distribution of certain pre-Christian Finnic personal names is related to those archaeological finds that are regarded as evidence of Late Iron Age settlement.
This method is shortly implemented in Articles I and V, whereas, in Articles III and IV, one of the main focuses is the archaeological aspect.
Morphological and semantical aspects of names are briefly analysed in each paper, but only in Article III they have their own section. Within the scope of the present study, morphological analysis considers three factors: name elements (i.e., personal name lexemes and personal name suffixes); analysis of how these elements were used to form personal names; and analysis of the popularity of specific name elements according to the collected name data.
Only pre-Christian names are analysed from the point of view of morphology.
The semantical analysis of the names studied entails the investigation of the meanings of singular personal name lexemes (e.g. Lempi ‘love, favorite’).
The semantic meanings of singular pre-Christian name lexemes are mostly based on definitions presented in onomastic literature. Even with the methods of historical-comparative linguistics, it is impossible to thoroughly determine what the name elements under study meant centuries ago. Furthermore, analysing the perceptions of name bestowers towards the name they gave is an aspect closely linked to semantics as well. This perspective is particularly relevant to foreign names since they were not semantically transparent to Finnic speakers. For example, Article II indicates that most of the popular medieval female names were names of well-known saints and Scandinavian queens. This indicates that the name givers wished their child to have the same qualities as the former bearer(s) had.
Statistical methods are implemented in each article. In papers I–IV, these methods are used to conduct simple statistics of the names studied. The last publication (V) is based on more advanced statistical techniques:
computational methods are utilized to find regional differences in the use of personal names. Measurements are performed by two different distance measuring and clustering procedures: the Jaccard distance metric with the average linkage and the Euclidean distance metric with the ward linkage.
2 ARTICLES AND THEIR MAIN RESULTS
The articles included in this dissertation concentrate on pre-Cristian Finnic personal names and early Christian Finnic names from different perspectives.
The most substantial differences are in the areal focus and the sources used.
The methods applied vary from article to article and especially in Article V, new ways to study ancient personal names are utilized. Publications are presented in order of completion, since they focus on the same time-period, that is approximately 1200–1550 AD.
Summaries are structured in the following manner. First, an overview of the aim of the study is given together with short descriptions of sources and methods used. Second, the main results of the study are presented. I will also make retrospective remarks on the individual papers where needed.